Research ArticleAn Analysis of Factors Affecting Flowslide Deposit MorphologyUsing Taguchi Method

Zhao Duan12 Yan-Bin Wu 12 Hao Tang12 Jian-Quan Ma 12 and Xing-Hua Zhu3

1College of Geology and Environment Xirsquoan University of Science and Technology Xirsquoan 710054 China2Shaanxi Provincial Key Laboratory of Geological Support for Coal Green Exploitation Xirsquoan 710055 China3College of Geological Engineering and Geomatics Changrsquoan University Xirsquoan 710064 China

Correspondence should be addressed to Yan-Bin Wu 19209071021stuxusteducn

Received 27 May 2020 Revised 28 June 2020 Accepted 7 July 2020 Published 27 August 2020

Academic Editor Haiyun Shi

Copyright copy 2020 Zhao Duan et al is is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work isproperly cited

Flowslides as one type of landslides are becoming a research hotspot due to their high speed and long runout distance which cancause tremendous damage and economic losse scale of damage and deposit morphology of flowslide is closely related to factorslike deposit volume slope height and slope angle In order to assess the influence of these factors a sandbox apparatus isdeveloped and the Taguchi method is used to design an experimental scheme to analyze the results of factors affecting the depositmorphology of flowslide e results show that the factor that has the greatest impact on flowslide deposit morphology is slopeangle followed by the influence of volume and slope height As slope angle increases the maximumwidth maximum length areaand length-width ratio of the deposit first increase and then decrease In addition there should be a critical angle in the changes ofdeposit morphology that is between 60deg and 70deg under the experimental conditions When the volume is 54times10minus 3m3 the slopeangle is 70deg the slope height is 090m and the changes of deposit morphology of the flowslide are the largest In this studyconsidering the slope angle as a single variable there is a single upheaval for a slope angle of 40deg and 50deg and a double upheaval at60deg and 70deg e formation mechanism of the upheaval is analyzed based on the Mohr-Coulomb criterion and consideredproperties of the material e apparent friction coefficient of a flowslide is spatially and lithologically different and increasesnonlinearly as the slope angle increases e initial benchmark of the slope angle and apparent friction coefficient curve areaffected by the friction coefficient of the material the position of the inflection point at which the curve increases rapidly is affectedby the coefficient of velocity restitution

1 Introduction

Landslides including slide and flowslide are often seen in theLoess Plateau of northwest China since the loess featuresmetastable microstructure high porosity and water sensi-tivity e hazards associated with flowslides are oftenclosely related to volume slope angle slope height and soon because these conditions significantly affect a flowslidersquosmoving distance [1ndash4] Flowslides are becoming a researchhotspot recently since it has the characteristics of high speedand long runout distance as well as leading to tremendousdamage and economic loss A large number of field in-vestigations have shown that the runout and area of a

flowslide increase with an increase in volume and slopeheight but decrease with an increase in slope angle [5ndash9] Infact each flowslide is unique and nonrepeatable and as aresult the characteristics and parameters of each are dif-ferent erefore in field research it is impossible to sep-arate out the various influencing factors for single factorresearch Similarly data obtained through field investiga-tions cannot accurately reflect the impact of various factorson the characteristics of flowslide movement Tomake up forthe shortcomings of field investigations physical model testswere introduced to study the deposit morphology andmovement rules of flowslides [10ndash12] In the course of theseinvestigations some rules and interesting characteristics of

HindawiAdvances in Civil EngineeringVolume 2020 Article ID 8844722 14 pageshttpsdoiorg10115520208844722

deposits that were in line with reality were reproduced wellby model tests ese modeling studies provide a referencefor understanding the dynamic characteristics of flowslidese rationality of experimental method directly determinesthe accuracy of experimental results [10 13ndash15] It is pre-vailing to study the runout distance influencing factorsstress during movement and other characteristics of flow-slides using model experiments [11] However in the re-search involving the field of flowslides currently the methodis relatively monotonous and those studies are conducted tothe movement and deposit characteristics under the influ-ence of a single factor [4] In addition when the order ofinfluencing factors was obtained a deep excavation on thechange laws of deposit morphology and the possible internalmechanical mechanism under the greatest influencing factorhas not been performed

e flowslide flows to the slope break where a change inits flow direction will inevitably cause collision and shearingthe intensity of this effect increases with an increase in slopebreak angularity and sharpness [13] e duration of aflowslide increases with angle and volume and decreaseswith aspect ratio [16] In the study of the deposit mor-phology of flowslides scale effects would affect the overallshape of the deposit including leading edge highest pointand trailing edge positions but could be minimized byaccurately controlling the test variables [17] Longitudinalridges developed on the surface of flowslide deposits causedby radial propagation of the flowslide during movement asdiscussed by Dufresne and Davies [18] e properties ofgranular materials are also an important factor affecting thecharacteristics of flowslide deposits including the granulardiameter the grading index and the roundness Differentgranular materials can lead to different granular effectswhich can increase the accumulation area and runout dis-tance [19 20] e apparent friction coefficient is an im-portant parameter used to characterize the movement of aflowslide which has been widely studied under differentvariable conditions [5 7 9 21] It is generally believed thatthe apparent friction coefficient decreases with increasingvolume and increases with increasing slope angle but therelated general relationship has not been proposed [3 22]Flowslide on alien planets shares the above-mentionedcharacteristics and rules [23ndash25] but shows longer runoutAt present research methods used in the study of flowslideare limited and research schemes are not systematic enoughIn addition better knowledge of the overall rules governingchanges in and causes of flowslide deposit morphology isneeded

In this paper the Taguchi method [15 26ndash29] is used todesign an experimental scheme and analyze the results offactors affecting the deposit morphology of flowslide eTaguchi method can be used to quantify test results and toobtain the combination factors that cause the greatestmorphological changes in the deposits under test conditionse specific objectives of this study are to (1) determine themaximum influencing factors of flowslide deposit mor-phology and the combination of factors that cause the largestchange in the morphology of a deposit using the Taguchimethod (2) analyze the changes of the surface and profile

morphology of the deposit under the most influential factorsand compare them with actual flowslides (3) explain therules governing changes in flowslide deposit morphologyand carry out the necessary stress analyses for particularmorphologies (4) obtain and confirm the distributioncharacteristics of the apparent friction coefficient by com-bining test results with natural flowslides

2 Material and Method

21 Sandbox Experiment Setup A sandbox experiment wasconducted to study flowslide deposit morphology Plexiglasswas chosen for the experimental devices which werecomposed of four parts a sand container an inclined plate ahorizontal plate and high-speed camerase length of boththe inclined plate and the horizontal plate is 15m and thewidth of each is 12m (Figure 1(a)) e specification vol-umes of each sand container with a gate in the bottom are18times10minus 3m3 36times10minus 3m3 and 54times10minus 3m3 e angle ofthe inclined plate can be adjusted freely through a rearbracket and the center gravity height of the sand containercan be adjusted through a track on the inclined plate High-speed cameras (120 framess) were used to record imagesduring the experiment

During the experiment medium-fine sand density wascontrolled to 151 kgmiddotmminus 6 e material flowed down an in-clined plate with no boundary constraints to the horizontalplate when the gate was opened When the sliding materialstopped moving a millimeter-level three-dimensional scannerwas used to collect digital elevation model (DEM) data of thedeposit surface (Figure 1(b)) through calibrated and extractedkey point coordinates to obtain flowslide deposit morphologydata

22 Material A medium-fine sand (Figure 2) was used asthe sliding material in the simulated sandbox experimentbecause it shows a similar flow characteristic during slidingcompared with natural flowslides [30 31] its particle-sizedistribution curve is shown in Figure 3 e uneven coef-ficient Cu is 239 the curvature coefficient Cc is 119 theaverage diameter is 2times10minus 4m and the specific surface area is002m2∙kgminus 4 e cumulative percentage of the particle sizein the range 0075ndash05mm is 8771 the internal frictionangle φ is 36deg the cohesion c is zero and the interface frictionparameter of plexiglass and sand is 0538

23 Method e experimental scheme was designed usingthe Taguchi orthogonal method (Table 1) to comprehen-sively analyze the influence of flowslide volume slope angleand slope height on deposit morphology e followingdesign variables were used to assess influencing factorsvolumes 18times10minus 3m3 36times10minus 3m3 and 54times10minus 3m3 slopeangles 50deg 60deg and 70deg and slope heights 07m 09m and11m

e principle of variable selection was as follows AK-means clustering of volume and slope height data from 28flowslides [30] in the Jingyang South platform Shaanxiprovince China was conducted Results showed that the

2 Advances in Civil Engineering

deposit volume was divided into three levels of131775times103m3 54544times103m3 and 18482times103m3 re-spectively so the similarity ratio was approximately 1 2times108 Slope height was divided into three levels of11345m 8427m and 6307m with a similarity ratio ofapproximately 1 100 e slope angle of the Jingyang Southplatformwasmainly within 40degndash70deg [31]ere were in totalnine groups of experiments two parallel experiments werecarried out for each group If the experimental results weregreater than millimeter-level in difference a third experi-ment was conducted eventually taking the average value oftwo small differences in the results to determine the depositmorphology of the flowslide

3 Results

31 Sandbox Experiment Results e Taguchi method is atest method commonly used in steel casting and other fieldswithin materials science to maximize production and re-duce consumables Its greatest advantage is to determinethe order of significance of evaluation indexes using fewertests and to determine an optimized combination schemeof influencing factors [15 26 27 29] e direct basis forthe analysis of the results is the signal-to-noise ratio (SN)which can be divided into three categories according tothree characteristics larger-the-better smaller-the-betterand nominal-the-better (washing the test results closest tothe target value) e larger the value of (SN) the greaterthe signal that the dependent variable responds to thechange of factors erefore the larger-the-better char-acteristic is used in this paper and the calculation ex-pression is as follows

S

N1113874 1113875

i minus 10lg

1n

1113944

n

j1Y2ij

⎛⎝ ⎞⎠ (1)

15m

12m

Sand container

Inclined plate

15m

High-speedcamera

(a)

Sample deposit

3D scaner

(b)

Figure 1 Flowslides experimental research design (a) structural of the experimental device (b) 3D scanner for acquisition of deposit data

Figure 2 Experimental material

0001 001 01 1

0

20

40

60

80

100

Massachusetts Institute of Technology (MIT) classification

CoarseMediumFineCoarseMediumSand

Perc

ent f

iner

by

wei

ght (

)

Particle diameter (mm)

SiltFine

Figure 3 Particle-size distribution curve for the experimentalmaterial

Advances in Civil Engineering 3

where (SN) denotes the signal-to-noise ratio n denotes therepeatable time of each variable combination Y1j denotesthe observed data of evaluation index i denotes a differentevaluation index and j is the test number from 1 to n

It is necessary to establish an evaluation index to analyzethe influencing factors of flowslide deposit morphology bythe Taguchi method because the deposit morphology offlowslide (Figure 4(a)) can be characterized by five elementsmaximum length (Ld) maximum width (Wd) maximumthickness (Dd) area (A) and length-width ratio (LdWd)

denoted by Y1j Y2j Y3j Y4j and Y5j respectively e fiveindexes used as evaluation indexes along with statisticalresults and signal-to-noise ratio calculation results areshown in Tables 2 and 3

It can be seen from Table 3 that when the volume is54times10minus 3m3 the slope angle is 70deg the slope height is 09mand the signal-to-noise ratio values of Ld Wd and A are thelargest 4055 4049 and minus 06 respectively erefore underthe influence of this combination the change responses ofLd Wd and A are the largest It can also be seen that whenthe volume is 54times10minus 3m3 the slope angle is 50deg the slopeheight is 11m and the change response of Dd is the largestWhen the volume is 18times10minus 3m3 the slope angle is 60deg theslope height is 09m and the change response of (LdWd) isthe largest Based on a comprehensive comparison of acombination of influencing factors A3B3C2mdashnamely whenthe volume is 54times10minus 3m3 the slope angle is 70deg and theslope height is 09mmdashthe change response of the flowslidedeposit morphology is the largest

32 Morphology of the Flowslide e calculation results ofthe Taguchi method and multivariate analysis of varianceboth indicate that slope angle has the greatest influence onthe deposit morphology of flowslide To further analyze themorphology of flowslide under the change of slope angleonly an experiment using four sets of sandboxes with aconstant volume of 36times10minus 3m3 a constant slope height of09m and slope angles of 40deg 50deg 60deg and 70deg was carriedout (Figure 4(b))

(1) Analysis based on Taguchi method According to thesignal-to-noise ratio values of different combinationsdetermined using equation (1) as listed in Table 3 the

values of the maximum minus minimum (extremedifference) of the signal-to-noise ratio between thethree levels under the same factor are obtained(Table 4) and shown in Figure 5 e bigger theextreme difference value of the influencing factorsthe greater the influence of this factor on the results[26 28]e results show that the values of extreme differenceof LdDdA and (LdWd) are 277 551 454 and 17respectively e largest value reflects the influence ofslope angle as compared with other factors e nextlargest value of extreme difference reflects the influ-ence of flowslide volume e value of extreme dif-ference as influenced by slope height is relatively thesmallest e value of Wd is most affected by volumethe corresponding value of extreme difference is 278In summary the most significant factor affecting thedeposit morphology of flowslide is slope angle fol-lowed by volume and slope height

(2) Multivariate analysis of variance To further verifyand analyze the impact of flowslide volume slopeangle and slope height on deposit morphology amultifactor analysis of variance is conducted on fiveevaluation indexes influenced by the three factors(Figure 5) the statistical results are shown in Table 5Partial η2 (the ratio of the within-group variance tothe overall variance of an independent variable)indicates the significance of the effect caused by theinfluence factor where the larger the value the moreobvious the effect of the influence factor e resultsof multivariate analysis of variance are consistentwith the calculation results of the Taguchi methode values of Ld Dd A and (LdWd) correspondingto the biggest partial η2 are 064 079 056 and 073all as a result of the effect of slope angle while forWd the biggest value is 087 which is the effect ofvolume of the flowslide Similarly the value of partialη2 under the influence of slope height is relatively thesmallest among the five evaluation indexes whichare 01 024 045 006 and 005 respectivelyerefore slope angle is the factor that has thegreatest influence on the deposit morphology offlowslide followed by volume and slope height

Table 1 Orthogonal experimental scheme

Number Orthogonal combination Flowslide volume (10minus 3m3) Slope angel (deg) Height (m)1 A1B1C1 18 50 072 A1B2C2 18 60 093 A1B3C3 18 70 114 A2B1C2 36 50 095 A2B2C3 36 60 116 A2B3C1 36 70 077 A3B1C3 54 50 118 A3B2C1 54 60 079 A3B3C2 54 70 09Note A B and C denote flowslide volume Slope angle and height respectively

4 Advances in Civil Engineering

33 Planar Morphology at Different Angles After correctingthe digital elevation model (DEM) data obtained by thescanner the three-dimensional coordinates of the flowslidedeposit are extracted and the contour maps of the surfacemorphology are drawn (Figure 6) At smaller slope angles of40deg and 50deg the deposit surface is relatively flat its long axis

lies along the width of the flowslide and it is generallytongue-shaped At 40deg the length width length-width ratioand area are 053m 070m 076 and 029m2 respectivelyAt 50deg the length width length-width ratio and area are089m 080m 111 and 044m2 respectively while at largerslope angles of 60deg and 70deg the deposits tend to be thickest at

Flowslide deposit

A

Ld

Dd

Wd

L

H

θ

(a)

H

θ1θ2

θ3

(b)

Figure 4 Schematic illustration of devices (a) e flowslide deposit characteristics Ld Maximum sliding length of flowslide deposit on thesurface Wd Maximum width of flowslide deposit Dd Maximum depth of flowslide deposit A Area of flowslide deposit projected onhorizontal plane θ slope angle HHeight of sandbox measuring from center of gravity L Total sliding distance of flowslide deposit (b) eprocess of center gravity constantly

Table 2 Statistics on flowslide deposit morphology

Orthogonal combination Y1jLd(10minus 2m) Y2jWd(10minus 2m) Y3jDd(10minus 2m) Y4jA(m2) Y5j(LdWd)(1)

A1B1C1 6542 6155 288 036 094A1B2C2 6833 7051 140 043 103A1B3C3 7463 7501 132 047 101A2B1C2 7333 5200 279 032 071A2B2C3 9413 9000 137 078 096A2B3C1 7927 6451 323 044 081A3B1C3 8812 6736 429 043 076A3B2C1 9508 8850 268 071 093A3B3C2 10576 10650 278 093 101Note Ld maximum sliding length of flowslide deposit on the surface Wd maximum width of flowslide deposit Dd maximum depth of flowslide deposit Aarea of flowslide deposit projected on horizontal plane θ slope angle (LdWd) length-width ratio

Table 3 e signal-to-noise ratio of elements of deposits morphology

Orthogonal combinationSignal-to-noise ratio (dB)

Ld (10-2m) Wd (10-2m) Dd (10-2m) A (m2) (LdWd) (1)

A1B1C1 3579 3631 919 minus 876 minus 053A1B2C2 3697 3669 292 minus 729 027A1B3C3 375 3746 241 minus 651 004A2B1C2 3432 3731 891 minus 991 minus 299A2B2C3 3909 3948 273 minus 21 minus 039A2B3C1 3619 3798 1018 minus 708 minus 179A3B1C3 3657 389 1265 minus 73 minus 233A3B2C1 3894 3956 856 minus 295 minus 062A3B3C2 4055 4049 888 minus 06 006Note Ld maximum sliding length of flowslide deposit on the surface Wd maximum width of flowslide deposit Dd maximum depth of flowslide deposit Aarea of flowslide deposit projected on horizontal plane θ slope angle (LdWd) length-width ratio

Advances in Civil Engineering 5

the front and back and are nearly round upheaval and washare arranged alternately on the surface At 60deg the lengthwidth length-width ratio and area are 090m 090m 10and 057m2 respectively At 70deg the length width length-width ratio and area are 074m 075m 099 and 042m2respectively e morphologies of the flowslides formed ateach of the four angles are symmetrically distributed aboutthe sliding central axis

Comparing the shape of the deposits formed at each ofthe four angles we can see that with an increase in slopeangle the values of Ld Wd A and (LdWd) for the flowslidedeposit first increase and then decrease e main reason isthat by increasing slope angle from 40deg to 60deg the sum ofkinetic energy loss by the sliding body due to friction andcollisional energy loss on the slope break is reducedHowever in increasing the slope angle from 60deg to 70deg thesum of the kinetic energy loss and the collisional energy lossincreases In the experiment the slope height was kept at aconstant value of 090m to each of the four angles so theinitial potential energy was the same for each (Figure 4(b))

e smaller the slope angle the greater the moving dis-placement of the sliding body on the inclined plate resultingin greater kinetic energy loss caused by friction When theflowslide flows to a slope break where there is a change in thedirection of motion to a horizontal direction and kineticenergy loss will inevitably occur kinetic energy loss ispositively related to the slope angle [32] Under the con-dition of changing angles the increase in the length of theflowslide deposit is significantly higher than the width isis explained by the feature of straightness [33] which theflowslide has during high-speed movement

ese laboratory findings can be examined in the contextof flowslides that occurred naturally In Xihetan Shaanxiprovince China a low-angle flowslide deposit formed at a45deg slope angle and had a length of 289m a width of 329mand a length-width ratio of 088 (Figure 7) Another low-angle deposit formed in Shenzhen China [34 35] on a slopeangle of 24deg with a length of 700m a width of about 550mand a length-width ratio of about 127 (Figure 8) A largeslope angle (more than 60deg) slide on the steep cliffs along thecoast of Northern Europe is shown in Figure 9e followingare shown in Figure 9(a) the deposit with a length 330m inFigure 9(b) the deposit with a length 150m in Figure 9(c)the deposit with a length 150m and width 120m inFigure 9(d) the deposit with a length 270m and width300m all of these deposits with a length-width ratio close to1 alternating upheaval and wash on the surface of thedeposit and an overall circular shape (Figure 9) [36 37] eactual flowslides at small angles and large angles are similarto the deposit morphology and state of motion formedduring the current experiment However due to the dif-ference in topography and the scale of the actual flowslides itis difficult to compare in terms of dimensional parameterssuch as the length and width of the deposit leaving thedimensionless length-width ratio surface morphology andoverall shape for comparison In both the natural and themodeled deposit morphology a flat surface and a generally

Table 4 e main factors of influence

Morphological elements of deposit Influencing factors Level1 (dB) Level2 (dB) Level3 (dB) Extreme difference (dB) Influence order

Ld (m)A 3675 3652 3869 216 2B 3556 3833 3808 278 1C 3697 3728 3772 075 3

Wd (m)A 3682 3825 3965 278 1B 3751 3858 3864 108 2C 3795 3816 3866 071 3

Dd (m)A 482 728 1003 521 2B 1025 474 716 551 1C 931 691 593 338 3

A (m2)A minus 752 minus 637 minus 362 390 2B minus 866 minus 411 minus 473 454 1C minus 627 minus 593 minus 531 096 3

(LdWd) (1)A minus 007 minus 172 minus 097 165 2B minus 195 minus 025 minus 561 170 1C minus 098 minus 090 minus 089 009 3

Note A B and C denote flowslide volume slope angle and height respectively Ld maximum sliding length of flowslide deposit on the surface Wdmaximum width of flowslide deposit Dd maximum depth of flowslide deposit A area of flowslide deposit projected on horizontal plane θ slope angle(LdWd) length-width ratio

A B C A B C A B C A B C A B C

0

1

2

3

4

5

6

Influencing factors

Extr

eme d

iffer

ence

(dB)

00

02

04

06

08

10

Ld Wd Dd LdWdA

Patr

ial η

2

Figure 5 Statistical results using Taguchi method and multivariateanalysis of variance

6 Advances in Civil Engineering

tongue-shaped deposit form at low angles while alternatingupheaval and wash on the deposit surface and a generallycircle-shaped deposit form at large angles

34 Profile Morphology at Different AnglesMultidimensional analysis of the morphology of flowslidedeposit can help researchers gain a deeper understanding ofmorphology characteristics Figure 10 shows the profiles ofthe four flowslides depicted in Figure 6 Combining theprofile (Figure 10) with planar morphology (Figure 6) it canbe seen that at smaller slope angles of 40deg and 50deg a generallysingle upheaval with heights of 0027m and 0018m can berecognized with corresponding peak positions of 028m and050m respectively although the surface of each deposit isslightly undulated e position of the center of gravity isconsistent with the peak position of the upheaval At largerslope angles of 60deg and 70deg the surfaces of the deposits areundulated and there is double upheaval e maximumheight differences between adjacent upheaval and wash are0005m and 0023m respectively e heights of the frontupheaval are 0009m and 0026m with corresponding peakpositions of 067m and 040m respectively e heights ofthe back upheaval are 0012m and 0035m with corre-sponding peak positions of 010m and 006m respectivelye heights of the middle wash are 0006m and 0011m withcorresponding positions of 041m and 022m respectivelye positions of the center of gravity are relatively consistentwith the position of the middle wash

Comparing the four profiles shows a single upheavalshape at smaller angles of 40deg and 50deg and a double upheavalshape at larger angles of 60deg and 70deg In the flowslides withdouble upheaval the height of the front upheaval is smallerthan that of the back upheaval As the slope angle increasesthe average thickness of the deposit first decreases and thenincreases and the position of the center of gravity firstadvances and then retreats Based on the above comparisonit can be considered that there should be a key angle in thechange laws of flowslide deposit morphology If the slopeangle is less than or larger than the key angle under the sameconditions the length and width of the deposit will decreasee key angle is between 60deg and 70deg under the experimentalconditions in the paper However actually the key angle offlowslides should be different values depending on thematerial of the sliding body and the terrain conditions

e profile figures of chalk flows [36] on the steep cliffs ofthe European coast show similarities with the profilemorphology of the sandbox experiment for the larger anglesespecially the feature of double upheaval (Figures 11 and 12)ese two deposits shown in Figures 11 and 12 are 320mand 200m in length respectively It can be seen from thepicture that the leading edge of the flowslide deposit is flatand the surface shows double upheaval e front upheavalis lower than the back upheaval e height difference be-tween the middle wash and the back upheaval is greater thanthat of the front upheaval e center of gravity is locatedslightly in front of the middle wash It is noticed that there isa difference although the morphology of the double

Table 5 Variance analysis statistics

Morphological elements of deposit Factors Quadratic sum DOF Mean square error Partial η2

Ld (m)

A 68628 2 34314 055B 98732 2 49366 064C 597 2 2985 01

Error 56458 2 28229Total 229787 8

Wd (m)

A 108303 2 54151 087B 22448 2 11224 059C 4897 2 2449 024

Error 15835 2 7917Total 151483 8

Dd (m)

A 289 2 144 076B 342 2 171 079C 073 2 037 045

Error 091 2 045Total 795 8

A (m2)

A 011 2 006 052B 013 2 007 056C 001 2 0 006

Error 011 2 005Total 036 8

(LdWd) (1)

A 004 2 002 069B 005 2 003 073C 0 2 0 005

Error 002 2 001Total 011 8

Note Ld maximum sliding length of flowslide deposit on the surface Wd maximum width of flowslide deposit Dd maximum depth of flowslide deposit Aarea of flowslide deposit projected on horizontal plane θ slope angle (LdWd) length-width ratio

Advances in Civil Engineering 7

upheaval is the same e center of gravity of flowslidedeposit is located in the middle wash in experimentalconditions However the center of gravity is located in frontof the middle wash in actual flowslides e possible reasonmainly is the difference of properties of granular materialsbecause the movement and accumulation process of flow-slides are also affected by internal factors such as the particlesize and roundness of sliding body materials [14]

35 Mechanical Analysis of the Deposit MorphologyBefore the mechanical analysis the properties of themedium-fine sand and plexiglass should be noticed as abasis for this analysis e friction coefficient 0727 of themedium-fine sand is larger than 0538 of the interface of

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

aprime

0

6

0

6121812

918120 6

(a)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

bprime0 4

8

80

412

0

04812

4

(b)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

cprime

0 46

28

8

9

9

68

4

2

02

6

84

0

20 468

8

10

8

(c)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

dprime

04

1216

20

242016128

24

16 12

048

04816

12

08

4

(d)

Figure 6 Contour map of flowslides topography at different slope angles (a) (b) (c) and (d) corresponding to 40deg 50deg 60deg and 70deg respectively

N

Figure 7 Xi Hetan flowslide in China (34deg29prime32PrimeN 108deg50prime13Prime E)

N

Figure 8 Shenzhen flowslide Source Aerial image form Sou-hucom in China (22deg43prime04Prime N 113deg56prime31Prime E)

8 Advances in Civil Engineering

the plexiglass with the medium-fine sand e differenceof the friction coefficient between the interface and thesliding body itself also existed in actual flowslides becausethe interface will be lubricated by water or the liquefactioneffect [30 38] erefore the sliding body tends to flowalong the interface of the plexiglass and the sand forms athin sand layer like an erodible bed before the main slidingbody arrived on both inclined and horizontal plates[39 40] It provides an essential condition of the shearaction Figure 13 shows the cross-sectional state of thesliding body at three phases in which phase 3 indicatedby a solid line represents the final deposit state of the

N

(a)

N

(b)

N

(c)

N

(d)

Figure 9 Topography of typical flowslides (a) and (b) located in Folkestone Warren (51deg05prime03Prime N 1deg20prime36Prime E) (c) and (d) located DoverKent (51deg07prime53Prime N 1deg20prime36Prime E) Source from Hutchinson [36]

0 200 400 600 800 10000

10

20

30

40

Dep

th o

f dep

osit

(mm

)

Length of deposit (mm)

40degprofile50degprofile

60degprofile70degprofile

Figure 10 e deposit profiles of flowslides at different angles

N

Front upheavalFront upheavalBack upheavalWash

Figure 11 Flowslide occurred in May 1981 at Le Chien Neuf(49deg46prime59Prime N 0deg25prime19Prime E) Source from Hutchinson [36]

N

Front upheavalBack upheaval

Wash

Figure 12 Flowslide occurred in May 1958 at Wissower KlinkenRugen Germany (54deg32prime09Prime N 13deg40prime42Prime E) Source fromHutchinson [36]

Advances in Civil Engineering 9

sliding body In this paper the Mohr-Coulomb criterion isused in the analysis of the formation of single and doubleupheaval as follows

τ c + σ tanφ (2)Taking the numbers① and② as two microelements in

the sliding body under a stress condition whereby themaximum principal stress σ1 of the microelement at theslope break is parallel to the direction of the inclined plateand the confining pressure σ2 σ3 is perpendicular to thedirection of the inclined plate (the microelements are smallenough so it is considered that σ2 σ3) shear behavioroccurs [22 30 31]

e angle between the shear plane and the maximumprincipal stress is defined as 45degminus (φ2) and the friction angleφ in the sand material is 36deg For a microelement ① atsmaller slope angles of 40deg and 50deg the counterclockwiseangles between the shear action direction and the overallmovement direction of the flowslide are 167deg and 157degrespectively tending to horizontal which is approximatelythe same as the direction of the maximum principal stress inthe state of the microelement ② which will be a source ofpower for the morphology of single upheaval Afterwardsthe microelements located on the horizontal plate movealong the shear plane at an angle of 27deg counterclockwisewith the movement direction of the sliding body on thehorizontal plate under the obstruction of the other particlesin the front and the push of rear particles similar to a thrustnappe in macroscopic performance resulting in depositmorphology as a single upheaval

e formation of double upheaval is more complicatedHere it will be described according to the order of the frontupheaval the wash and the back upheaval First based on thesame calculation method as for the smaller angles thecounterclockwise angles between the shear-action direction ofthe microelement ① and the movement direction of thesliding body on the horizontal plate are 93deg and 83deg at thelarger slope angles of 60deg and 70deg respectively At this pointthe sliding body deposits part of its load under the actions ofshear force and thrust force to form a rudiment of the frontupheaval while the part of the sliding body on the horizontalplate keeps moving under the inertial force and the thrust

force of the back part sliding body (the force state is similar tothe microelement②) and continues development of the frontupheaval gradually Next themicroelements at the back of thefront upheaval stop their shearing motion when the equi-librium condition of equation (3) is satisfied climbing isstopped and the wash is formed However mechanicalconditions still allow the microelements to move horizontallyWhen the equilibrium condition of equation (4) is satisfiedthe horizontal movement stops Assuming that the micro-elements are small enough the equilibrium condition can belisted as follows

m middot g τ middot sin 45deg minusφ2

1113874 1113875 (3)

m middot a τ middot cos 45deg minusφ2

1113874 1113875 (4)

where m is the mass of the microelement g is the accel-eration of gravity τ is shear stress a is the horizontal ac-celeration of the microelement and φ is the friction angle ofsand material

Eventually under the balance of equations (3) and (4)the sliding body that is still moving at the trailing edge is verysmall as the speed so the shear force provided can be re-duced accordingly At this time if the internal shearing forceof the sliding body at the trailing edge is greater than theshearing strength of the sliding body the subsequent slidingbody is deposited at the foot of the slope to form the backupheaval and conducts a localized shear motion based on theshear force similar to the microelement ① however if theshear force is less than the shear strength of the sliding bodythen the sliding body deposits at the slope break to form theback upheaval In general the order of the formation processof the double upheaval is front upheaval wash and backupheaval

e formation of double upheaval morphology is alsoowing to properties of the granular material In this studythere is a double upheaval morphology nevertheless this isnot in the study by Hu [14] e range of the granularparticle from 5times10minus 6m to 4times10minus 4m in this article is ob-viously less than the range of the granular particle from10minus 3m to 16times10minus 2m in the research by Hu [14] ereforethe internal friction angle and cohesion of the experimentalmaterial in this study are larger than those in the research byHu [14]emore the internal friction angle and cohesion ofgranular materials are the more the difficult to disturb thestate of the deposit morphology [19] is When the frontupheaval is formed the granular material with a particle sizeof 5times10minus 6m to 4times10minus 4m is more likely to keep its mor-phology of the front upheaval in the disturbance of the backsliding body compared with a particle size of 10minus 3m to16times10minus 2m If the backsliding body can not disturb theformed front upheaval it will accumulate in the slope breakforming the back upheaval meanwhile whereupon thedouble upheaval is formed

For these change laws as the slope angle increases themaximum width maximum length and area of the depositfirst increase and then decrease similar to the results by Hu[14] at the condition of the mass of fine particles changes

Phase 1Phase 2Phase 3

m1 m2

σ1

σ1

σ2

σ2

α

ατ1 τ1

12

Figure 13 Schematic illustration of shear force of flowslide

10 Advances in Civil Engineering

Based on those results a conclusion was proposed that there isan optimum proportion of fine and coarse particles in thegranular materials because the shear strength and frictioncoefficient are increasing with the mass of fine particles in-creases [41 42] In the condition that the internal frictionangle and cohesion are the smallest the maximum widthmaximum length and area are maximum Hence the de-duction mentioned before is reasonable that there should be acritical slope angle in the changes of the deposit morphology

36 Apparent Friction Coefficient To further understand therules that govern changes in the apparent friction coefficientwith slope angle and the distribution characteristics of theapparent friction coefficient in different lithologies and dif-ferent regions the apparent friction coefficients and corre-sponding slope angle data were collected for research fromthree places ere are 28 soil flowslides in the Jingyang Southplatform (34deg29prime32PrimeN 108deg50prime13Prime E) Shaanxi province China[30] consisting of loesse accumulation area is the even riverterrace comprised of the loess deposit underlain by the siltyclay deposit with interbedded sand layers occasionally Alongthe coast of northern Europe(51deg05prime03Prime N 1deg20prime36Prime E) thereare 27 chalk flows [22] with a porosity ranging from 9 to 52and dry density ranging from 2413 to 1265 kNm3 InWenchuan (31deg00prime00Prime N 103deg40prime00Prime E) Sichuan provinceChina [2] there are 44 rock flowslides that occurred in themountainous area and consisted of the sand-mudstonemainlye accumulation area is relatively flat though it is hard tocompare with flowslides that occurred in Jingyang and coast ofnorthern Europe An important reason why the above threeflowslide groups are taken as research objects is that they havesimilar flow characteristic and long runout distance In ad-dition the topographical condition of the accumulation area issimilar to the model test conditions It can be seen fromFigure 14 that the overall relationship between the apparentfriction coefficient of flowslide and the slope angles shows anonlinear growth relationship is conclusion is consistentwith Crosta [22] e growth relationship is as follows

H

L

tan θ1 + (cos θ + ε sin θ)2 middot ((tan θμ) minus 1)1113960 1113961

(5)

where θ represents the slope angle H represents the slopeheight L represents the moving distance of the flowslide ε isthe coefficient of velocity restitution when the sliding bodyreaches the slope break to change its moving direction tohorizontal μ is the friction coefficient

e difference in regional lithology determines thedifference between the friction coefficient μ and the co-efficient of velocity restitution ε so two different μ valuesand three different ε values are set in this paper When μ is02 and ε is 0 02 or 04 the curve better reflects theoverall trend of the three natural flowslide groups Whenthe friction coefficient μ between the inclined plate andthe experimental material is 055 and ε is 04 the curveshows good correlation with the experimental points Itcan be said that the friction coefficient determines theheight of the initial benchmark of the curve e larger thevalue of the friction coefficient the higher the initial

benchmark of the curve e coefficient of velocity res-titution determines the change of the inflection pointwhen the curve changes from flat to rapid e larger thevalue of the coefficient of velocity restitution the largerthe horizontal coordinates of the inflection point

From the point of energy conversion energy loss at theslope break is closely related to terrain and lithology ekinetic energy loss is small when the junction of slope and theaccumulation area is a curved slope break when the junctionhas sharp edges the kinetic energy loss is larger [13 22]Under the experimental conditions in which slope height Hand volume are constant as the slope angle increases energyloss also increases due to the collision of the flowslide and thehorizontal plate at the slope break [43 44] which results inshorter distance of movement of the sliding body e slopeangle in the chalk flow area is too large and the energy loss isexcessive though the sliding path is lubricated by water be-neath the accumulation area 1-2m e flowslides in Wen-chuan affected by lubrication due to water are few Besidesthere are many obstacles in the movement path because theseflowslides are located in a mountainous area though theaccumulation area is relatively flat so energy loss is largeHowever for the flowslides in Jingyang the path is flat andstraight as well as lubricated by the liquefied effect [31] andthe energy loss is small e flowslides in the model exper-iment fall between the three natural flowslide groups to theleft of chalk flows above loess flowslides and intersectingwith rock flowslides e apparent friction coefficient in thecondition of this model experiment is not the smallest al-though the condition is ideal flat for the reason that there areno lubrications erefore the lubrication effect plays a keyrole in affecting the apparent friction coefficient

4 Discussion

In this paper the Taguchi method was employed inno-vatively to study quantitatively the order of significance

0 20 40 60 80 10000

05

10

15

20

Wenchuan rock flowslidesJingyang loess flowslides

Chalk flowsExperimental flowslides

HL

Slope angle (deg)

ε = 0

ε = 02

ε = 0

2

ε = 04

ε = 0

4

ε = 0μ = 055

μ = 02

tan θ1 + [(cos θ + εsin θ)2 middot (tan θμ ndash1)]

ndashHL =

Figure 14 Comprehensive relationship between slope angle andapparent friction coefficient Formula comes from Crosta [22]

Advances in Civil Engineering 11

and the maximal responses of combinations of factorsaffecting flowslide deposit morphology e calculationresults in Figure 5 confirm that slope angle is the mostinfluential factor e maximum influencing combinationis a volume of 54 times10minus 3 m3 at a slope angle of 70deg and aslope height of 090m which will be useful to evaluate thehazards of potential flowslides Based on the acceptedknowledge that slope angle is the largest influencingfactor further experimental research was performed econditions of small slope angles and large slope anglesshow interesting phenomena of single upheaval anddouble upheaval respectively the Mohr-Coulomb crite-rion is used to explain the action of the shear motionduring the process e comprehensive study of apparentfriction coefficient shows a nonlinear increase with slopeangle as well as spatial and lithological differences indistribution which are affected by the friction coefficientand the coefficient of velocity restitution

Similar model experiments have been carried out toanalyze the movement rule of flowslide [4 12 21] anddeposit morphology [3 16 18 22 25] Previous studiesfocused on the influence of a single factor and the researchcontent is more in-depth [8 11 45 46] but the researchprocess is not systematic enough as flowslide movement anddeposit morphology are determined by multiple factorsincluding volume slope angle slope height sliding bodymaterial and slip bed material Of course some scholarshave conducted multifactor experiments for exampleManzella [13] studied the influence of flowslide volumeinitial height slope of slip bed and the slope break onmovement characteristics but the relative significance ofvarious factors governing the movement of flowslide has notbeen revealed although the factors considered are com-prehensive An orthogonal experiment was used to study themotion characteristics (velocity acceleration) of flowslideunder the influence of various factors and found that slopeangle was the largest influencing factor [33] which isconsistent with the findings of this article though the re-search content was different

e morphology of the double upheaval of flowslide wasalso observed in an experiment similar to the current study[1] which examined wave propagation in the macroscopicview at experiment was also carried out with bulk ma-terials A wave would be formed when the experimentalmaterials slid or fell into the accumulation area the initialwave would accelerate suddenly as it was pushed by asubsequent wave propagating forward e wave crestformed when motion stopped due to resistance may be oneof the reasons for the formation of upheaval and the wavetrough can be regarded as wash [1] in this paper Accordingto the motion rule of the object the flowslide is partiallyreflected at the slope break [22] and then propagates in theform of waves which were observed somewhat during theexperiment of larger slope angles Macroscopic wavepropagation may be the cause of the formation of upheavaland wash on the surface of flowslide but if the phenomenonis deeply understood from the interior of sliding body theinternal shear behavior in the movement of flowslide may bereasonable In addition the morphology of actual flowslides

is also affected by properties of sliding materials [14 19 20]With an increase of the mass of fine particles the length andwidth of the flowslide deposit increase first and then de-crease [14] erefore the difference of properties ofgranular materials is contributing to the formation of thedouble upheaval

e apparent friction coefficient is often involved in thestudy of flowslide but the focus here is on its relationshipwith volume [5 7 9 21 38] ere are few studies that useslope angle as a variable including Lied and Bakkehoi [47]and Okura [3] both of which proposed that the apparentfriction coefficient of flowslide was directly proportional tothe slope angle but the results were based on the analysis ofonly about 20 flowslides so the conclusion was limited eanalysis in this paper involved 99 landslides in three areasand the conclusion is consistent with the viewpoint of Crosta[22] in finding that the apparent friction coefficient increasesnonlinearly as the slope angle increases In the comparisonmade in this study spatial and lithological differences in theapparent friction coefficient are also found

In this paper an experimental idea of constant height ofthe center of gravity is provided but when the angle changesthe sliding distance of the flowslide on the inclined plate(Figure 4(b)) will inevitably increase and the loss of kineticenergy due to friction will increase accordingly ereforefurther improvement of the experimental device is needed toreduce the kinetic energy loss due to the increase in thesliding distance on the inclined plate Modeling should alsoinclude more influencing factors

5 Conclusions

(1) e combined influence factor obtained by theTaguchi method demonstrates that when flowslidevolume is 54times10minus 3m3 the slope angle is 70deg theslope height is 090m and the changes of depositmorphology of the flowslide is the largest e ex-treme differences of maximum length (Ld) maxi-mum thickness (Dd) area (A) and the length-widthratio (LdWd) at three different levels under variousfactors are 277 551 454 and 17 respectively ecalculation results indicate that the factor that has thegreatest impact on the deposit morphology offlowslide is slope angle followed by volume andslope height

(2) With an increase in slope angle the width lengtharea and length-width ratio of flowslide deposits firstincrease and then decrease and the maximum ac-cumulation thickness first decreases and then in-creases And there should be a critical angle in thechange laws of the flowslide deposit morphologythat is between 60deg and 70deg under the experimentalconditions When the volume is 36times10minus 3m3 andthe slope height is 090m the slope angles at 40deg and50deg both result in a single upheaval while slopeangles of 60deg and 70deg both result in a double upheavalwhich has been explained by shear behavior in thesliding body based on the Mohr-Coulomb criterion

12 Advances in Civil Engineering

(3) e apparent friction coefficient of flowslide in-creases nonlinearly as the slope angle increases andshows spatial and lithological difference e greaterthe friction coefficient of the sliding material thehigher the initial benchmark of the curve the largerthe coefficient of velocity restitution the larger theabscissa corresponding to the position of inflectionpoint when the curve changes from flat to rapid

Data Availability

e data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

e authors declare that they have no conflicts of interest

Acknowledgments

is study would not have been possible without financialsupport from the Special Fund for the National NaturalScience Foundation of China under Grant nos 4170229841790442 and 41602359 as well as the Project Supported byNatural Science Basic Research Plan in Shaanxi Province ofChina under Grant no 2017JQ4020

Supplementary Materials

In this section the data on the slope angles and the apparentfriction coefficients are provided as a formation of tableese data are related to 44 rock flowslides in theWenchuanSichuan province China 28 soil flowslides in in the JingyangSouth platform Shaanxi province China and 27 chalkflowslides along the coast of northern Europe (Supple-mentary Materials)

References

[1] O Roche M Attali A Mangeney and A Lucas ldquoOn the run-out distance of geophysical gravitational flows insight fromfluidized granular collapse experimentsrdquo Earth PlanetaryScience Letters vol 311 no 3-4 pp 375ndash385 2011

[2] D Guo M Hamada C He Y Wang and Y Zou ldquoAnempirical model for landslide travel distance prediction inWenchuan earthquake areardquo Landslides vol 11 no 2pp 281ndash291 2014

[3] Y Okura H Kitahara A Kawanami and U KurokawaldquoTopography and volume effects on travel distance of surfacefailurerdquo Engineering Geology vol 67 no 3-4 pp 243ndash2542003

[4] Z Han Y Li J Huang et al ldquoNumerical simulation for run-out extent of debris flows using an improved cellular au-tomaton modelrdquo Bulletin of Engineering Geology and theEnvironment vol 76 no 3 pp 961ndash974 2017

[5] F Guzzeti F Ardizzone M Cardinali M Rossi andD Valigi ldquoLandslide volumes and landslide mobilizationrates in Umbria Central Italyrdquo Earth Planetary Science Let-ters vol 279 no 3-4 pp 222ndash229 2009

[6] J D Chaytor S Uri A R Solow and B Andrews ldquoSizedistribution of submarine landslides along the US Atlanticmarginrdquo Marine Geology vol 264 no 1-2 pp 16ndash27 2009

[7] H Qiu P Cui A D Regmi et al ldquoInfluence of topographyand volume on mobility of loess slides within different slipsurfacesrdquo Catena vol 157 pp 180ndash188 2017

[8] J Hu J Peng G Wang I Javed Y Wang and W LildquoDistribution and characteristics of landslide in LoessPlateau a case study in Shaanxi provincerdquo EngineeringGeology vol 236 pp 89ndash96 2018

[9] L Xu F C Dai Q M Gong L G am and H MinldquoIrrigation-induced loess flow failure in Heifangtai Platformnorth-west Chinardquo Environmental Earth Sciences vol 66no 6 pp 1707ndash1713 2012

[10] R Q Huang W H Liu J P Zhou and X J Pei ldquoRolling testson movement characteristics of rock blocksrdquo Journal ofGeotechnical Engineering vol 29 no 9 pp 1296ndash1302 2007in Chinese

[11] I Manzella and V Labiouse ldquoPhysical modelling to betterunderstand rock avalanchesrdquo in Physical Modelling in Geo-technics Two Volume Set Proceedings of the 7th InternationalConference on Physical Modelling in Geotechnics (ICPMG2010) vol 2 pp 1259ndash1265 Taylor amp Francis Group LondonUK 2010

[12] G B Crosta F V De Blasio M De Caro G VolpiS Imposimato and D Roddeman ldquoModes of propagationand deposition of granular flows onto an erodible substrateexperimental analytical and numerical studyrdquo Landslidesvol 14 no 1 pp 47ndash68 2017

[13] I Manzella and V Labiouse ldquoEmpirical and analytical ana-lyses of laboratory granular flows to investigate rock ava-lanche propagationrdquo Landslides vol 10 no 1 pp 23ndash362013

[14] Y-X Hu H-B Li S-C Qi G Fan and J-W ZhouldquoGranular effects on depositional processes of debris ava-lanchesrdquo KSCE Journal of Civil Engineering vol 24 no 4pp 1116ndash1127 2020

[15] I Goktepeli U Atmaca and A Cakan ldquoInvestigation of heattransfer augmentation between the ribbed plates via Taguchiapproach and computational fluid dynamicsrdquo Journal ofEermal Science vol 29 no 3 pp 647ndash666 2020

[16] M Farin A Mangeney and O Roche ldquoFundamental changesof granular flow dynamics deposition and erosion processesat high slope angles insights from laboratory experimentsrdquoJournal of Geophysical Research Earth Surface vol 119 no 3pp 504ndash532 2014

[17] M Kesseler V Heller and B Turnbull ldquoA laboratory-nu-merical approach for modelling scale effects in dry granularslidesrdquo Landslides vol 15 no 11 pp 2145ndash2159 2018

[18] A Dufresne and T R Davies ldquoLongitudinal ridges in massmovement depositsrdquo Geomorphology vol 105 no 3-4pp 171ndash181 2009

[19] T De Haas L Braat J R F W Leuven I R Lokhorst andM G Kleinhans ldquoEffects of debris flow composition onrunout depositional mechanisms and deposit morphology inlaboratory experimentsrdquo Journal of Geophysical ResearchEarth Surface vol 120 no 9 pp 1949ndash1972 2015

[20] C Goujon N omas and B Dalloz-Dubrujeaud ldquoMono-disperse dry granular flows on inclined planes role ofroughnessrdquo Ee European Physical Journal E vol 11 no 2pp 147ndash157 2003

[21] X-Y Fan S-J Tian and Y-Y Zhang ldquoMass-front velocity ofdry granular flows influenced by the angle of the slope to therunout plane and particle size gradationrdquo Journal of MountainScience vol 13 no 2 pp 234ndash245 2016

[22] G B Crosta F V D Blasio M Locatelli S Imposimato andD Roddeman ldquoLandslides falling onto a shallow erodible

Advances in Civil Engineering 13

deposit volume was divided into three levels of131775times103m3 54544times103m3 and 18482times103m3 re-spectively so the similarity ratio was approximately 1 2times108 Slope height was divided into three levels of11345m 8427m and 6307m with a similarity ratio ofapproximately 1 100 e slope angle of the Jingyang Southplatformwasmainly within 40degndash70deg [31]ere were in totalnine groups of experiments two parallel experiments werecarried out for each group If the experimental results weregreater than millimeter-level in difference a third experi-ment was conducted eventually taking the average value oftwo small differences in the results to determine the depositmorphology of the flowslide

3 Results

31 Sandbox Experiment Results e Taguchi method is atest method commonly used in steel casting and other fieldswithin materials science to maximize production and re-duce consumables Its greatest advantage is to determinethe order of significance of evaluation indexes using fewertests and to determine an optimized combination schemeof influencing factors [15 26 27 29] e direct basis forthe analysis of the results is the signal-to-noise ratio (SN)which can be divided into three categories according tothree characteristics larger-the-better smaller-the-betterand nominal-the-better (washing the test results closest tothe target value) e larger the value of (SN) the greaterthe signal that the dependent variable responds to thechange of factors erefore the larger-the-better char-acteristic is used in this paper and the calculation ex-pression is as follows

S

N1113874 1113875

i minus 10lg

1n

1113944

n

j1Y2ij

⎛⎝ ⎞⎠ (1)

15m

12m

Sand container

Inclined plate

15m

High-speedcamera

(a)

Sample deposit

3D scaner

(b)

Figure 1 Flowslides experimental research design (a) structural of the experimental device (b) 3D scanner for acquisition of deposit data

Figure 2 Experimental material

0001 001 01 1

0

20

40

60

80

100

Massachusetts Institute of Technology (MIT) classification

CoarseMediumFineCoarseMediumSand

Perc

ent f

iner

by

wei

ght (

)

Particle diameter (mm)

SiltFine

Figure 3 Particle-size distribution curve for the experimentalmaterial

Advances in Civil Engineering 3

where (SN) denotes the signal-to-noise ratio n denotes therepeatable time of each variable combination Y1j denotesthe observed data of evaluation index i denotes a differentevaluation index and j is the test number from 1 to n

It is necessary to establish an evaluation index to analyzethe influencing factors of flowslide deposit morphology bythe Taguchi method because the deposit morphology offlowslide (Figure 4(a)) can be characterized by five elementsmaximum length (Ld) maximum width (Wd) maximumthickness (Dd) area (A) and length-width ratio (LdWd)

denoted by Y1j Y2j Y3j Y4j and Y5j respectively e fiveindexes used as evaluation indexes along with statisticalresults and signal-to-noise ratio calculation results areshown in Tables 2 and 3

It can be seen from Table 3 that when the volume is54times10minus 3m3 the slope angle is 70deg the slope height is 09mand the signal-to-noise ratio values of Ld Wd and A are thelargest 4055 4049 and minus 06 respectively erefore underthe influence of this combination the change responses ofLd Wd and A are the largest It can also be seen that whenthe volume is 54times10minus 3m3 the slope angle is 50deg the slopeheight is 11m and the change response of Dd is the largestWhen the volume is 18times10minus 3m3 the slope angle is 60deg theslope height is 09m and the change response of (LdWd) isthe largest Based on a comprehensive comparison of acombination of influencing factors A3B3C2mdashnamely whenthe volume is 54times10minus 3m3 the slope angle is 70deg and theslope height is 09mmdashthe change response of the flowslidedeposit morphology is the largest

32 Morphology of the Flowslide e calculation results ofthe Taguchi method and multivariate analysis of varianceboth indicate that slope angle has the greatest influence onthe deposit morphology of flowslide To further analyze themorphology of flowslide under the change of slope angleonly an experiment using four sets of sandboxes with aconstant volume of 36times10minus 3m3 a constant slope height of09m and slope angles of 40deg 50deg 60deg and 70deg was carriedout (Figure 4(b))

(1) Analysis based on Taguchi method According to thesignal-to-noise ratio values of different combinationsdetermined using equation (1) as listed in Table 3 the

values of the maximum minus minimum (extremedifference) of the signal-to-noise ratio between thethree levels under the same factor are obtained(Table 4) and shown in Figure 5 e bigger theextreme difference value of the influencing factorsthe greater the influence of this factor on the results[26 28]e results show that the values of extreme differenceof LdDdA and (LdWd) are 277 551 454 and 17respectively e largest value reflects the influence ofslope angle as compared with other factors e nextlargest value of extreme difference reflects the influ-ence of flowslide volume e value of extreme dif-ference as influenced by slope height is relatively thesmallest e value of Wd is most affected by volumethe corresponding value of extreme difference is 278In summary the most significant factor affecting thedeposit morphology of flowslide is slope angle fol-lowed by volume and slope height

(2) Multivariate analysis of variance To further verifyand analyze the impact of flowslide volume slopeangle and slope height on deposit morphology amultifactor analysis of variance is conducted on fiveevaluation indexes influenced by the three factors(Figure 5) the statistical results are shown in Table 5Partial η2 (the ratio of the within-group variance tothe overall variance of an independent variable)indicates the significance of the effect caused by theinfluence factor where the larger the value the moreobvious the effect of the influence factor e resultsof multivariate analysis of variance are consistentwith the calculation results of the Taguchi methode values of Ld Dd A and (LdWd) correspondingto the biggest partial η2 are 064 079 056 and 073all as a result of the effect of slope angle while forWd the biggest value is 087 which is the effect ofvolume of the flowslide Similarly the value of partialη2 under the influence of slope height is relatively thesmallest among the five evaluation indexes whichare 01 024 045 006 and 005 respectivelyerefore slope angle is the factor that has thegreatest influence on the deposit morphology offlowslide followed by volume and slope height

Table 1 Orthogonal experimental scheme

Number Orthogonal combination Flowslide volume (10minus 3m3) Slope angel (deg) Height (m)1 A1B1C1 18 50 072 A1B2C2 18 60 093 A1B3C3 18 70 114 A2B1C2 36 50 095 A2B2C3 36 60 116 A2B3C1 36 70 077 A3B1C3 54 50 118 A3B2C1 54 60 079 A3B3C2 54 70 09Note A B and C denote flowslide volume Slope angle and height respectively

4 Advances in Civil Engineering

33 Planar Morphology at Different Angles After correctingthe digital elevation model (DEM) data obtained by thescanner the three-dimensional coordinates of the flowslidedeposit are extracted and the contour maps of the surfacemorphology are drawn (Figure 6) At smaller slope angles of40deg and 50deg the deposit surface is relatively flat its long axis

lies along the width of the flowslide and it is generallytongue-shaped At 40deg the length width length-width ratioand area are 053m 070m 076 and 029m2 respectivelyAt 50deg the length width length-width ratio and area are089m 080m 111 and 044m2 respectively while at largerslope angles of 60deg and 70deg the deposits tend to be thickest at

Flowslide deposit

A

Ld

Dd

Wd

L

H

θ

(a)

H

θ1θ2

θ3

(b)

Figure 4 Schematic illustration of devices (a) e flowslide deposit characteristics Ld Maximum sliding length of flowslide deposit on thesurface Wd Maximum width of flowslide deposit Dd Maximum depth of flowslide deposit A Area of flowslide deposit projected onhorizontal plane θ slope angle HHeight of sandbox measuring from center of gravity L Total sliding distance of flowslide deposit (b) eprocess of center gravity constantly

Table 2 Statistics on flowslide deposit morphology

Orthogonal combination Y1jLd(10minus 2m) Y2jWd(10minus 2m) Y3jDd(10minus 2m) Y4jA(m2) Y5j(LdWd)(1)

A1B1C1 6542 6155 288 036 094A1B2C2 6833 7051 140 043 103A1B3C3 7463 7501 132 047 101A2B1C2 7333 5200 279 032 071A2B2C3 9413 9000 137 078 096A2B3C1 7927 6451 323 044 081A3B1C3 8812 6736 429 043 076A3B2C1 9508 8850 268 071 093A3B3C2 10576 10650 278 093 101Note Ld maximum sliding length of flowslide deposit on the surface Wd maximum width of flowslide deposit Dd maximum depth of flowslide deposit Aarea of flowslide deposit projected on horizontal plane θ slope angle (LdWd) length-width ratio

Table 3 e signal-to-noise ratio of elements of deposits morphology

Orthogonal combinationSignal-to-noise ratio (dB)

Ld (10-2m) Wd (10-2m) Dd (10-2m) A (m2) (LdWd) (1)

A1B1C1 3579 3631 919 minus 876 minus 053A1B2C2 3697 3669 292 minus 729 027A1B3C3 375 3746 241 minus 651 004A2B1C2 3432 3731 891 minus 991 minus 299A2B2C3 3909 3948 273 minus 21 minus 039A2B3C1 3619 3798 1018 minus 708 minus 179A3B1C3 3657 389 1265 minus 73 minus 233A3B2C1 3894 3956 856 minus 295 minus 062A3B3C2 4055 4049 888 minus 06 006Note Ld maximum sliding length of flowslide deposit on the surface Wd maximum width of flowslide deposit Dd maximum depth of flowslide deposit Aarea of flowslide deposit projected on horizontal plane θ slope angle (LdWd) length-width ratio

Advances in Civil Engineering 5

the front and back and are nearly round upheaval and washare arranged alternately on the surface At 60deg the lengthwidth length-width ratio and area are 090m 090m 10and 057m2 respectively At 70deg the length width length-width ratio and area are 074m 075m 099 and 042m2respectively e morphologies of the flowslides formed ateach of the four angles are symmetrically distributed aboutthe sliding central axis

Comparing the shape of the deposits formed at each ofthe four angles we can see that with an increase in slopeangle the values of Ld Wd A and (LdWd) for the flowslidedeposit first increase and then decrease e main reason isthat by increasing slope angle from 40deg to 60deg the sum ofkinetic energy loss by the sliding body due to friction andcollisional energy loss on the slope break is reducedHowever in increasing the slope angle from 60deg to 70deg thesum of the kinetic energy loss and the collisional energy lossincreases In the experiment the slope height was kept at aconstant value of 090m to each of the four angles so theinitial potential energy was the same for each (Figure 4(b))

e smaller the slope angle the greater the moving dis-placement of the sliding body on the inclined plate resultingin greater kinetic energy loss caused by friction When theflowslide flows to a slope break where there is a change in thedirection of motion to a horizontal direction and kineticenergy loss will inevitably occur kinetic energy loss ispositively related to the slope angle [32] Under the con-dition of changing angles the increase in the length of theflowslide deposit is significantly higher than the width isis explained by the feature of straightness [33] which theflowslide has during high-speed movement

ese laboratory findings can be examined in the contextof flowslides that occurred naturally In Xihetan Shaanxiprovince China a low-angle flowslide deposit formed at a45deg slope angle and had a length of 289m a width of 329mand a length-width ratio of 088 (Figure 7) Another low-angle deposit formed in Shenzhen China [34 35] on a slopeangle of 24deg with a length of 700m a width of about 550mand a length-width ratio of about 127 (Figure 8) A largeslope angle (more than 60deg) slide on the steep cliffs along thecoast of Northern Europe is shown in Figure 9e followingare shown in Figure 9(a) the deposit with a length 330m inFigure 9(b) the deposit with a length 150m in Figure 9(c)the deposit with a length 150m and width 120m inFigure 9(d) the deposit with a length 270m and width300m all of these deposits with a length-width ratio close to1 alternating upheaval and wash on the surface of thedeposit and an overall circular shape (Figure 9) [36 37] eactual flowslides at small angles and large angles are similarto the deposit morphology and state of motion formedduring the current experiment However due to the dif-ference in topography and the scale of the actual flowslides itis difficult to compare in terms of dimensional parameterssuch as the length and width of the deposit leaving thedimensionless length-width ratio surface morphology andoverall shape for comparison In both the natural and themodeled deposit morphology a flat surface and a generally

Table 4 e main factors of influence

Morphological elements of deposit Influencing factors Level1 (dB) Level2 (dB) Level3 (dB) Extreme difference (dB) Influence order

Ld (m)A 3675 3652 3869 216 2B 3556 3833 3808 278 1C 3697 3728 3772 075 3

Wd (m)A 3682 3825 3965 278 1B 3751 3858 3864 108 2C 3795 3816 3866 071 3

Dd (m)A 482 728 1003 521 2B 1025 474 716 551 1C 931 691 593 338 3

A (m2)A minus 752 minus 637 minus 362 390 2B minus 866 minus 411 minus 473 454 1C minus 627 minus 593 minus 531 096 3

(LdWd) (1)A minus 007 minus 172 minus 097 165 2B minus 195 minus 025 minus 561 170 1C minus 098 minus 090 minus 089 009 3

Note A B and C denote flowslide volume slope angle and height respectively Ld maximum sliding length of flowslide deposit on the surface Wdmaximum width of flowslide deposit Dd maximum depth of flowslide deposit A area of flowslide deposit projected on horizontal plane θ slope angle(LdWd) length-width ratio

A B C A B C A B C A B C A B C

0

1

2

3

4

5

6

Influencing factors

Extr

eme d

iffer

ence

(dB)

00

02

04

06

08

10

Ld Wd Dd LdWdA

Patr

ial η

2

Figure 5 Statistical results using Taguchi method and multivariateanalysis of variance

6 Advances in Civil Engineering

tongue-shaped deposit form at low angles while alternatingupheaval and wash on the deposit surface and a generallycircle-shaped deposit form at large angles

34 Profile Morphology at Different AnglesMultidimensional analysis of the morphology of flowslidedeposit can help researchers gain a deeper understanding ofmorphology characteristics Figure 10 shows the profiles ofthe four flowslides depicted in Figure 6 Combining theprofile (Figure 10) with planar morphology (Figure 6) it canbe seen that at smaller slope angles of 40deg and 50deg a generallysingle upheaval with heights of 0027m and 0018m can berecognized with corresponding peak positions of 028m and050m respectively although the surface of each deposit isslightly undulated e position of the center of gravity isconsistent with the peak position of the upheaval At largerslope angles of 60deg and 70deg the surfaces of the deposits areundulated and there is double upheaval e maximumheight differences between adjacent upheaval and wash are0005m and 0023m respectively e heights of the frontupheaval are 0009m and 0026m with corresponding peakpositions of 067m and 040m respectively e heights ofthe back upheaval are 0012m and 0035m with corre-sponding peak positions of 010m and 006m respectivelye heights of the middle wash are 0006m and 0011m withcorresponding positions of 041m and 022m respectivelye positions of the center of gravity are relatively consistentwith the position of the middle wash

Comparing the four profiles shows a single upheavalshape at smaller angles of 40deg and 50deg and a double upheavalshape at larger angles of 60deg and 70deg In the flowslides withdouble upheaval the height of the front upheaval is smallerthan that of the back upheaval As the slope angle increasesthe average thickness of the deposit first decreases and thenincreases and the position of the center of gravity firstadvances and then retreats Based on the above comparisonit can be considered that there should be a key angle in thechange laws of flowslide deposit morphology If the slopeangle is less than or larger than the key angle under the sameconditions the length and width of the deposit will decreasee key angle is between 60deg and 70deg under the experimentalconditions in the paper However actually the key angle offlowslides should be different values depending on thematerial of the sliding body and the terrain conditions

e profile figures of chalk flows [36] on the steep cliffs ofthe European coast show similarities with the profilemorphology of the sandbox experiment for the larger anglesespecially the feature of double upheaval (Figures 11 and 12)ese two deposits shown in Figures 11 and 12 are 320mand 200m in length respectively It can be seen from thepicture that the leading edge of the flowslide deposit is flatand the surface shows double upheaval e front upheavalis lower than the back upheaval e height difference be-tween the middle wash and the back upheaval is greater thanthat of the front upheaval e center of gravity is locatedslightly in front of the middle wash It is noticed that there isa difference although the morphology of the double

Table 5 Variance analysis statistics

Morphological elements of deposit Factors Quadratic sum DOF Mean square error Partial η2

Ld (m)

A 68628 2 34314 055B 98732 2 49366 064C 597 2 2985 01

Error 56458 2 28229Total 229787 8

Wd (m)

A 108303 2 54151 087B 22448 2 11224 059C 4897 2 2449 024

Error 15835 2 7917Total 151483 8

Dd (m)

A 289 2 144 076B 342 2 171 079C 073 2 037 045

Error 091 2 045Total 795 8

A (m2)

A 011 2 006 052B 013 2 007 056C 001 2 0 006

Error 011 2 005Total 036 8

(LdWd) (1)

A 004 2 002 069B 005 2 003 073C 0 2 0 005

Error 002 2 001Total 011 8

Note Ld maximum sliding length of flowslide deposit on the surface Wd maximum width of flowslide deposit Dd maximum depth of flowslide deposit Aarea of flowslide deposit projected on horizontal plane θ slope angle (LdWd) length-width ratio

Advances in Civil Engineering 7

upheaval is the same e center of gravity of flowslidedeposit is located in the middle wash in experimentalconditions However the center of gravity is located in frontof the middle wash in actual flowslides e possible reasonmainly is the difference of properties of granular materialsbecause the movement and accumulation process of flow-slides are also affected by internal factors such as the particlesize and roundness of sliding body materials [14]

35 Mechanical Analysis of the Deposit MorphologyBefore the mechanical analysis the properties of themedium-fine sand and plexiglass should be noticed as abasis for this analysis e friction coefficient 0727 of themedium-fine sand is larger than 0538 of the interface of

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

aprime

0

6

0

6121812

918120 6

(a)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

bprime0 4

8

80

412

0

04812

4

(b)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

cprime

0 46

28

8

9

9

68

4

2

02

6

84

0

20 468

8

10

8

(c)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

dprime

04

1216

20

242016128

24

16 12

048

04816

12

08

4

(d)

Figure 6 Contour map of flowslides topography at different slope angles (a) (b) (c) and (d) corresponding to 40deg 50deg 60deg and 70deg respectively

N

Figure 7 Xi Hetan flowslide in China (34deg29prime32PrimeN 108deg50prime13Prime E)

N

Figure 8 Shenzhen flowslide Source Aerial image form Sou-hucom in China (22deg43prime04Prime N 113deg56prime31Prime E)

8 Advances in Civil Engineering

the plexiglass with the medium-fine sand e differenceof the friction coefficient between the interface and thesliding body itself also existed in actual flowslides becausethe interface will be lubricated by water or the liquefactioneffect [30 38] erefore the sliding body tends to flowalong the interface of the plexiglass and the sand forms athin sand layer like an erodible bed before the main slidingbody arrived on both inclined and horizontal plates[39 40] It provides an essential condition of the shearaction Figure 13 shows the cross-sectional state of thesliding body at three phases in which phase 3 indicatedby a solid line represents the final deposit state of the

N

(a)

N

(b)

N

(c)

N

(d)

Figure 9 Topography of typical flowslides (a) and (b) located in Folkestone Warren (51deg05prime03Prime N 1deg20prime36Prime E) (c) and (d) located DoverKent (51deg07prime53Prime N 1deg20prime36Prime E) Source from Hutchinson [36]

0 200 400 600 800 10000

10

20

30

40

Dep

th o

f dep

osit

(mm

)

Length of deposit (mm)

40degprofile50degprofile

60degprofile70degprofile

Figure 10 e deposit profiles of flowslides at different angles

N

Front upheavalFront upheavalBack upheavalWash

Figure 11 Flowslide occurred in May 1981 at Le Chien Neuf(49deg46prime59Prime N 0deg25prime19Prime E) Source from Hutchinson [36]

N

Front upheavalBack upheaval

Wash

Figure 12 Flowslide occurred in May 1958 at Wissower KlinkenRugen Germany (54deg32prime09Prime N 13deg40prime42Prime E) Source fromHutchinson [36]

Advances in Civil Engineering 9

sliding body In this paper the Mohr-Coulomb criterion isused in the analysis of the formation of single and doubleupheaval as follows

τ c + σ tanφ (2)Taking the numbers① and② as two microelements in

the sliding body under a stress condition whereby themaximum principal stress σ1 of the microelement at theslope break is parallel to the direction of the inclined plateand the confining pressure σ2 σ3 is perpendicular to thedirection of the inclined plate (the microelements are smallenough so it is considered that σ2 σ3) shear behavioroccurs [22 30 31]

e angle between the shear plane and the maximumprincipal stress is defined as 45degminus (φ2) and the friction angleφ in the sand material is 36deg For a microelement ① atsmaller slope angles of 40deg and 50deg the counterclockwiseangles between the shear action direction and the overallmovement direction of the flowslide are 167deg and 157degrespectively tending to horizontal which is approximatelythe same as the direction of the maximum principal stress inthe state of the microelement ② which will be a source ofpower for the morphology of single upheaval Afterwardsthe microelements located on the horizontal plate movealong the shear plane at an angle of 27deg counterclockwisewith the movement direction of the sliding body on thehorizontal plate under the obstruction of the other particlesin the front and the push of rear particles similar to a thrustnappe in macroscopic performance resulting in depositmorphology as a single upheaval

e formation of double upheaval is more complicatedHere it will be described according to the order of the frontupheaval the wash and the back upheaval First based on thesame calculation method as for the smaller angles thecounterclockwise angles between the shear-action direction ofthe microelement ① and the movement direction of thesliding body on the horizontal plate are 93deg and 83deg at thelarger slope angles of 60deg and 70deg respectively At this pointthe sliding body deposits part of its load under the actions ofshear force and thrust force to form a rudiment of the frontupheaval while the part of the sliding body on the horizontalplate keeps moving under the inertial force and the thrust

force of the back part sliding body (the force state is similar tothe microelement②) and continues development of the frontupheaval gradually Next themicroelements at the back of thefront upheaval stop their shearing motion when the equi-librium condition of equation (3) is satisfied climbing isstopped and the wash is formed However mechanicalconditions still allow the microelements to move horizontallyWhen the equilibrium condition of equation (4) is satisfiedthe horizontal movement stops Assuming that the micro-elements are small enough the equilibrium condition can belisted as follows

m middot g τ middot sin 45deg minusφ2

1113874 1113875 (3)

m middot a τ middot cos 45deg minusφ2

1113874 1113875 (4)

where m is the mass of the microelement g is the accel-eration of gravity τ is shear stress a is the horizontal ac-celeration of the microelement and φ is the friction angle ofsand material

Eventually under the balance of equations (3) and (4)the sliding body that is still moving at the trailing edge is verysmall as the speed so the shear force provided can be re-duced accordingly At this time if the internal shearing forceof the sliding body at the trailing edge is greater than theshearing strength of the sliding body the subsequent slidingbody is deposited at the foot of the slope to form the backupheaval and conducts a localized shear motion based on theshear force similar to the microelement ① however if theshear force is less than the shear strength of the sliding bodythen the sliding body deposits at the slope break to form theback upheaval In general the order of the formation processof the double upheaval is front upheaval wash and backupheaval

e formation of double upheaval morphology is alsoowing to properties of the granular material In this studythere is a double upheaval morphology nevertheless this isnot in the study by Hu [14] e range of the granularparticle from 5times10minus 6m to 4times10minus 4m in this article is ob-viously less than the range of the granular particle from10minus 3m to 16times10minus 2m in the research by Hu [14] ereforethe internal friction angle and cohesion of the experimentalmaterial in this study are larger than those in the research byHu [14]emore the internal friction angle and cohesion ofgranular materials are the more the difficult to disturb thestate of the deposit morphology [19] is When the frontupheaval is formed the granular material with a particle sizeof 5times10minus 6m to 4times10minus 4m is more likely to keep its mor-phology of the front upheaval in the disturbance of the backsliding body compared with a particle size of 10minus 3m to16times10minus 2m If the backsliding body can not disturb theformed front upheaval it will accumulate in the slope breakforming the back upheaval meanwhile whereupon thedouble upheaval is formed

For these change laws as the slope angle increases themaximum width maximum length and area of the depositfirst increase and then decrease similar to the results by Hu[14] at the condition of the mass of fine particles changes

Phase 1Phase 2Phase 3

m1 m2

σ1

σ1

σ2

σ2

α

ατ1 τ1

12

Figure 13 Schematic illustration of shear force of flowslide

10 Advances in Civil Engineering

Based on those results a conclusion was proposed that there isan optimum proportion of fine and coarse particles in thegranular materials because the shear strength and frictioncoefficient are increasing with the mass of fine particles in-creases [41 42] In the condition that the internal frictionangle and cohesion are the smallest the maximum widthmaximum length and area are maximum Hence the de-duction mentioned before is reasonable that there should be acritical slope angle in the changes of the deposit morphology

36 Apparent Friction Coefficient To further understand therules that govern changes in the apparent friction coefficientwith slope angle and the distribution characteristics of theapparent friction coefficient in different lithologies and dif-ferent regions the apparent friction coefficients and corre-sponding slope angle data were collected for research fromthree places ere are 28 soil flowslides in the Jingyang Southplatform (34deg29prime32PrimeN 108deg50prime13Prime E) Shaanxi province China[30] consisting of loesse accumulation area is the even riverterrace comprised of the loess deposit underlain by the siltyclay deposit with interbedded sand layers occasionally Alongthe coast of northern Europe(51deg05prime03Prime N 1deg20prime36Prime E) thereare 27 chalk flows [22] with a porosity ranging from 9 to 52and dry density ranging from 2413 to 1265 kNm3 InWenchuan (31deg00prime00Prime N 103deg40prime00Prime E) Sichuan provinceChina [2] there are 44 rock flowslides that occurred in themountainous area and consisted of the sand-mudstonemainlye accumulation area is relatively flat though it is hard tocompare with flowslides that occurred in Jingyang and coast ofnorthern Europe An important reason why the above threeflowslide groups are taken as research objects is that they havesimilar flow characteristic and long runout distance In ad-dition the topographical condition of the accumulation area issimilar to the model test conditions It can be seen fromFigure 14 that the overall relationship between the apparentfriction coefficient of flowslide and the slope angles shows anonlinear growth relationship is conclusion is consistentwith Crosta [22] e growth relationship is as follows

H

L

tan θ1 + (cos θ + ε sin θ)2 middot ((tan θμ) minus 1)1113960 1113961

(5)

where θ represents the slope angle H represents the slopeheight L represents the moving distance of the flowslide ε isthe coefficient of velocity restitution when the sliding bodyreaches the slope break to change its moving direction tohorizontal μ is the friction coefficient

e difference in regional lithology determines thedifference between the friction coefficient μ and the co-efficient of velocity restitution ε so two different μ valuesand three different ε values are set in this paper When μ is02 and ε is 0 02 or 04 the curve better reflects theoverall trend of the three natural flowslide groups Whenthe friction coefficient μ between the inclined plate andthe experimental material is 055 and ε is 04 the curveshows good correlation with the experimental points Itcan be said that the friction coefficient determines theheight of the initial benchmark of the curve e larger thevalue of the friction coefficient the higher the initial

benchmark of the curve e coefficient of velocity res-titution determines the change of the inflection pointwhen the curve changes from flat to rapid e larger thevalue of the coefficient of velocity restitution the largerthe horizontal coordinates of the inflection point

From the point of energy conversion energy loss at theslope break is closely related to terrain and lithology ekinetic energy loss is small when the junction of slope and theaccumulation area is a curved slope break when the junctionhas sharp edges the kinetic energy loss is larger [13 22]Under the experimental conditions in which slope height Hand volume are constant as the slope angle increases energyloss also increases due to the collision of the flowslide and thehorizontal plate at the slope break [43 44] which results inshorter distance of movement of the sliding body e slopeangle in the chalk flow area is too large and the energy loss isexcessive though the sliding path is lubricated by water be-neath the accumulation area 1-2m e flowslides in Wen-chuan affected by lubrication due to water are few Besidesthere are many obstacles in the movement path because theseflowslides are located in a mountainous area though theaccumulation area is relatively flat so energy loss is largeHowever for the flowslides in Jingyang the path is flat andstraight as well as lubricated by the liquefied effect [31] andthe energy loss is small e flowslides in the model exper-iment fall between the three natural flowslide groups to theleft of chalk flows above loess flowslides and intersectingwith rock flowslides e apparent friction coefficient in thecondition of this model experiment is not the smallest al-though the condition is ideal flat for the reason that there areno lubrications erefore the lubrication effect plays a keyrole in affecting the apparent friction coefficient

4 Discussion

In this paper the Taguchi method was employed inno-vatively to study quantitatively the order of significance

0 20 40 60 80 10000

05

10

15

20

Wenchuan rock flowslidesJingyang loess flowslides

Chalk flowsExperimental flowslides

HL

Slope angle (deg)

ε = 0

ε = 02

ε = 0

2

ε = 04

ε = 0

4

ε = 0μ = 055

μ = 02

tan θ1 + [(cos θ + εsin θ)2 middot (tan θμ ndash1)]

ndashHL =

Figure 14 Comprehensive relationship between slope angle andapparent friction coefficient Formula comes from Crosta [22]

Advances in Civil Engineering 11

and the maximal responses of combinations of factorsaffecting flowslide deposit morphology e calculationresults in Figure 5 confirm that slope angle is the mostinfluential factor e maximum influencing combinationis a volume of 54 times10minus 3 m3 at a slope angle of 70deg and aslope height of 090m which will be useful to evaluate thehazards of potential flowslides Based on the acceptedknowledge that slope angle is the largest influencingfactor further experimental research was performed econditions of small slope angles and large slope anglesshow interesting phenomena of single upheaval anddouble upheaval respectively the Mohr-Coulomb crite-rion is used to explain the action of the shear motionduring the process e comprehensive study of apparentfriction coefficient shows a nonlinear increase with slopeangle as well as spatial and lithological differences indistribution which are affected by the friction coefficientand the coefficient of velocity restitution

Similar model experiments have been carried out toanalyze the movement rule of flowslide [4 12 21] anddeposit morphology [3 16 18 22 25] Previous studiesfocused on the influence of a single factor and the researchcontent is more in-depth [8 11 45 46] but the researchprocess is not systematic enough as flowslide movement anddeposit morphology are determined by multiple factorsincluding volume slope angle slope height sliding bodymaterial and slip bed material Of course some scholarshave conducted multifactor experiments for exampleManzella [13] studied the influence of flowslide volumeinitial height slope of slip bed and the slope break onmovement characteristics but the relative significance ofvarious factors governing the movement of flowslide has notbeen revealed although the factors considered are com-prehensive An orthogonal experiment was used to study themotion characteristics (velocity acceleration) of flowslideunder the influence of various factors and found that slopeangle was the largest influencing factor [33] which isconsistent with the findings of this article though the re-search content was different

e morphology of the double upheaval of flowslide wasalso observed in an experiment similar to the current study[1] which examined wave propagation in the macroscopicview at experiment was also carried out with bulk ma-terials A wave would be formed when the experimentalmaterials slid or fell into the accumulation area the initialwave would accelerate suddenly as it was pushed by asubsequent wave propagating forward e wave crestformed when motion stopped due to resistance may be oneof the reasons for the formation of upheaval and the wavetrough can be regarded as wash [1] in this paper Accordingto the motion rule of the object the flowslide is partiallyreflected at the slope break [22] and then propagates in theform of waves which were observed somewhat during theexperiment of larger slope angles Macroscopic wavepropagation may be the cause of the formation of upheavaland wash on the surface of flowslide but if the phenomenonis deeply understood from the interior of sliding body theinternal shear behavior in the movement of flowslide may bereasonable In addition the morphology of actual flowslides

is also affected by properties of sliding materials [14 19 20]With an increase of the mass of fine particles the length andwidth of the flowslide deposit increase first and then de-crease [14] erefore the difference of properties ofgranular materials is contributing to the formation of thedouble upheaval

e apparent friction coefficient is often involved in thestudy of flowslide but the focus here is on its relationshipwith volume [5 7 9 21 38] ere are few studies that useslope angle as a variable including Lied and Bakkehoi [47]and Okura [3] both of which proposed that the apparentfriction coefficient of flowslide was directly proportional tothe slope angle but the results were based on the analysis ofonly about 20 flowslides so the conclusion was limited eanalysis in this paper involved 99 landslides in three areasand the conclusion is consistent with the viewpoint of Crosta[22] in finding that the apparent friction coefficient increasesnonlinearly as the slope angle increases In the comparisonmade in this study spatial and lithological differences in theapparent friction coefficient are also found

In this paper an experimental idea of constant height ofthe center of gravity is provided but when the angle changesthe sliding distance of the flowslide on the inclined plate(Figure 4(b)) will inevitably increase and the loss of kineticenergy due to friction will increase accordingly ereforefurther improvement of the experimental device is needed toreduce the kinetic energy loss due to the increase in thesliding distance on the inclined plate Modeling should alsoinclude more influencing factors

5 Conclusions

(1) e combined influence factor obtained by theTaguchi method demonstrates that when flowslidevolume is 54times10minus 3m3 the slope angle is 70deg theslope height is 090m and the changes of depositmorphology of the flowslide is the largest e ex-treme differences of maximum length (Ld) maxi-mum thickness (Dd) area (A) and the length-widthratio (LdWd) at three different levels under variousfactors are 277 551 454 and 17 respectively ecalculation results indicate that the factor that has thegreatest impact on the deposit morphology offlowslide is slope angle followed by volume andslope height

(2) With an increase in slope angle the width lengtharea and length-width ratio of flowslide deposits firstincrease and then decrease and the maximum ac-cumulation thickness first decreases and then in-creases And there should be a critical angle in thechange laws of the flowslide deposit morphologythat is between 60deg and 70deg under the experimentalconditions When the volume is 36times10minus 3m3 andthe slope height is 090m the slope angles at 40deg and50deg both result in a single upheaval while slopeangles of 60deg and 70deg both result in a double upheavalwhich has been explained by shear behavior in thesliding body based on the Mohr-Coulomb criterion

12 Advances in Civil Engineering

(3) e apparent friction coefficient of flowslide in-creases nonlinearly as the slope angle increases andshows spatial and lithological difference e greaterthe friction coefficient of the sliding material thehigher the initial benchmark of the curve the largerthe coefficient of velocity restitution the larger theabscissa corresponding to the position of inflectionpoint when the curve changes from flat to rapid

Data Availability

e data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

e authors declare that they have no conflicts of interest

Acknowledgments

is study would not have been possible without financialsupport from the Special Fund for the National NaturalScience Foundation of China under Grant nos 4170229841790442 and 41602359 as well as the Project Supported byNatural Science Basic Research Plan in Shaanxi Province ofChina under Grant no 2017JQ4020

Supplementary Materials

In this section the data on the slope angles and the apparentfriction coefficients are provided as a formation of tableese data are related to 44 rock flowslides in theWenchuanSichuan province China 28 soil flowslides in in the JingyangSouth platform Shaanxi province China and 27 chalkflowslides along the coast of northern Europe (Supple-mentary Materials)

References

[1] O Roche M Attali A Mangeney and A Lucas ldquoOn the run-out distance of geophysical gravitational flows insight fromfluidized granular collapse experimentsrdquo Earth PlanetaryScience Letters vol 311 no 3-4 pp 375ndash385 2011

[2] D Guo M Hamada C He Y Wang and Y Zou ldquoAnempirical model for landslide travel distance prediction inWenchuan earthquake areardquo Landslides vol 11 no 2pp 281ndash291 2014

[3] Y Okura H Kitahara A Kawanami and U KurokawaldquoTopography and volume effects on travel distance of surfacefailurerdquo Engineering Geology vol 67 no 3-4 pp 243ndash2542003

[4] Z Han Y Li J Huang et al ldquoNumerical simulation for run-out extent of debris flows using an improved cellular au-tomaton modelrdquo Bulletin of Engineering Geology and theEnvironment vol 76 no 3 pp 961ndash974 2017

[5] F Guzzeti F Ardizzone M Cardinali M Rossi andD Valigi ldquoLandslide volumes and landslide mobilizationrates in Umbria Central Italyrdquo Earth Planetary Science Let-ters vol 279 no 3-4 pp 222ndash229 2009

[6] J D Chaytor S Uri A R Solow and B Andrews ldquoSizedistribution of submarine landslides along the US Atlanticmarginrdquo Marine Geology vol 264 no 1-2 pp 16ndash27 2009

[7] H Qiu P Cui A D Regmi et al ldquoInfluence of topographyand volume on mobility of loess slides within different slipsurfacesrdquo Catena vol 157 pp 180ndash188 2017

[8] J Hu J Peng G Wang I Javed Y Wang and W LildquoDistribution and characteristics of landslide in LoessPlateau a case study in Shaanxi provincerdquo EngineeringGeology vol 236 pp 89ndash96 2018

[9] L Xu F C Dai Q M Gong L G am and H MinldquoIrrigation-induced loess flow failure in Heifangtai Platformnorth-west Chinardquo Environmental Earth Sciences vol 66no 6 pp 1707ndash1713 2012

[10] R Q Huang W H Liu J P Zhou and X J Pei ldquoRolling testson movement characteristics of rock blocksrdquo Journal ofGeotechnical Engineering vol 29 no 9 pp 1296ndash1302 2007in Chinese

[11] I Manzella and V Labiouse ldquoPhysical modelling to betterunderstand rock avalanchesrdquo in Physical Modelling in Geo-technics Two Volume Set Proceedings of the 7th InternationalConference on Physical Modelling in Geotechnics (ICPMG2010) vol 2 pp 1259ndash1265 Taylor amp Francis Group LondonUK 2010

[12] G B Crosta F V De Blasio M De Caro G VolpiS Imposimato and D Roddeman ldquoModes of propagationand deposition of granular flows onto an erodible substrateexperimental analytical and numerical studyrdquo Landslidesvol 14 no 1 pp 47ndash68 2017

[13] I Manzella and V Labiouse ldquoEmpirical and analytical ana-lyses of laboratory granular flows to investigate rock ava-lanche propagationrdquo Landslides vol 10 no 1 pp 23ndash362013

[14] Y-X Hu H-B Li S-C Qi G Fan and J-W ZhouldquoGranular effects on depositional processes of debris ava-lanchesrdquo KSCE Journal of Civil Engineering vol 24 no 4pp 1116ndash1127 2020

[15] I Goktepeli U Atmaca and A Cakan ldquoInvestigation of heattransfer augmentation between the ribbed plates via Taguchiapproach and computational fluid dynamicsrdquo Journal ofEermal Science vol 29 no 3 pp 647ndash666 2020

[16] M Farin A Mangeney and O Roche ldquoFundamental changesof granular flow dynamics deposition and erosion processesat high slope angles insights from laboratory experimentsrdquoJournal of Geophysical Research Earth Surface vol 119 no 3pp 504ndash532 2014

[17] M Kesseler V Heller and B Turnbull ldquoA laboratory-nu-merical approach for modelling scale effects in dry granularslidesrdquo Landslides vol 15 no 11 pp 2145ndash2159 2018

[18] A Dufresne and T R Davies ldquoLongitudinal ridges in massmovement depositsrdquo Geomorphology vol 105 no 3-4pp 171ndash181 2009

[19] T De Haas L Braat J R F W Leuven I R Lokhorst andM G Kleinhans ldquoEffects of debris flow composition onrunout depositional mechanisms and deposit morphology inlaboratory experimentsrdquo Journal of Geophysical ResearchEarth Surface vol 120 no 9 pp 1949ndash1972 2015

[20] C Goujon N omas and B Dalloz-Dubrujeaud ldquoMono-disperse dry granular flows on inclined planes role ofroughnessrdquo Ee European Physical Journal E vol 11 no 2pp 147ndash157 2003

[21] X-Y Fan S-J Tian and Y-Y Zhang ldquoMass-front velocity ofdry granular flows influenced by the angle of the slope to therunout plane and particle size gradationrdquo Journal of MountainScience vol 13 no 2 pp 234ndash245 2016

[22] G B Crosta F V D Blasio M Locatelli S Imposimato andD Roddeman ldquoLandslides falling onto a shallow erodible

Advances in Civil Engineering 13

where (SN) denotes the signal-to-noise ratio n denotes therepeatable time of each variable combination Y1j denotesthe observed data of evaluation index i denotes a differentevaluation index and j is the test number from 1 to n

It is necessary to establish an evaluation index to analyzethe influencing factors of flowslide deposit morphology bythe Taguchi method because the deposit morphology offlowslide (Figure 4(a)) can be characterized by five elementsmaximum length (Ld) maximum width (Wd) maximumthickness (Dd) area (A) and length-width ratio (LdWd)

denoted by Y1j Y2j Y3j Y4j and Y5j respectively e fiveindexes used as evaluation indexes along with statisticalresults and signal-to-noise ratio calculation results areshown in Tables 2 and 3

It can be seen from Table 3 that when the volume is54times10minus 3m3 the slope angle is 70deg the slope height is 09mand the signal-to-noise ratio values of Ld Wd and A are thelargest 4055 4049 and minus 06 respectively erefore underthe influence of this combination the change responses ofLd Wd and A are the largest It can also be seen that whenthe volume is 54times10minus 3m3 the slope angle is 50deg the slopeheight is 11m and the change response of Dd is the largestWhen the volume is 18times10minus 3m3 the slope angle is 60deg theslope height is 09m and the change response of (LdWd) isthe largest Based on a comprehensive comparison of acombination of influencing factors A3B3C2mdashnamely whenthe volume is 54times10minus 3m3 the slope angle is 70deg and theslope height is 09mmdashthe change response of the flowslidedeposit morphology is the largest

32 Morphology of the Flowslide e calculation results ofthe Taguchi method and multivariate analysis of varianceboth indicate that slope angle has the greatest influence onthe deposit morphology of flowslide To further analyze themorphology of flowslide under the change of slope angleonly an experiment using four sets of sandboxes with aconstant volume of 36times10minus 3m3 a constant slope height of09m and slope angles of 40deg 50deg 60deg and 70deg was carriedout (Figure 4(b))

(1) Analysis based on Taguchi method According to thesignal-to-noise ratio values of different combinationsdetermined using equation (1) as listed in Table 3 the

values of the maximum minus minimum (extremedifference) of the signal-to-noise ratio between thethree levels under the same factor are obtained(Table 4) and shown in Figure 5 e bigger theextreme difference value of the influencing factorsthe greater the influence of this factor on the results[26 28]e results show that the values of extreme differenceof LdDdA and (LdWd) are 277 551 454 and 17respectively e largest value reflects the influence ofslope angle as compared with other factors e nextlargest value of extreme difference reflects the influ-ence of flowslide volume e value of extreme dif-ference as influenced by slope height is relatively thesmallest e value of Wd is most affected by volumethe corresponding value of extreme difference is 278In summary the most significant factor affecting thedeposit morphology of flowslide is slope angle fol-lowed by volume and slope height

(2) Multivariate analysis of variance To further verifyand analyze the impact of flowslide volume slopeangle and slope height on deposit morphology amultifactor analysis of variance is conducted on fiveevaluation indexes influenced by the three factors(Figure 5) the statistical results are shown in Table 5Partial η2 (the ratio of the within-group variance tothe overall variance of an independent variable)indicates the significance of the effect caused by theinfluence factor where the larger the value the moreobvious the effect of the influence factor e resultsof multivariate analysis of variance are consistentwith the calculation results of the Taguchi methode values of Ld Dd A and (LdWd) correspondingto the biggest partial η2 are 064 079 056 and 073all as a result of the effect of slope angle while forWd the biggest value is 087 which is the effect ofvolume of the flowslide Similarly the value of partialη2 under the influence of slope height is relatively thesmallest among the five evaluation indexes whichare 01 024 045 006 and 005 respectivelyerefore slope angle is the factor that has thegreatest influence on the deposit morphology offlowslide followed by volume and slope height

Table 1 Orthogonal experimental scheme

Number Orthogonal combination Flowslide volume (10minus 3m3) Slope angel (deg) Height (m)1 A1B1C1 18 50 072 A1B2C2 18 60 093 A1B3C3 18 70 114 A2B1C2 36 50 095 A2B2C3 36 60 116 A2B3C1 36 70 077 A3B1C3 54 50 118 A3B2C1 54 60 079 A3B3C2 54 70 09Note A B and C denote flowslide volume Slope angle and height respectively

4 Advances in Civil Engineering

33 Planar Morphology at Different Angles After correctingthe digital elevation model (DEM) data obtained by thescanner the three-dimensional coordinates of the flowslidedeposit are extracted and the contour maps of the surfacemorphology are drawn (Figure 6) At smaller slope angles of40deg and 50deg the deposit surface is relatively flat its long axis

lies along the width of the flowslide and it is generallytongue-shaped At 40deg the length width length-width ratioand area are 053m 070m 076 and 029m2 respectivelyAt 50deg the length width length-width ratio and area are089m 080m 111 and 044m2 respectively while at largerslope angles of 60deg and 70deg the deposits tend to be thickest at

Flowslide deposit

A

Ld

Dd

Wd

L

H

θ

(a)

H

θ1θ2

θ3

(b)

Figure 4 Schematic illustration of devices (a) e flowslide deposit characteristics Ld Maximum sliding length of flowslide deposit on thesurface Wd Maximum width of flowslide deposit Dd Maximum depth of flowslide deposit A Area of flowslide deposit projected onhorizontal plane θ slope angle HHeight of sandbox measuring from center of gravity L Total sliding distance of flowslide deposit (b) eprocess of center gravity constantly

Table 2 Statistics on flowslide deposit morphology

Orthogonal combination Y1jLd(10minus 2m) Y2jWd(10minus 2m) Y3jDd(10minus 2m) Y4jA(m2) Y5j(LdWd)(1)

A1B1C1 6542 6155 288 036 094A1B2C2 6833 7051 140 043 103A1B3C3 7463 7501 132 047 101A2B1C2 7333 5200 279 032 071A2B2C3 9413 9000 137 078 096A2B3C1 7927 6451 323 044 081A3B1C3 8812 6736 429 043 076A3B2C1 9508 8850 268 071 093A3B3C2 10576 10650 278 093 101Note Ld maximum sliding length of flowslide deposit on the surface Wd maximum width of flowslide deposit Dd maximum depth of flowslide deposit Aarea of flowslide deposit projected on horizontal plane θ slope angle (LdWd) length-width ratio

Table 3 e signal-to-noise ratio of elements of deposits morphology

Orthogonal combinationSignal-to-noise ratio (dB)

Ld (10-2m) Wd (10-2m) Dd (10-2m) A (m2) (LdWd) (1)

A1B1C1 3579 3631 919 minus 876 minus 053A1B2C2 3697 3669 292 minus 729 027A1B3C3 375 3746 241 minus 651 004A2B1C2 3432 3731 891 minus 991 minus 299A2B2C3 3909 3948 273 minus 21 minus 039A2B3C1 3619 3798 1018 minus 708 minus 179A3B1C3 3657 389 1265 minus 73 minus 233A3B2C1 3894 3956 856 minus 295 minus 062A3B3C2 4055 4049 888 minus 06 006Note Ld maximum sliding length of flowslide deposit on the surface Wd maximum width of flowslide deposit Dd maximum depth of flowslide deposit Aarea of flowslide deposit projected on horizontal plane θ slope angle (LdWd) length-width ratio

Advances in Civil Engineering 5

the front and back and are nearly round upheaval and washare arranged alternately on the surface At 60deg the lengthwidth length-width ratio and area are 090m 090m 10and 057m2 respectively At 70deg the length width length-width ratio and area are 074m 075m 099 and 042m2respectively e morphologies of the flowslides formed ateach of the four angles are symmetrically distributed aboutthe sliding central axis

Comparing the shape of the deposits formed at each ofthe four angles we can see that with an increase in slopeangle the values of Ld Wd A and (LdWd) for the flowslidedeposit first increase and then decrease e main reason isthat by increasing slope angle from 40deg to 60deg the sum ofkinetic energy loss by the sliding body due to friction andcollisional energy loss on the slope break is reducedHowever in increasing the slope angle from 60deg to 70deg thesum of the kinetic energy loss and the collisional energy lossincreases In the experiment the slope height was kept at aconstant value of 090m to each of the four angles so theinitial potential energy was the same for each (Figure 4(b))

e smaller the slope angle the greater the moving dis-placement of the sliding body on the inclined plate resultingin greater kinetic energy loss caused by friction When theflowslide flows to a slope break where there is a change in thedirection of motion to a horizontal direction and kineticenergy loss will inevitably occur kinetic energy loss ispositively related to the slope angle [32] Under the con-dition of changing angles the increase in the length of theflowslide deposit is significantly higher than the width isis explained by the feature of straightness [33] which theflowslide has during high-speed movement

ese laboratory findings can be examined in the contextof flowslides that occurred naturally In Xihetan Shaanxiprovince China a low-angle flowslide deposit formed at a45deg slope angle and had a length of 289m a width of 329mand a length-width ratio of 088 (Figure 7) Another low-angle deposit formed in Shenzhen China [34 35] on a slopeangle of 24deg with a length of 700m a width of about 550mand a length-width ratio of about 127 (Figure 8) A largeslope angle (more than 60deg) slide on the steep cliffs along thecoast of Northern Europe is shown in Figure 9e followingare shown in Figure 9(a) the deposit with a length 330m inFigure 9(b) the deposit with a length 150m in Figure 9(c)the deposit with a length 150m and width 120m inFigure 9(d) the deposit with a length 270m and width300m all of these deposits with a length-width ratio close to1 alternating upheaval and wash on the surface of thedeposit and an overall circular shape (Figure 9) [36 37] eactual flowslides at small angles and large angles are similarto the deposit morphology and state of motion formedduring the current experiment However due to the dif-ference in topography and the scale of the actual flowslides itis difficult to compare in terms of dimensional parameterssuch as the length and width of the deposit leaving thedimensionless length-width ratio surface morphology andoverall shape for comparison In both the natural and themodeled deposit morphology a flat surface and a generally

Table 4 e main factors of influence

Morphological elements of deposit Influencing factors Level1 (dB) Level2 (dB) Level3 (dB) Extreme difference (dB) Influence order

Ld (m)A 3675 3652 3869 216 2B 3556 3833 3808 278 1C 3697 3728 3772 075 3

Wd (m)A 3682 3825 3965 278 1B 3751 3858 3864 108 2C 3795 3816 3866 071 3

Dd (m)A 482 728 1003 521 2B 1025 474 716 551 1C 931 691 593 338 3

A (m2)A minus 752 minus 637 minus 362 390 2B minus 866 minus 411 minus 473 454 1C minus 627 minus 593 minus 531 096 3

(LdWd) (1)A minus 007 minus 172 minus 097 165 2B minus 195 minus 025 minus 561 170 1C minus 098 minus 090 minus 089 009 3

Note A B and C denote flowslide volume slope angle and height respectively Ld maximum sliding length of flowslide deposit on the surface Wdmaximum width of flowslide deposit Dd maximum depth of flowslide deposit A area of flowslide deposit projected on horizontal plane θ slope angle(LdWd) length-width ratio

A B C A B C A B C A B C A B C

0

1

2

3

4

5

6

Influencing factors

Extr

eme d

iffer

ence

(dB)

00

02

04

06

08

10

Ld Wd Dd LdWdA

Patr

ial η

2

Figure 5 Statistical results using Taguchi method and multivariateanalysis of variance

6 Advances in Civil Engineering

tongue-shaped deposit form at low angles while alternatingupheaval and wash on the deposit surface and a generallycircle-shaped deposit form at large angles

34 Profile Morphology at Different AnglesMultidimensional analysis of the morphology of flowslidedeposit can help researchers gain a deeper understanding ofmorphology characteristics Figure 10 shows the profiles ofthe four flowslides depicted in Figure 6 Combining theprofile (Figure 10) with planar morphology (Figure 6) it canbe seen that at smaller slope angles of 40deg and 50deg a generallysingle upheaval with heights of 0027m and 0018m can berecognized with corresponding peak positions of 028m and050m respectively although the surface of each deposit isslightly undulated e position of the center of gravity isconsistent with the peak position of the upheaval At largerslope angles of 60deg and 70deg the surfaces of the deposits areundulated and there is double upheaval e maximumheight differences between adjacent upheaval and wash are0005m and 0023m respectively e heights of the frontupheaval are 0009m and 0026m with corresponding peakpositions of 067m and 040m respectively e heights ofthe back upheaval are 0012m and 0035m with corre-sponding peak positions of 010m and 006m respectivelye heights of the middle wash are 0006m and 0011m withcorresponding positions of 041m and 022m respectivelye positions of the center of gravity are relatively consistentwith the position of the middle wash

Comparing the four profiles shows a single upheavalshape at smaller angles of 40deg and 50deg and a double upheavalshape at larger angles of 60deg and 70deg In the flowslides withdouble upheaval the height of the front upheaval is smallerthan that of the back upheaval As the slope angle increasesthe average thickness of the deposit first decreases and thenincreases and the position of the center of gravity firstadvances and then retreats Based on the above comparisonit can be considered that there should be a key angle in thechange laws of flowslide deposit morphology If the slopeangle is less than or larger than the key angle under the sameconditions the length and width of the deposit will decreasee key angle is between 60deg and 70deg under the experimentalconditions in the paper However actually the key angle offlowslides should be different values depending on thematerial of the sliding body and the terrain conditions

e profile figures of chalk flows [36] on the steep cliffs ofthe European coast show similarities with the profilemorphology of the sandbox experiment for the larger anglesespecially the feature of double upheaval (Figures 11 and 12)ese two deposits shown in Figures 11 and 12 are 320mand 200m in length respectively It can be seen from thepicture that the leading edge of the flowslide deposit is flatand the surface shows double upheaval e front upheavalis lower than the back upheaval e height difference be-tween the middle wash and the back upheaval is greater thanthat of the front upheaval e center of gravity is locatedslightly in front of the middle wash It is noticed that there isa difference although the morphology of the double

Table 5 Variance analysis statistics

Morphological elements of deposit Factors Quadratic sum DOF Mean square error Partial η2

Ld (m)

A 68628 2 34314 055B 98732 2 49366 064C 597 2 2985 01

Error 56458 2 28229Total 229787 8

Wd (m)

A 108303 2 54151 087B 22448 2 11224 059C 4897 2 2449 024

Error 15835 2 7917Total 151483 8

Dd (m)

A 289 2 144 076B 342 2 171 079C 073 2 037 045

Error 091 2 045Total 795 8

A (m2)

A 011 2 006 052B 013 2 007 056C 001 2 0 006

Error 011 2 005Total 036 8

(LdWd) (1)

A 004 2 002 069B 005 2 003 073C 0 2 0 005

Error 002 2 001Total 011 8

Note Ld maximum sliding length of flowslide deposit on the surface Wd maximum width of flowslide deposit Dd maximum depth of flowslide deposit Aarea of flowslide deposit projected on horizontal plane θ slope angle (LdWd) length-width ratio

Advances in Civil Engineering 7

upheaval is the same e center of gravity of flowslidedeposit is located in the middle wash in experimentalconditions However the center of gravity is located in frontof the middle wash in actual flowslides e possible reasonmainly is the difference of properties of granular materialsbecause the movement and accumulation process of flow-slides are also affected by internal factors such as the particlesize and roundness of sliding body materials [14]

35 Mechanical Analysis of the Deposit MorphologyBefore the mechanical analysis the properties of themedium-fine sand and plexiglass should be noticed as abasis for this analysis e friction coefficient 0727 of themedium-fine sand is larger than 0538 of the interface of

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

aprime

0

6

0

6121812

918120 6

(a)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

bprime0 4

8

80

412

0

04812

4

(b)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

cprime

0 46

28

8

9

9

68

4

2

02

6

84

0

20 468

8

10

8

(c)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

dprime

04

1216

20

242016128

24

16 12

048

04816

12

08

4

(d)

Figure 6 Contour map of flowslides topography at different slope angles (a) (b) (c) and (d) corresponding to 40deg 50deg 60deg and 70deg respectively

N

Figure 7 Xi Hetan flowslide in China (34deg29prime32PrimeN 108deg50prime13Prime E)

N

Figure 8 Shenzhen flowslide Source Aerial image form Sou-hucom in China (22deg43prime04Prime N 113deg56prime31Prime E)

8 Advances in Civil Engineering

the plexiglass with the medium-fine sand e differenceof the friction coefficient between the interface and thesliding body itself also existed in actual flowslides becausethe interface will be lubricated by water or the liquefactioneffect [30 38] erefore the sliding body tends to flowalong the interface of the plexiglass and the sand forms athin sand layer like an erodible bed before the main slidingbody arrived on both inclined and horizontal plates[39 40] It provides an essential condition of the shearaction Figure 13 shows the cross-sectional state of thesliding body at three phases in which phase 3 indicatedby a solid line represents the final deposit state of the

N

(a)

N

(b)

N

(c)

N

(d)

Figure 9 Topography of typical flowslides (a) and (b) located in Folkestone Warren (51deg05prime03Prime N 1deg20prime36Prime E) (c) and (d) located DoverKent (51deg07prime53Prime N 1deg20prime36Prime E) Source from Hutchinson [36]

0 200 400 600 800 10000

10

20

30

40

Dep

th o

f dep

osit

(mm

)

Length of deposit (mm)

40degprofile50degprofile

60degprofile70degprofile

Figure 10 e deposit profiles of flowslides at different angles

N

Front upheavalFront upheavalBack upheavalWash

Figure 11 Flowslide occurred in May 1981 at Le Chien Neuf(49deg46prime59Prime N 0deg25prime19Prime E) Source from Hutchinson [36]

N

Front upheavalBack upheaval

Wash

Figure 12 Flowslide occurred in May 1958 at Wissower KlinkenRugen Germany (54deg32prime09Prime N 13deg40prime42Prime E) Source fromHutchinson [36]

Advances in Civil Engineering 9

sliding body In this paper the Mohr-Coulomb criterion isused in the analysis of the formation of single and doubleupheaval as follows

τ c + σ tanφ (2)Taking the numbers① and② as two microelements in

the sliding body under a stress condition whereby themaximum principal stress σ1 of the microelement at theslope break is parallel to the direction of the inclined plateand the confining pressure σ2 σ3 is perpendicular to thedirection of the inclined plate (the microelements are smallenough so it is considered that σ2 σ3) shear behavioroccurs [22 30 31]

e angle between the shear plane and the maximumprincipal stress is defined as 45degminus (φ2) and the friction angleφ in the sand material is 36deg For a microelement ① atsmaller slope angles of 40deg and 50deg the counterclockwiseangles between the shear action direction and the overallmovement direction of the flowslide are 167deg and 157degrespectively tending to horizontal which is approximatelythe same as the direction of the maximum principal stress inthe state of the microelement ② which will be a source ofpower for the morphology of single upheaval Afterwardsthe microelements located on the horizontal plate movealong the shear plane at an angle of 27deg counterclockwisewith the movement direction of the sliding body on thehorizontal plate under the obstruction of the other particlesin the front and the push of rear particles similar to a thrustnappe in macroscopic performance resulting in depositmorphology as a single upheaval

e formation of double upheaval is more complicatedHere it will be described according to the order of the frontupheaval the wash and the back upheaval First based on thesame calculation method as for the smaller angles thecounterclockwise angles between the shear-action direction ofthe microelement ① and the movement direction of thesliding body on the horizontal plate are 93deg and 83deg at thelarger slope angles of 60deg and 70deg respectively At this pointthe sliding body deposits part of its load under the actions ofshear force and thrust force to form a rudiment of the frontupheaval while the part of the sliding body on the horizontalplate keeps moving under the inertial force and the thrust

force of the back part sliding body (the force state is similar tothe microelement②) and continues development of the frontupheaval gradually Next themicroelements at the back of thefront upheaval stop their shearing motion when the equi-librium condition of equation (3) is satisfied climbing isstopped and the wash is formed However mechanicalconditions still allow the microelements to move horizontallyWhen the equilibrium condition of equation (4) is satisfiedthe horizontal movement stops Assuming that the micro-elements are small enough the equilibrium condition can belisted as follows

m middot g τ middot sin 45deg minusφ2

1113874 1113875 (3)

m middot a τ middot cos 45deg minusφ2

1113874 1113875 (4)

where m is the mass of the microelement g is the accel-eration of gravity τ is shear stress a is the horizontal ac-celeration of the microelement and φ is the friction angle ofsand material

Eventually under the balance of equations (3) and (4)the sliding body that is still moving at the trailing edge is verysmall as the speed so the shear force provided can be re-duced accordingly At this time if the internal shearing forceof the sliding body at the trailing edge is greater than theshearing strength of the sliding body the subsequent slidingbody is deposited at the foot of the slope to form the backupheaval and conducts a localized shear motion based on theshear force similar to the microelement ① however if theshear force is less than the shear strength of the sliding bodythen the sliding body deposits at the slope break to form theback upheaval In general the order of the formation processof the double upheaval is front upheaval wash and backupheaval

e formation of double upheaval morphology is alsoowing to properties of the granular material In this studythere is a double upheaval morphology nevertheless this isnot in the study by Hu [14] e range of the granularparticle from 5times10minus 6m to 4times10minus 4m in this article is ob-viously less than the range of the granular particle from10minus 3m to 16times10minus 2m in the research by Hu [14] ereforethe internal friction angle and cohesion of the experimentalmaterial in this study are larger than those in the research byHu [14]emore the internal friction angle and cohesion ofgranular materials are the more the difficult to disturb thestate of the deposit morphology [19] is When the frontupheaval is formed the granular material with a particle sizeof 5times10minus 6m to 4times10minus 4m is more likely to keep its mor-phology of the front upheaval in the disturbance of the backsliding body compared with a particle size of 10minus 3m to16times10minus 2m If the backsliding body can not disturb theformed front upheaval it will accumulate in the slope breakforming the back upheaval meanwhile whereupon thedouble upheaval is formed

For these change laws as the slope angle increases themaximum width maximum length and area of the depositfirst increase and then decrease similar to the results by Hu[14] at the condition of the mass of fine particles changes

Phase 1Phase 2Phase 3

m1 m2

σ1

σ1

σ2

σ2

α

ατ1 τ1

12

Figure 13 Schematic illustration of shear force of flowslide

10 Advances in Civil Engineering

Based on those results a conclusion was proposed that there isan optimum proportion of fine and coarse particles in thegranular materials because the shear strength and frictioncoefficient are increasing with the mass of fine particles in-creases [41 42] In the condition that the internal frictionangle and cohesion are the smallest the maximum widthmaximum length and area are maximum Hence the de-duction mentioned before is reasonable that there should be acritical slope angle in the changes of the deposit morphology

36 Apparent Friction Coefficient To further understand therules that govern changes in the apparent friction coefficientwith slope angle and the distribution characteristics of theapparent friction coefficient in different lithologies and dif-ferent regions the apparent friction coefficients and corre-sponding slope angle data were collected for research fromthree places ere are 28 soil flowslides in the Jingyang Southplatform (34deg29prime32PrimeN 108deg50prime13Prime E) Shaanxi province China[30] consisting of loesse accumulation area is the even riverterrace comprised of the loess deposit underlain by the siltyclay deposit with interbedded sand layers occasionally Alongthe coast of northern Europe(51deg05prime03Prime N 1deg20prime36Prime E) thereare 27 chalk flows [22] with a porosity ranging from 9 to 52and dry density ranging from 2413 to 1265 kNm3 InWenchuan (31deg00prime00Prime N 103deg40prime00Prime E) Sichuan provinceChina [2] there are 44 rock flowslides that occurred in themountainous area and consisted of the sand-mudstonemainlye accumulation area is relatively flat though it is hard tocompare with flowslides that occurred in Jingyang and coast ofnorthern Europe An important reason why the above threeflowslide groups are taken as research objects is that they havesimilar flow characteristic and long runout distance In ad-dition the topographical condition of the accumulation area issimilar to the model test conditions It can be seen fromFigure 14 that the overall relationship between the apparentfriction coefficient of flowslide and the slope angles shows anonlinear growth relationship is conclusion is consistentwith Crosta [22] e growth relationship is as follows

H

L

tan θ1 + (cos θ + ε sin θ)2 middot ((tan θμ) minus 1)1113960 1113961

(5)

where θ represents the slope angle H represents the slopeheight L represents the moving distance of the flowslide ε isthe coefficient of velocity restitution when the sliding bodyreaches the slope break to change its moving direction tohorizontal μ is the friction coefficient

e difference in regional lithology determines thedifference between the friction coefficient μ and the co-efficient of velocity restitution ε so two different μ valuesand three different ε values are set in this paper When μ is02 and ε is 0 02 or 04 the curve better reflects theoverall trend of the three natural flowslide groups Whenthe friction coefficient μ between the inclined plate andthe experimental material is 055 and ε is 04 the curveshows good correlation with the experimental points Itcan be said that the friction coefficient determines theheight of the initial benchmark of the curve e larger thevalue of the friction coefficient the higher the initial

benchmark of the curve e coefficient of velocity res-titution determines the change of the inflection pointwhen the curve changes from flat to rapid e larger thevalue of the coefficient of velocity restitution the largerthe horizontal coordinates of the inflection point

From the point of energy conversion energy loss at theslope break is closely related to terrain and lithology ekinetic energy loss is small when the junction of slope and theaccumulation area is a curved slope break when the junctionhas sharp edges the kinetic energy loss is larger [13 22]Under the experimental conditions in which slope height Hand volume are constant as the slope angle increases energyloss also increases due to the collision of the flowslide and thehorizontal plate at the slope break [43 44] which results inshorter distance of movement of the sliding body e slopeangle in the chalk flow area is too large and the energy loss isexcessive though the sliding path is lubricated by water be-neath the accumulation area 1-2m e flowslides in Wen-chuan affected by lubrication due to water are few Besidesthere are many obstacles in the movement path because theseflowslides are located in a mountainous area though theaccumulation area is relatively flat so energy loss is largeHowever for the flowslides in Jingyang the path is flat andstraight as well as lubricated by the liquefied effect [31] andthe energy loss is small e flowslides in the model exper-iment fall between the three natural flowslide groups to theleft of chalk flows above loess flowslides and intersectingwith rock flowslides e apparent friction coefficient in thecondition of this model experiment is not the smallest al-though the condition is ideal flat for the reason that there areno lubrications erefore the lubrication effect plays a keyrole in affecting the apparent friction coefficient

4 Discussion

In this paper the Taguchi method was employed inno-vatively to study quantitatively the order of significance

0 20 40 60 80 10000

05

10

15

20

Wenchuan rock flowslidesJingyang loess flowslides

Chalk flowsExperimental flowslides

HL

Slope angle (deg)

ε = 0

ε = 02

ε = 0

2

ε = 04

ε = 0

4

ε = 0μ = 055

μ = 02

tan θ1 + [(cos θ + εsin θ)2 middot (tan θμ ndash1)]

ndashHL =

Figure 14 Comprehensive relationship between slope angle andapparent friction coefficient Formula comes from Crosta [22]

Advances in Civil Engineering 11

and the maximal responses of combinations of factorsaffecting flowslide deposit morphology e calculationresults in Figure 5 confirm that slope angle is the mostinfluential factor e maximum influencing combinationis a volume of 54 times10minus 3 m3 at a slope angle of 70deg and aslope height of 090m which will be useful to evaluate thehazards of potential flowslides Based on the acceptedknowledge that slope angle is the largest influencingfactor further experimental research was performed econditions of small slope angles and large slope anglesshow interesting phenomena of single upheaval anddouble upheaval respectively the Mohr-Coulomb crite-rion is used to explain the action of the shear motionduring the process e comprehensive study of apparentfriction coefficient shows a nonlinear increase with slopeangle as well as spatial and lithological differences indistribution which are affected by the friction coefficientand the coefficient of velocity restitution

Similar model experiments have been carried out toanalyze the movement rule of flowslide [4 12 21] anddeposit morphology [3 16 18 22 25] Previous studiesfocused on the influence of a single factor and the researchcontent is more in-depth [8 11 45 46] but the researchprocess is not systematic enough as flowslide movement anddeposit morphology are determined by multiple factorsincluding volume slope angle slope height sliding bodymaterial and slip bed material Of course some scholarshave conducted multifactor experiments for exampleManzella [13] studied the influence of flowslide volumeinitial height slope of slip bed and the slope break onmovement characteristics but the relative significance ofvarious factors governing the movement of flowslide has notbeen revealed although the factors considered are com-prehensive An orthogonal experiment was used to study themotion characteristics (velocity acceleration) of flowslideunder the influence of various factors and found that slopeangle was the largest influencing factor [33] which isconsistent with the findings of this article though the re-search content was different

e morphology of the double upheaval of flowslide wasalso observed in an experiment similar to the current study[1] which examined wave propagation in the macroscopicview at experiment was also carried out with bulk ma-terials A wave would be formed when the experimentalmaterials slid or fell into the accumulation area the initialwave would accelerate suddenly as it was pushed by asubsequent wave propagating forward e wave crestformed when motion stopped due to resistance may be oneof the reasons for the formation of upheaval and the wavetrough can be regarded as wash [1] in this paper Accordingto the motion rule of the object the flowslide is partiallyreflected at the slope break [22] and then propagates in theform of waves which were observed somewhat during theexperiment of larger slope angles Macroscopic wavepropagation may be the cause of the formation of upheavaland wash on the surface of flowslide but if the phenomenonis deeply understood from the interior of sliding body theinternal shear behavior in the movement of flowslide may bereasonable In addition the morphology of actual flowslides

is also affected by properties of sliding materials [14 19 20]With an increase of the mass of fine particles the length andwidth of the flowslide deposit increase first and then de-crease [14] erefore the difference of properties ofgranular materials is contributing to the formation of thedouble upheaval

e apparent friction coefficient is often involved in thestudy of flowslide but the focus here is on its relationshipwith volume [5 7 9 21 38] ere are few studies that useslope angle as a variable including Lied and Bakkehoi [47]and Okura [3] both of which proposed that the apparentfriction coefficient of flowslide was directly proportional tothe slope angle but the results were based on the analysis ofonly about 20 flowslides so the conclusion was limited eanalysis in this paper involved 99 landslides in three areasand the conclusion is consistent with the viewpoint of Crosta[22] in finding that the apparent friction coefficient increasesnonlinearly as the slope angle increases In the comparisonmade in this study spatial and lithological differences in theapparent friction coefficient are also found

In this paper an experimental idea of constant height ofthe center of gravity is provided but when the angle changesthe sliding distance of the flowslide on the inclined plate(Figure 4(b)) will inevitably increase and the loss of kineticenergy due to friction will increase accordingly ereforefurther improvement of the experimental device is needed toreduce the kinetic energy loss due to the increase in thesliding distance on the inclined plate Modeling should alsoinclude more influencing factors

5 Conclusions

(1) e combined influence factor obtained by theTaguchi method demonstrates that when flowslidevolume is 54times10minus 3m3 the slope angle is 70deg theslope height is 090m and the changes of depositmorphology of the flowslide is the largest e ex-treme differences of maximum length (Ld) maxi-mum thickness (Dd) area (A) and the length-widthratio (LdWd) at three different levels under variousfactors are 277 551 454 and 17 respectively ecalculation results indicate that the factor that has thegreatest impact on the deposit morphology offlowslide is slope angle followed by volume andslope height

(2) With an increase in slope angle the width lengtharea and length-width ratio of flowslide deposits firstincrease and then decrease and the maximum ac-cumulation thickness first decreases and then in-creases And there should be a critical angle in thechange laws of the flowslide deposit morphologythat is between 60deg and 70deg under the experimentalconditions When the volume is 36times10minus 3m3 andthe slope height is 090m the slope angles at 40deg and50deg both result in a single upheaval while slopeangles of 60deg and 70deg both result in a double upheavalwhich has been explained by shear behavior in thesliding body based on the Mohr-Coulomb criterion

12 Advances in Civil Engineering

(3) e apparent friction coefficient of flowslide in-creases nonlinearly as the slope angle increases andshows spatial and lithological difference e greaterthe friction coefficient of the sliding material thehigher the initial benchmark of the curve the largerthe coefficient of velocity restitution the larger theabscissa corresponding to the position of inflectionpoint when the curve changes from flat to rapid

Data Availability

e data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

e authors declare that they have no conflicts of interest

Acknowledgments

is study would not have been possible without financialsupport from the Special Fund for the National NaturalScience Foundation of China under Grant nos 4170229841790442 and 41602359 as well as the Project Supported byNatural Science Basic Research Plan in Shaanxi Province ofChina under Grant no 2017JQ4020

Supplementary Materials

In this section the data on the slope angles and the apparentfriction coefficients are provided as a formation of tableese data are related to 44 rock flowslides in theWenchuanSichuan province China 28 soil flowslides in in the JingyangSouth platform Shaanxi province China and 27 chalkflowslides along the coast of northern Europe (Supple-mentary Materials)

References

[1] O Roche M Attali A Mangeney and A Lucas ldquoOn the run-out distance of geophysical gravitational flows insight fromfluidized granular collapse experimentsrdquo Earth PlanetaryScience Letters vol 311 no 3-4 pp 375ndash385 2011

[2] D Guo M Hamada C He Y Wang and Y Zou ldquoAnempirical model for landslide travel distance prediction inWenchuan earthquake areardquo Landslides vol 11 no 2pp 281ndash291 2014

[3] Y Okura H Kitahara A Kawanami and U KurokawaldquoTopography and volume effects on travel distance of surfacefailurerdquo Engineering Geology vol 67 no 3-4 pp 243ndash2542003

[4] Z Han Y Li J Huang et al ldquoNumerical simulation for run-out extent of debris flows using an improved cellular au-tomaton modelrdquo Bulletin of Engineering Geology and theEnvironment vol 76 no 3 pp 961ndash974 2017

[5] F Guzzeti F Ardizzone M Cardinali M Rossi andD Valigi ldquoLandslide volumes and landslide mobilizationrates in Umbria Central Italyrdquo Earth Planetary Science Let-ters vol 279 no 3-4 pp 222ndash229 2009

[6] J D Chaytor S Uri A R Solow and B Andrews ldquoSizedistribution of submarine landslides along the US Atlanticmarginrdquo Marine Geology vol 264 no 1-2 pp 16ndash27 2009

[7] H Qiu P Cui A D Regmi et al ldquoInfluence of topographyand volume on mobility of loess slides within different slipsurfacesrdquo Catena vol 157 pp 180ndash188 2017

[8] J Hu J Peng G Wang I Javed Y Wang and W LildquoDistribution and characteristics of landslide in LoessPlateau a case study in Shaanxi provincerdquo EngineeringGeology vol 236 pp 89ndash96 2018

[9] L Xu F C Dai Q M Gong L G am and H MinldquoIrrigation-induced loess flow failure in Heifangtai Platformnorth-west Chinardquo Environmental Earth Sciences vol 66no 6 pp 1707ndash1713 2012

[10] R Q Huang W H Liu J P Zhou and X J Pei ldquoRolling testson movement characteristics of rock blocksrdquo Journal ofGeotechnical Engineering vol 29 no 9 pp 1296ndash1302 2007in Chinese

[11] I Manzella and V Labiouse ldquoPhysical modelling to betterunderstand rock avalanchesrdquo in Physical Modelling in Geo-technics Two Volume Set Proceedings of the 7th InternationalConference on Physical Modelling in Geotechnics (ICPMG2010) vol 2 pp 1259ndash1265 Taylor amp Francis Group LondonUK 2010

[12] G B Crosta F V De Blasio M De Caro G VolpiS Imposimato and D Roddeman ldquoModes of propagationand deposition of granular flows onto an erodible substrateexperimental analytical and numerical studyrdquo Landslidesvol 14 no 1 pp 47ndash68 2017

[13] I Manzella and V Labiouse ldquoEmpirical and analytical ana-lyses of laboratory granular flows to investigate rock ava-lanche propagationrdquo Landslides vol 10 no 1 pp 23ndash362013

[14] Y-X Hu H-B Li S-C Qi G Fan and J-W ZhouldquoGranular effects on depositional processes of debris ava-lanchesrdquo KSCE Journal of Civil Engineering vol 24 no 4pp 1116ndash1127 2020

[15] I Goktepeli U Atmaca and A Cakan ldquoInvestigation of heattransfer augmentation between the ribbed plates via Taguchiapproach and computational fluid dynamicsrdquo Journal ofEermal Science vol 29 no 3 pp 647ndash666 2020

[16] M Farin A Mangeney and O Roche ldquoFundamental changesof granular flow dynamics deposition and erosion processesat high slope angles insights from laboratory experimentsrdquoJournal of Geophysical Research Earth Surface vol 119 no 3pp 504ndash532 2014

[17] M Kesseler V Heller and B Turnbull ldquoA laboratory-nu-merical approach for modelling scale effects in dry granularslidesrdquo Landslides vol 15 no 11 pp 2145ndash2159 2018

[18] A Dufresne and T R Davies ldquoLongitudinal ridges in massmovement depositsrdquo Geomorphology vol 105 no 3-4pp 171ndash181 2009

[19] T De Haas L Braat J R F W Leuven I R Lokhorst andM G Kleinhans ldquoEffects of debris flow composition onrunout depositional mechanisms and deposit morphology inlaboratory experimentsrdquo Journal of Geophysical ResearchEarth Surface vol 120 no 9 pp 1949ndash1972 2015

[20] C Goujon N omas and B Dalloz-Dubrujeaud ldquoMono-disperse dry granular flows on inclined planes role ofroughnessrdquo Ee European Physical Journal E vol 11 no 2pp 147ndash157 2003

[21] X-Y Fan S-J Tian and Y-Y Zhang ldquoMass-front velocity ofdry granular flows influenced by the angle of the slope to therunout plane and particle size gradationrdquo Journal of MountainScience vol 13 no 2 pp 234ndash245 2016

[22] G B Crosta F V D Blasio M Locatelli S Imposimato andD Roddeman ldquoLandslides falling onto a shallow erodible

Advances in Civil Engineering 13

33 Planar Morphology at Different Angles After correctingthe digital elevation model (DEM) data obtained by thescanner the three-dimensional coordinates of the flowslidedeposit are extracted and the contour maps of the surfacemorphology are drawn (Figure 6) At smaller slope angles of40deg and 50deg the deposit surface is relatively flat its long axis

lies along the width of the flowslide and it is generallytongue-shaped At 40deg the length width length-width ratioand area are 053m 070m 076 and 029m2 respectivelyAt 50deg the length width length-width ratio and area are089m 080m 111 and 044m2 respectively while at largerslope angles of 60deg and 70deg the deposits tend to be thickest at

Flowslide deposit

A

Ld

Dd

Wd

L

H

θ

(a)

H

θ1θ2

θ3

(b)

Figure 4 Schematic illustration of devices (a) e flowslide deposit characteristics Ld Maximum sliding length of flowslide deposit on thesurface Wd Maximum width of flowslide deposit Dd Maximum depth of flowslide deposit A Area of flowslide deposit projected onhorizontal plane θ slope angle HHeight of sandbox measuring from center of gravity L Total sliding distance of flowslide deposit (b) eprocess of center gravity constantly

Table 2 Statistics on flowslide deposit morphology

Orthogonal combination Y1jLd(10minus 2m) Y2jWd(10minus 2m) Y3jDd(10minus 2m) Y4jA(m2) Y5j(LdWd)(1)

A1B1C1 6542 6155 288 036 094A1B2C2 6833 7051 140 043 103A1B3C3 7463 7501 132 047 101A2B1C2 7333 5200 279 032 071A2B2C3 9413 9000 137 078 096A2B3C1 7927 6451 323 044 081A3B1C3 8812 6736 429 043 076A3B2C1 9508 8850 268 071 093A3B3C2 10576 10650 278 093 101Note Ld maximum sliding length of flowslide deposit on the surface Wd maximum width of flowslide deposit Dd maximum depth of flowslide deposit Aarea of flowslide deposit projected on horizontal plane θ slope angle (LdWd) length-width ratio

Table 3 e signal-to-noise ratio of elements of deposits morphology

Orthogonal combinationSignal-to-noise ratio (dB)

Ld (10-2m) Wd (10-2m) Dd (10-2m) A (m2) (LdWd) (1)

A1B1C1 3579 3631 919 minus 876 minus 053A1B2C2 3697 3669 292 minus 729 027A1B3C3 375 3746 241 minus 651 004A2B1C2 3432 3731 891 minus 991 minus 299A2B2C3 3909 3948 273 minus 21 minus 039A2B3C1 3619 3798 1018 minus 708 minus 179A3B1C3 3657 389 1265 minus 73 minus 233A3B2C1 3894 3956 856 minus 295 minus 062A3B3C2 4055 4049 888 minus 06 006Note Ld maximum sliding length of flowslide deposit on the surface Wd maximum width of flowslide deposit Dd maximum depth of flowslide deposit Aarea of flowslide deposit projected on horizontal plane θ slope angle (LdWd) length-width ratio

Advances in Civil Engineering 5

the front and back and are nearly round upheaval and washare arranged alternately on the surface At 60deg the lengthwidth length-width ratio and area are 090m 090m 10and 057m2 respectively At 70deg the length width length-width ratio and area are 074m 075m 099 and 042m2respectively e morphologies of the flowslides formed ateach of the four angles are symmetrically distributed aboutthe sliding central axis

Comparing the shape of the deposits formed at each ofthe four angles we can see that with an increase in slopeangle the values of Ld Wd A and (LdWd) for the flowslidedeposit first increase and then decrease e main reason isthat by increasing slope angle from 40deg to 60deg the sum ofkinetic energy loss by the sliding body due to friction andcollisional energy loss on the slope break is reducedHowever in increasing the slope angle from 60deg to 70deg thesum of the kinetic energy loss and the collisional energy lossincreases In the experiment the slope height was kept at aconstant value of 090m to each of the four angles so theinitial potential energy was the same for each (Figure 4(b))

e smaller the slope angle the greater the moving dis-placement of the sliding body on the inclined plate resultingin greater kinetic energy loss caused by friction When theflowslide flows to a slope break where there is a change in thedirection of motion to a horizontal direction and kineticenergy loss will inevitably occur kinetic energy loss ispositively related to the slope angle [32] Under the con-dition of changing angles the increase in the length of theflowslide deposit is significantly higher than the width isis explained by the feature of straightness [33] which theflowslide has during high-speed movement

ese laboratory findings can be examined in the contextof flowslides that occurred naturally In Xihetan Shaanxiprovince China a low-angle flowslide deposit formed at a45deg slope angle and had a length of 289m a width of 329mand a length-width ratio of 088 (Figure 7) Another low-angle deposit formed in Shenzhen China [34 35] on a slopeangle of 24deg with a length of 700m a width of about 550mand a length-width ratio of about 127 (Figure 8) A largeslope angle (more than 60deg) slide on the steep cliffs along thecoast of Northern Europe is shown in Figure 9e followingare shown in Figure 9(a) the deposit with a length 330m inFigure 9(b) the deposit with a length 150m in Figure 9(c)the deposit with a length 150m and width 120m inFigure 9(d) the deposit with a length 270m and width300m all of these deposits with a length-width ratio close to1 alternating upheaval and wash on the surface of thedeposit and an overall circular shape (Figure 9) [36 37] eactual flowslides at small angles and large angles are similarto the deposit morphology and state of motion formedduring the current experiment However due to the dif-ference in topography and the scale of the actual flowslides itis difficult to compare in terms of dimensional parameterssuch as the length and width of the deposit leaving thedimensionless length-width ratio surface morphology andoverall shape for comparison In both the natural and themodeled deposit morphology a flat surface and a generally

Table 4 e main factors of influence

Morphological elements of deposit Influencing factors Level1 (dB) Level2 (dB) Level3 (dB) Extreme difference (dB) Influence order

Ld (m)A 3675 3652 3869 216 2B 3556 3833 3808 278 1C 3697 3728 3772 075 3

Wd (m)A 3682 3825 3965 278 1B 3751 3858 3864 108 2C 3795 3816 3866 071 3

Dd (m)A 482 728 1003 521 2B 1025 474 716 551 1C 931 691 593 338 3

A (m2)A minus 752 minus 637 minus 362 390 2B minus 866 minus 411 minus 473 454 1C minus 627 minus 593 minus 531 096 3

(LdWd) (1)A minus 007 minus 172 minus 097 165 2B minus 195 minus 025 minus 561 170 1C minus 098 minus 090 minus 089 009 3

Note A B and C denote flowslide volume slope angle and height respectively Ld maximum sliding length of flowslide deposit on the surface Wdmaximum width of flowslide deposit Dd maximum depth of flowslide deposit A area of flowslide deposit projected on horizontal plane θ slope angle(LdWd) length-width ratio

A B C A B C A B C A B C A B C

0

1

2

3

4

5

6

Influencing factors

Extr

eme d

iffer

ence

(dB)

00

02

04

06

08

10

Ld Wd Dd LdWdA

Patr

ial η

2

Figure 5 Statistical results using Taguchi method and multivariateanalysis of variance

6 Advances in Civil Engineering

tongue-shaped deposit form at low angles while alternatingupheaval and wash on the deposit surface and a generallycircle-shaped deposit form at large angles

34 Profile Morphology at Different AnglesMultidimensional analysis of the morphology of flowslidedeposit can help researchers gain a deeper understanding ofmorphology characteristics Figure 10 shows the profiles ofthe four flowslides depicted in Figure 6 Combining theprofile (Figure 10) with planar morphology (Figure 6) it canbe seen that at smaller slope angles of 40deg and 50deg a generallysingle upheaval with heights of 0027m and 0018m can berecognized with corresponding peak positions of 028m and050m respectively although the surface of each deposit isslightly undulated e position of the center of gravity isconsistent with the peak position of the upheaval At largerslope angles of 60deg and 70deg the surfaces of the deposits areundulated and there is double upheaval e maximumheight differences between adjacent upheaval and wash are0005m and 0023m respectively e heights of the frontupheaval are 0009m and 0026m with corresponding peakpositions of 067m and 040m respectively e heights ofthe back upheaval are 0012m and 0035m with corre-sponding peak positions of 010m and 006m respectivelye heights of the middle wash are 0006m and 0011m withcorresponding positions of 041m and 022m respectivelye positions of the center of gravity are relatively consistentwith the position of the middle wash

Comparing the four profiles shows a single upheavalshape at smaller angles of 40deg and 50deg and a double upheavalshape at larger angles of 60deg and 70deg In the flowslides withdouble upheaval the height of the front upheaval is smallerthan that of the back upheaval As the slope angle increasesthe average thickness of the deposit first decreases and thenincreases and the position of the center of gravity firstadvances and then retreats Based on the above comparisonit can be considered that there should be a key angle in thechange laws of flowslide deposit morphology If the slopeangle is less than or larger than the key angle under the sameconditions the length and width of the deposit will decreasee key angle is between 60deg and 70deg under the experimentalconditions in the paper However actually the key angle offlowslides should be different values depending on thematerial of the sliding body and the terrain conditions

e profile figures of chalk flows [36] on the steep cliffs ofthe European coast show similarities with the profilemorphology of the sandbox experiment for the larger anglesespecially the feature of double upheaval (Figures 11 and 12)ese two deposits shown in Figures 11 and 12 are 320mand 200m in length respectively It can be seen from thepicture that the leading edge of the flowslide deposit is flatand the surface shows double upheaval e front upheavalis lower than the back upheaval e height difference be-tween the middle wash and the back upheaval is greater thanthat of the front upheaval e center of gravity is locatedslightly in front of the middle wash It is noticed that there isa difference although the morphology of the double

Table 5 Variance analysis statistics

Morphological elements of deposit Factors Quadratic sum DOF Mean square error Partial η2

Ld (m)

A 68628 2 34314 055B 98732 2 49366 064C 597 2 2985 01

Error 56458 2 28229Total 229787 8

Wd (m)

A 108303 2 54151 087B 22448 2 11224 059C 4897 2 2449 024

Error 15835 2 7917Total 151483 8

Dd (m)

A 289 2 144 076B 342 2 171 079C 073 2 037 045

Error 091 2 045Total 795 8

A (m2)

A 011 2 006 052B 013 2 007 056C 001 2 0 006

Error 011 2 005Total 036 8

(LdWd) (1)

A 004 2 002 069B 005 2 003 073C 0 2 0 005

Error 002 2 001Total 011 8

Note Ld maximum sliding length of flowslide deposit on the surface Wd maximum width of flowslide deposit Dd maximum depth of flowslide deposit Aarea of flowslide deposit projected on horizontal plane θ slope angle (LdWd) length-width ratio

Advances in Civil Engineering 7

upheaval is the same e center of gravity of flowslidedeposit is located in the middle wash in experimentalconditions However the center of gravity is located in frontof the middle wash in actual flowslides e possible reasonmainly is the difference of properties of granular materialsbecause the movement and accumulation process of flow-slides are also affected by internal factors such as the particlesize and roundness of sliding body materials [14]

35 Mechanical Analysis of the Deposit MorphologyBefore the mechanical analysis the properties of themedium-fine sand and plexiglass should be noticed as abasis for this analysis e friction coefficient 0727 of themedium-fine sand is larger than 0538 of the interface of

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

aprime

0

6

0

6121812

918120 6

(a)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

bprime0 4

8

80

412

0

04812

4

(b)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

cprime

0 46

28

8

9

9

68

4

2

02

6

84

0

20 468

8

10

8

(c)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

dprime

04

1216

20

242016128

24

16 12

048

04816

12

08

4

(d)

Figure 6 Contour map of flowslides topography at different slope angles (a) (b) (c) and (d) corresponding to 40deg 50deg 60deg and 70deg respectively

N

Figure 7 Xi Hetan flowslide in China (34deg29prime32PrimeN 108deg50prime13Prime E)

N

Figure 8 Shenzhen flowslide Source Aerial image form Sou-hucom in China (22deg43prime04Prime N 113deg56prime31Prime E)

8 Advances in Civil Engineering

the plexiglass with the medium-fine sand e differenceof the friction coefficient between the interface and thesliding body itself also existed in actual flowslides becausethe interface will be lubricated by water or the liquefactioneffect [30 38] erefore the sliding body tends to flowalong the interface of the plexiglass and the sand forms athin sand layer like an erodible bed before the main slidingbody arrived on both inclined and horizontal plates[39 40] It provides an essential condition of the shearaction Figure 13 shows the cross-sectional state of thesliding body at three phases in which phase 3 indicatedby a solid line represents the final deposit state of the

N

(a)

N

(b)

N

(c)

N

(d)

Figure 9 Topography of typical flowslides (a) and (b) located in Folkestone Warren (51deg05prime03Prime N 1deg20prime36Prime E) (c) and (d) located DoverKent (51deg07prime53Prime N 1deg20prime36Prime E) Source from Hutchinson [36]

0 200 400 600 800 10000

10

20

30

40

Dep

th o

f dep

osit

(mm

)

Length of deposit (mm)

40degprofile50degprofile

60degprofile70degprofile

Figure 10 e deposit profiles of flowslides at different angles

N

Front upheavalFront upheavalBack upheavalWash

Figure 11 Flowslide occurred in May 1981 at Le Chien Neuf(49deg46prime59Prime N 0deg25prime19Prime E) Source from Hutchinson [36]

N

Front upheavalBack upheaval

Wash

Figure 12 Flowslide occurred in May 1958 at Wissower KlinkenRugen Germany (54deg32prime09Prime N 13deg40prime42Prime E) Source fromHutchinson [36]

Advances in Civil Engineering 9

sliding body In this paper the Mohr-Coulomb criterion isused in the analysis of the formation of single and doubleupheaval as follows

τ c + σ tanφ (2)Taking the numbers① and② as two microelements in

the sliding body under a stress condition whereby themaximum principal stress σ1 of the microelement at theslope break is parallel to the direction of the inclined plateand the confining pressure σ2 σ3 is perpendicular to thedirection of the inclined plate (the microelements are smallenough so it is considered that σ2 σ3) shear behavioroccurs [22 30 31]

e angle between the shear plane and the maximumprincipal stress is defined as 45degminus (φ2) and the friction angleφ in the sand material is 36deg For a microelement ① atsmaller slope angles of 40deg and 50deg the counterclockwiseangles between the shear action direction and the overallmovement direction of the flowslide are 167deg and 157degrespectively tending to horizontal which is approximatelythe same as the direction of the maximum principal stress inthe state of the microelement ② which will be a source ofpower for the morphology of single upheaval Afterwardsthe microelements located on the horizontal plate movealong the shear plane at an angle of 27deg counterclockwisewith the movement direction of the sliding body on thehorizontal plate under the obstruction of the other particlesin the front and the push of rear particles similar to a thrustnappe in macroscopic performance resulting in depositmorphology as a single upheaval

e formation of double upheaval is more complicatedHere it will be described according to the order of the frontupheaval the wash and the back upheaval First based on thesame calculation method as for the smaller angles thecounterclockwise angles between the shear-action direction ofthe microelement ① and the movement direction of thesliding body on the horizontal plate are 93deg and 83deg at thelarger slope angles of 60deg and 70deg respectively At this pointthe sliding body deposits part of its load under the actions ofshear force and thrust force to form a rudiment of the frontupheaval while the part of the sliding body on the horizontalplate keeps moving under the inertial force and the thrust

force of the back part sliding body (the force state is similar tothe microelement②) and continues development of the frontupheaval gradually Next themicroelements at the back of thefront upheaval stop their shearing motion when the equi-librium condition of equation (3) is satisfied climbing isstopped and the wash is formed However mechanicalconditions still allow the microelements to move horizontallyWhen the equilibrium condition of equation (4) is satisfiedthe horizontal movement stops Assuming that the micro-elements are small enough the equilibrium condition can belisted as follows

m middot g τ middot sin 45deg minusφ2

1113874 1113875 (3)

m middot a τ middot cos 45deg minusφ2

1113874 1113875 (4)

where m is the mass of the microelement g is the accel-eration of gravity τ is shear stress a is the horizontal ac-celeration of the microelement and φ is the friction angle ofsand material

Eventually under the balance of equations (3) and (4)the sliding body that is still moving at the trailing edge is verysmall as the speed so the shear force provided can be re-duced accordingly At this time if the internal shearing forceof the sliding body at the trailing edge is greater than theshearing strength of the sliding body the subsequent slidingbody is deposited at the foot of the slope to form the backupheaval and conducts a localized shear motion based on theshear force similar to the microelement ① however if theshear force is less than the shear strength of the sliding bodythen the sliding body deposits at the slope break to form theback upheaval In general the order of the formation processof the double upheaval is front upheaval wash and backupheaval

e formation of double upheaval morphology is alsoowing to properties of the granular material In this studythere is a double upheaval morphology nevertheless this isnot in the study by Hu [14] e range of the granularparticle from 5times10minus 6m to 4times10minus 4m in this article is ob-viously less than the range of the granular particle from10minus 3m to 16times10minus 2m in the research by Hu [14] ereforethe internal friction angle and cohesion of the experimentalmaterial in this study are larger than those in the research byHu [14]emore the internal friction angle and cohesion ofgranular materials are the more the difficult to disturb thestate of the deposit morphology [19] is When the frontupheaval is formed the granular material with a particle sizeof 5times10minus 6m to 4times10minus 4m is more likely to keep its mor-phology of the front upheaval in the disturbance of the backsliding body compared with a particle size of 10minus 3m to16times10minus 2m If the backsliding body can not disturb theformed front upheaval it will accumulate in the slope breakforming the back upheaval meanwhile whereupon thedouble upheaval is formed

For these change laws as the slope angle increases themaximum width maximum length and area of the depositfirst increase and then decrease similar to the results by Hu[14] at the condition of the mass of fine particles changes

Phase 1Phase 2Phase 3

m1 m2

σ1

σ1

σ2

σ2

α

ατ1 τ1

12

Figure 13 Schematic illustration of shear force of flowslide

10 Advances in Civil Engineering

Based on those results a conclusion was proposed that there isan optimum proportion of fine and coarse particles in thegranular materials because the shear strength and frictioncoefficient are increasing with the mass of fine particles in-creases [41 42] In the condition that the internal frictionangle and cohesion are the smallest the maximum widthmaximum length and area are maximum Hence the de-duction mentioned before is reasonable that there should be acritical slope angle in the changes of the deposit morphology

36 Apparent Friction Coefficient To further understand therules that govern changes in the apparent friction coefficientwith slope angle and the distribution characteristics of theapparent friction coefficient in different lithologies and dif-ferent regions the apparent friction coefficients and corre-sponding slope angle data were collected for research fromthree places ere are 28 soil flowslides in the Jingyang Southplatform (34deg29prime32PrimeN 108deg50prime13Prime E) Shaanxi province China[30] consisting of loesse accumulation area is the even riverterrace comprised of the loess deposit underlain by the siltyclay deposit with interbedded sand layers occasionally Alongthe coast of northern Europe(51deg05prime03Prime N 1deg20prime36Prime E) thereare 27 chalk flows [22] with a porosity ranging from 9 to 52and dry density ranging from 2413 to 1265 kNm3 InWenchuan (31deg00prime00Prime N 103deg40prime00Prime E) Sichuan provinceChina [2] there are 44 rock flowslides that occurred in themountainous area and consisted of the sand-mudstonemainlye accumulation area is relatively flat though it is hard tocompare with flowslides that occurred in Jingyang and coast ofnorthern Europe An important reason why the above threeflowslide groups are taken as research objects is that they havesimilar flow characteristic and long runout distance In ad-dition the topographical condition of the accumulation area issimilar to the model test conditions It can be seen fromFigure 14 that the overall relationship between the apparentfriction coefficient of flowslide and the slope angles shows anonlinear growth relationship is conclusion is consistentwith Crosta [22] e growth relationship is as follows

H

L

tan θ1 + (cos θ + ε sin θ)2 middot ((tan θμ) minus 1)1113960 1113961

(5)

where θ represents the slope angle H represents the slopeheight L represents the moving distance of the flowslide ε isthe coefficient of velocity restitution when the sliding bodyreaches the slope break to change its moving direction tohorizontal μ is the friction coefficient

e difference in regional lithology determines thedifference between the friction coefficient μ and the co-efficient of velocity restitution ε so two different μ valuesand three different ε values are set in this paper When μ is02 and ε is 0 02 or 04 the curve better reflects theoverall trend of the three natural flowslide groups Whenthe friction coefficient μ between the inclined plate andthe experimental material is 055 and ε is 04 the curveshows good correlation with the experimental points Itcan be said that the friction coefficient determines theheight of the initial benchmark of the curve e larger thevalue of the friction coefficient the higher the initial

benchmark of the curve e coefficient of velocity res-titution determines the change of the inflection pointwhen the curve changes from flat to rapid e larger thevalue of the coefficient of velocity restitution the largerthe horizontal coordinates of the inflection point

From the point of energy conversion energy loss at theslope break is closely related to terrain and lithology ekinetic energy loss is small when the junction of slope and theaccumulation area is a curved slope break when the junctionhas sharp edges the kinetic energy loss is larger [13 22]Under the experimental conditions in which slope height Hand volume are constant as the slope angle increases energyloss also increases due to the collision of the flowslide and thehorizontal plate at the slope break [43 44] which results inshorter distance of movement of the sliding body e slopeangle in the chalk flow area is too large and the energy loss isexcessive though the sliding path is lubricated by water be-neath the accumulation area 1-2m e flowslides in Wen-chuan affected by lubrication due to water are few Besidesthere are many obstacles in the movement path because theseflowslides are located in a mountainous area though theaccumulation area is relatively flat so energy loss is largeHowever for the flowslides in Jingyang the path is flat andstraight as well as lubricated by the liquefied effect [31] andthe energy loss is small e flowslides in the model exper-iment fall between the three natural flowslide groups to theleft of chalk flows above loess flowslides and intersectingwith rock flowslides e apparent friction coefficient in thecondition of this model experiment is not the smallest al-though the condition is ideal flat for the reason that there areno lubrications erefore the lubrication effect plays a keyrole in affecting the apparent friction coefficient

4 Discussion

In this paper the Taguchi method was employed inno-vatively to study quantitatively the order of significance

0 20 40 60 80 10000

05

10

15

20

Wenchuan rock flowslidesJingyang loess flowslides

Chalk flowsExperimental flowslides

HL

Slope angle (deg)

ε = 0

ε = 02

ε = 0

2

ε = 04

ε = 0

4

ε = 0μ = 055

μ = 02

tan θ1 + [(cos θ + εsin θ)2 middot (tan θμ ndash1)]

ndashHL =

Figure 14 Comprehensive relationship between slope angle andapparent friction coefficient Formula comes from Crosta [22]

Advances in Civil Engineering 11

and the maximal responses of combinations of factorsaffecting flowslide deposit morphology e calculationresults in Figure 5 confirm that slope angle is the mostinfluential factor e maximum influencing combinationis a volume of 54 times10minus 3 m3 at a slope angle of 70deg and aslope height of 090m which will be useful to evaluate thehazards of potential flowslides Based on the acceptedknowledge that slope angle is the largest influencingfactor further experimental research was performed econditions of small slope angles and large slope anglesshow interesting phenomena of single upheaval anddouble upheaval respectively the Mohr-Coulomb crite-rion is used to explain the action of the shear motionduring the process e comprehensive study of apparentfriction coefficient shows a nonlinear increase with slopeangle as well as spatial and lithological differences indistribution which are affected by the friction coefficientand the coefficient of velocity restitution

Similar model experiments have been carried out toanalyze the movement rule of flowslide [4 12 21] anddeposit morphology [3 16 18 22 25] Previous studiesfocused on the influence of a single factor and the researchcontent is more in-depth [8 11 45 46] but the researchprocess is not systematic enough as flowslide movement anddeposit morphology are determined by multiple factorsincluding volume slope angle slope height sliding bodymaterial and slip bed material Of course some scholarshave conducted multifactor experiments for exampleManzella [13] studied the influence of flowslide volumeinitial height slope of slip bed and the slope break onmovement characteristics but the relative significance ofvarious factors governing the movement of flowslide has notbeen revealed although the factors considered are com-prehensive An orthogonal experiment was used to study themotion characteristics (velocity acceleration) of flowslideunder the influence of various factors and found that slopeangle was the largest influencing factor [33] which isconsistent with the findings of this article though the re-search content was different

e morphology of the double upheaval of flowslide wasalso observed in an experiment similar to the current study[1] which examined wave propagation in the macroscopicview at experiment was also carried out with bulk ma-terials A wave would be formed when the experimentalmaterials slid or fell into the accumulation area the initialwave would accelerate suddenly as it was pushed by asubsequent wave propagating forward e wave crestformed when motion stopped due to resistance may be oneof the reasons for the formation of upheaval and the wavetrough can be regarded as wash [1] in this paper Accordingto the motion rule of the object the flowslide is partiallyreflected at the slope break [22] and then propagates in theform of waves which were observed somewhat during theexperiment of larger slope angles Macroscopic wavepropagation may be the cause of the formation of upheavaland wash on the surface of flowslide but if the phenomenonis deeply understood from the interior of sliding body theinternal shear behavior in the movement of flowslide may bereasonable In addition the morphology of actual flowslides

is also affected by properties of sliding materials [14 19 20]With an increase of the mass of fine particles the length andwidth of the flowslide deposit increase first and then de-crease [14] erefore the difference of properties ofgranular materials is contributing to the formation of thedouble upheaval

e apparent friction coefficient is often involved in thestudy of flowslide but the focus here is on its relationshipwith volume [5 7 9 21 38] ere are few studies that useslope angle as a variable including Lied and Bakkehoi [47]and Okura [3] both of which proposed that the apparentfriction coefficient of flowslide was directly proportional tothe slope angle but the results were based on the analysis ofonly about 20 flowslides so the conclusion was limited eanalysis in this paper involved 99 landslides in three areasand the conclusion is consistent with the viewpoint of Crosta[22] in finding that the apparent friction coefficient increasesnonlinearly as the slope angle increases In the comparisonmade in this study spatial and lithological differences in theapparent friction coefficient are also found

In this paper an experimental idea of constant height ofthe center of gravity is provided but when the angle changesthe sliding distance of the flowslide on the inclined plate(Figure 4(b)) will inevitably increase and the loss of kineticenergy due to friction will increase accordingly ereforefurther improvement of the experimental device is needed toreduce the kinetic energy loss due to the increase in thesliding distance on the inclined plate Modeling should alsoinclude more influencing factors

5 Conclusions

(1) e combined influence factor obtained by theTaguchi method demonstrates that when flowslidevolume is 54times10minus 3m3 the slope angle is 70deg theslope height is 090m and the changes of depositmorphology of the flowslide is the largest e ex-treme differences of maximum length (Ld) maxi-mum thickness (Dd) area (A) and the length-widthratio (LdWd) at three different levels under variousfactors are 277 551 454 and 17 respectively ecalculation results indicate that the factor that has thegreatest impact on the deposit morphology offlowslide is slope angle followed by volume andslope height

(2) With an increase in slope angle the width lengtharea and length-width ratio of flowslide deposits firstincrease and then decrease and the maximum ac-cumulation thickness first decreases and then in-creases And there should be a critical angle in thechange laws of the flowslide deposit morphologythat is between 60deg and 70deg under the experimentalconditions When the volume is 36times10minus 3m3 andthe slope height is 090m the slope angles at 40deg and50deg both result in a single upheaval while slopeangles of 60deg and 70deg both result in a double upheavalwhich has been explained by shear behavior in thesliding body based on the Mohr-Coulomb criterion

12 Advances in Civil Engineering

(3) e apparent friction coefficient of flowslide in-creases nonlinearly as the slope angle increases andshows spatial and lithological difference e greaterthe friction coefficient of the sliding material thehigher the initial benchmark of the curve the largerthe coefficient of velocity restitution the larger theabscissa corresponding to the position of inflectionpoint when the curve changes from flat to rapid

Data Availability

e data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

e authors declare that they have no conflicts of interest

Acknowledgments

is study would not have been possible without financialsupport from the Special Fund for the National NaturalScience Foundation of China under Grant nos 4170229841790442 and 41602359 as well as the Project Supported byNatural Science Basic Research Plan in Shaanxi Province ofChina under Grant no 2017JQ4020

Supplementary Materials

In this section the data on the slope angles and the apparentfriction coefficients are provided as a formation of tableese data are related to 44 rock flowslides in theWenchuanSichuan province China 28 soil flowslides in in the JingyangSouth platform Shaanxi province China and 27 chalkflowslides along the coast of northern Europe (Supple-mentary Materials)

References

[1] O Roche M Attali A Mangeney and A Lucas ldquoOn the run-out distance of geophysical gravitational flows insight fromfluidized granular collapse experimentsrdquo Earth PlanetaryScience Letters vol 311 no 3-4 pp 375ndash385 2011

[2] D Guo M Hamada C He Y Wang and Y Zou ldquoAnempirical model for landslide travel distance prediction inWenchuan earthquake areardquo Landslides vol 11 no 2pp 281ndash291 2014

[3] Y Okura H Kitahara A Kawanami and U KurokawaldquoTopography and volume effects on travel distance of surfacefailurerdquo Engineering Geology vol 67 no 3-4 pp 243ndash2542003

[4] Z Han Y Li J Huang et al ldquoNumerical simulation for run-out extent of debris flows using an improved cellular au-tomaton modelrdquo Bulletin of Engineering Geology and theEnvironment vol 76 no 3 pp 961ndash974 2017

[5] F Guzzeti F Ardizzone M Cardinali M Rossi andD Valigi ldquoLandslide volumes and landslide mobilizationrates in Umbria Central Italyrdquo Earth Planetary Science Let-ters vol 279 no 3-4 pp 222ndash229 2009

[6] J D Chaytor S Uri A R Solow and B Andrews ldquoSizedistribution of submarine landslides along the US Atlanticmarginrdquo Marine Geology vol 264 no 1-2 pp 16ndash27 2009

[7] H Qiu P Cui A D Regmi et al ldquoInfluence of topographyand volume on mobility of loess slides within different slipsurfacesrdquo Catena vol 157 pp 180ndash188 2017

[8] J Hu J Peng G Wang I Javed Y Wang and W LildquoDistribution and characteristics of landslide in LoessPlateau a case study in Shaanxi provincerdquo EngineeringGeology vol 236 pp 89ndash96 2018

[9] L Xu F C Dai Q M Gong L G am and H MinldquoIrrigation-induced loess flow failure in Heifangtai Platformnorth-west Chinardquo Environmental Earth Sciences vol 66no 6 pp 1707ndash1713 2012

[10] R Q Huang W H Liu J P Zhou and X J Pei ldquoRolling testson movement characteristics of rock blocksrdquo Journal ofGeotechnical Engineering vol 29 no 9 pp 1296ndash1302 2007in Chinese

[11] I Manzella and V Labiouse ldquoPhysical modelling to betterunderstand rock avalanchesrdquo in Physical Modelling in Geo-technics Two Volume Set Proceedings of the 7th InternationalConference on Physical Modelling in Geotechnics (ICPMG2010) vol 2 pp 1259ndash1265 Taylor amp Francis Group LondonUK 2010

[12] G B Crosta F V De Blasio M De Caro G VolpiS Imposimato and D Roddeman ldquoModes of propagationand deposition of granular flows onto an erodible substrateexperimental analytical and numerical studyrdquo Landslidesvol 14 no 1 pp 47ndash68 2017

[13] I Manzella and V Labiouse ldquoEmpirical and analytical ana-lyses of laboratory granular flows to investigate rock ava-lanche propagationrdquo Landslides vol 10 no 1 pp 23ndash362013

[14] Y-X Hu H-B Li S-C Qi G Fan and J-W ZhouldquoGranular effects on depositional processes of debris ava-lanchesrdquo KSCE Journal of Civil Engineering vol 24 no 4pp 1116ndash1127 2020

[15] I Goktepeli U Atmaca and A Cakan ldquoInvestigation of heattransfer augmentation between the ribbed plates via Taguchiapproach and computational fluid dynamicsrdquo Journal ofEermal Science vol 29 no 3 pp 647ndash666 2020

[16] M Farin A Mangeney and O Roche ldquoFundamental changesof granular flow dynamics deposition and erosion processesat high slope angles insights from laboratory experimentsrdquoJournal of Geophysical Research Earth Surface vol 119 no 3pp 504ndash532 2014

[17] M Kesseler V Heller and B Turnbull ldquoA laboratory-nu-merical approach for modelling scale effects in dry granularslidesrdquo Landslides vol 15 no 11 pp 2145ndash2159 2018

[18] A Dufresne and T R Davies ldquoLongitudinal ridges in massmovement depositsrdquo Geomorphology vol 105 no 3-4pp 171ndash181 2009

[19] T De Haas L Braat J R F W Leuven I R Lokhorst andM G Kleinhans ldquoEffects of debris flow composition onrunout depositional mechanisms and deposit morphology inlaboratory experimentsrdquo Journal of Geophysical ResearchEarth Surface vol 120 no 9 pp 1949ndash1972 2015

[20] C Goujon N omas and B Dalloz-Dubrujeaud ldquoMono-disperse dry granular flows on inclined planes role ofroughnessrdquo Ee European Physical Journal E vol 11 no 2pp 147ndash157 2003

[21] X-Y Fan S-J Tian and Y-Y Zhang ldquoMass-front velocity ofdry granular flows influenced by the angle of the slope to therunout plane and particle size gradationrdquo Journal of MountainScience vol 13 no 2 pp 234ndash245 2016

[22] G B Crosta F V D Blasio M Locatelli S Imposimato andD Roddeman ldquoLandslides falling onto a shallow erodible

Advances in Civil Engineering 13

the front and back and are nearly round upheaval and washare arranged alternately on the surface At 60deg the lengthwidth length-width ratio and area are 090m 090m 10and 057m2 respectively At 70deg the length width length-width ratio and area are 074m 075m 099 and 042m2respectively e morphologies of the flowslides formed ateach of the four angles are symmetrically distributed aboutthe sliding central axis

Comparing the shape of the deposits formed at each ofthe four angles we can see that with an increase in slopeangle the values of Ld Wd A and (LdWd) for the flowslidedeposit first increase and then decrease e main reason isthat by increasing slope angle from 40deg to 60deg the sum ofkinetic energy loss by the sliding body due to friction andcollisional energy loss on the slope break is reducedHowever in increasing the slope angle from 60deg to 70deg thesum of the kinetic energy loss and the collisional energy lossincreases In the experiment the slope height was kept at aconstant value of 090m to each of the four angles so theinitial potential energy was the same for each (Figure 4(b))

e smaller the slope angle the greater the moving dis-placement of the sliding body on the inclined plate resultingin greater kinetic energy loss caused by friction When theflowslide flows to a slope break where there is a change in thedirection of motion to a horizontal direction and kineticenergy loss will inevitably occur kinetic energy loss ispositively related to the slope angle [32] Under the con-dition of changing angles the increase in the length of theflowslide deposit is significantly higher than the width isis explained by the feature of straightness [33] which theflowslide has during high-speed movement

ese laboratory findings can be examined in the contextof flowslides that occurred naturally In Xihetan Shaanxiprovince China a low-angle flowslide deposit formed at a45deg slope angle and had a length of 289m a width of 329mand a length-width ratio of 088 (Figure 7) Another low-angle deposit formed in Shenzhen China [34 35] on a slopeangle of 24deg with a length of 700m a width of about 550mand a length-width ratio of about 127 (Figure 8) A largeslope angle (more than 60deg) slide on the steep cliffs along thecoast of Northern Europe is shown in Figure 9e followingare shown in Figure 9(a) the deposit with a length 330m inFigure 9(b) the deposit with a length 150m in Figure 9(c)the deposit with a length 150m and width 120m inFigure 9(d) the deposit with a length 270m and width300m all of these deposits with a length-width ratio close to1 alternating upheaval and wash on the surface of thedeposit and an overall circular shape (Figure 9) [36 37] eactual flowslides at small angles and large angles are similarto the deposit morphology and state of motion formedduring the current experiment However due to the dif-ference in topography and the scale of the actual flowslides itis difficult to compare in terms of dimensional parameterssuch as the length and width of the deposit leaving thedimensionless length-width ratio surface morphology andoverall shape for comparison In both the natural and themodeled deposit morphology a flat surface and a generally

Table 4 e main factors of influence

Morphological elements of deposit Influencing factors Level1 (dB) Level2 (dB) Level3 (dB) Extreme difference (dB) Influence order

Ld (m)A 3675 3652 3869 216 2B 3556 3833 3808 278 1C 3697 3728 3772 075 3

Wd (m)A 3682 3825 3965 278 1B 3751 3858 3864 108 2C 3795 3816 3866 071 3

Dd (m)A 482 728 1003 521 2B 1025 474 716 551 1C 931 691 593 338 3

A (m2)A minus 752 minus 637 minus 362 390 2B minus 866 minus 411 minus 473 454 1C minus 627 minus 593 minus 531 096 3

(LdWd) (1)A minus 007 minus 172 minus 097 165 2B minus 195 minus 025 minus 561 170 1C minus 098 minus 090 minus 089 009 3

Note A B and C denote flowslide volume slope angle and height respectively Ld maximum sliding length of flowslide deposit on the surface Wdmaximum width of flowslide deposit Dd maximum depth of flowslide deposit A area of flowslide deposit projected on horizontal plane θ slope angle(LdWd) length-width ratio

A B C A B C A B C A B C A B C

0

1

2

3

4

5

6

Influencing factors

Extr

eme d

iffer

ence

(dB)

00

02

04

06

08

10

Ld Wd Dd LdWdA

Patr

ial η

2

Figure 5 Statistical results using Taguchi method and multivariateanalysis of variance

6 Advances in Civil Engineering

tongue-shaped deposit form at low angles while alternatingupheaval and wash on the deposit surface and a generallycircle-shaped deposit form at large angles

34 Profile Morphology at Different AnglesMultidimensional analysis of the morphology of flowslidedeposit can help researchers gain a deeper understanding ofmorphology characteristics Figure 10 shows the profiles ofthe four flowslides depicted in Figure 6 Combining theprofile (Figure 10) with planar morphology (Figure 6) it canbe seen that at smaller slope angles of 40deg and 50deg a generallysingle upheaval with heights of 0027m and 0018m can berecognized with corresponding peak positions of 028m and050m respectively although the surface of each deposit isslightly undulated e position of the center of gravity isconsistent with the peak position of the upheaval At largerslope angles of 60deg and 70deg the surfaces of the deposits areundulated and there is double upheaval e maximumheight differences between adjacent upheaval and wash are0005m and 0023m respectively e heights of the frontupheaval are 0009m and 0026m with corresponding peakpositions of 067m and 040m respectively e heights ofthe back upheaval are 0012m and 0035m with corre-sponding peak positions of 010m and 006m respectivelye heights of the middle wash are 0006m and 0011m withcorresponding positions of 041m and 022m respectivelye positions of the center of gravity are relatively consistentwith the position of the middle wash

Comparing the four profiles shows a single upheavalshape at smaller angles of 40deg and 50deg and a double upheavalshape at larger angles of 60deg and 70deg In the flowslides withdouble upheaval the height of the front upheaval is smallerthan that of the back upheaval As the slope angle increasesthe average thickness of the deposit first decreases and thenincreases and the position of the center of gravity firstadvances and then retreats Based on the above comparisonit can be considered that there should be a key angle in thechange laws of flowslide deposit morphology If the slopeangle is less than or larger than the key angle under the sameconditions the length and width of the deposit will decreasee key angle is between 60deg and 70deg under the experimentalconditions in the paper However actually the key angle offlowslides should be different values depending on thematerial of the sliding body and the terrain conditions

e profile figures of chalk flows [36] on the steep cliffs ofthe European coast show similarities with the profilemorphology of the sandbox experiment for the larger anglesespecially the feature of double upheaval (Figures 11 and 12)ese two deposits shown in Figures 11 and 12 are 320mand 200m in length respectively It can be seen from thepicture that the leading edge of the flowslide deposit is flatand the surface shows double upheaval e front upheavalis lower than the back upheaval e height difference be-tween the middle wash and the back upheaval is greater thanthat of the front upheaval e center of gravity is locatedslightly in front of the middle wash It is noticed that there isa difference although the morphology of the double

Table 5 Variance analysis statistics

Morphological elements of deposit Factors Quadratic sum DOF Mean square error Partial η2

Ld (m)

A 68628 2 34314 055B 98732 2 49366 064C 597 2 2985 01

Error 56458 2 28229Total 229787 8

Wd (m)

A 108303 2 54151 087B 22448 2 11224 059C 4897 2 2449 024

Error 15835 2 7917Total 151483 8

Dd (m)

A 289 2 144 076B 342 2 171 079C 073 2 037 045

Error 091 2 045Total 795 8

A (m2)

A 011 2 006 052B 013 2 007 056C 001 2 0 006

Error 011 2 005Total 036 8

(LdWd) (1)

A 004 2 002 069B 005 2 003 073C 0 2 0 005

Error 002 2 001Total 011 8

Note Ld maximum sliding length of flowslide deposit on the surface Wd maximum width of flowslide deposit Dd maximum depth of flowslide deposit Aarea of flowslide deposit projected on horizontal plane θ slope angle (LdWd) length-width ratio

Advances in Civil Engineering 7

upheaval is the same e center of gravity of flowslidedeposit is located in the middle wash in experimentalconditions However the center of gravity is located in frontof the middle wash in actual flowslides e possible reasonmainly is the difference of properties of granular materialsbecause the movement and accumulation process of flow-slides are also affected by internal factors such as the particlesize and roundness of sliding body materials [14]

35 Mechanical Analysis of the Deposit MorphologyBefore the mechanical analysis the properties of themedium-fine sand and plexiglass should be noticed as abasis for this analysis e friction coefficient 0727 of themedium-fine sand is larger than 0538 of the interface of

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

aprime

0

6

0

6121812

918120 6

(a)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

bprime0 4

8

80

412

0

04812

4

(b)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

cprime

0 46

28

8

9

9

68

4

2

02

6

84

0

20 468

8

10

8

(c)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

dprime

04

1216

20

242016128

24

16 12

048

04816

12

08

4

(d)

Figure 6 Contour map of flowslides topography at different slope angles (a) (b) (c) and (d) corresponding to 40deg 50deg 60deg and 70deg respectively

N

Figure 7 Xi Hetan flowslide in China (34deg29prime32PrimeN 108deg50prime13Prime E)

N

Figure 8 Shenzhen flowslide Source Aerial image form Sou-hucom in China (22deg43prime04Prime N 113deg56prime31Prime E)

8 Advances in Civil Engineering

the plexiglass with the medium-fine sand e differenceof the friction coefficient between the interface and thesliding body itself also existed in actual flowslides becausethe interface will be lubricated by water or the liquefactioneffect [30 38] erefore the sliding body tends to flowalong the interface of the plexiglass and the sand forms athin sand layer like an erodible bed before the main slidingbody arrived on both inclined and horizontal plates[39 40] It provides an essential condition of the shearaction Figure 13 shows the cross-sectional state of thesliding body at three phases in which phase 3 indicatedby a solid line represents the final deposit state of the

N

(a)

N

(b)

N

(c)

N

(d)

Figure 9 Topography of typical flowslides (a) and (b) located in Folkestone Warren (51deg05prime03Prime N 1deg20prime36Prime E) (c) and (d) located DoverKent (51deg07prime53Prime N 1deg20prime36Prime E) Source from Hutchinson [36]

0 200 400 600 800 10000

10

20

30

40

Dep

th o

f dep

osit

(mm

)

Length of deposit (mm)

40degprofile50degprofile

60degprofile70degprofile

Figure 10 e deposit profiles of flowslides at different angles

N

Front upheavalFront upheavalBack upheavalWash

Figure 11 Flowslide occurred in May 1981 at Le Chien Neuf(49deg46prime59Prime N 0deg25prime19Prime E) Source from Hutchinson [36]

N

Front upheavalBack upheaval

Wash

Figure 12 Flowslide occurred in May 1958 at Wissower KlinkenRugen Germany (54deg32prime09Prime N 13deg40prime42Prime E) Source fromHutchinson [36]

Advances in Civil Engineering 9

sliding body In this paper the Mohr-Coulomb criterion isused in the analysis of the formation of single and doubleupheaval as follows

τ c + σ tanφ (2)Taking the numbers① and② as two microelements in

the sliding body under a stress condition whereby themaximum principal stress σ1 of the microelement at theslope break is parallel to the direction of the inclined plateand the confining pressure σ2 σ3 is perpendicular to thedirection of the inclined plate (the microelements are smallenough so it is considered that σ2 σ3) shear behavioroccurs [22 30 31]

e angle between the shear plane and the maximumprincipal stress is defined as 45degminus (φ2) and the friction angleφ in the sand material is 36deg For a microelement ① atsmaller slope angles of 40deg and 50deg the counterclockwiseangles between the shear action direction and the overallmovement direction of the flowslide are 167deg and 157degrespectively tending to horizontal which is approximatelythe same as the direction of the maximum principal stress inthe state of the microelement ② which will be a source ofpower for the morphology of single upheaval Afterwardsthe microelements located on the horizontal plate movealong the shear plane at an angle of 27deg counterclockwisewith the movement direction of the sliding body on thehorizontal plate under the obstruction of the other particlesin the front and the push of rear particles similar to a thrustnappe in macroscopic performance resulting in depositmorphology as a single upheaval

e formation of double upheaval is more complicatedHere it will be described according to the order of the frontupheaval the wash and the back upheaval First based on thesame calculation method as for the smaller angles thecounterclockwise angles between the shear-action direction ofthe microelement ① and the movement direction of thesliding body on the horizontal plate are 93deg and 83deg at thelarger slope angles of 60deg and 70deg respectively At this pointthe sliding body deposits part of its load under the actions ofshear force and thrust force to form a rudiment of the frontupheaval while the part of the sliding body on the horizontalplate keeps moving under the inertial force and the thrust

force of the back part sliding body (the force state is similar tothe microelement②) and continues development of the frontupheaval gradually Next themicroelements at the back of thefront upheaval stop their shearing motion when the equi-librium condition of equation (3) is satisfied climbing isstopped and the wash is formed However mechanicalconditions still allow the microelements to move horizontallyWhen the equilibrium condition of equation (4) is satisfiedthe horizontal movement stops Assuming that the micro-elements are small enough the equilibrium condition can belisted as follows

m middot g τ middot sin 45deg minusφ2

1113874 1113875 (3)

m middot a τ middot cos 45deg minusφ2

1113874 1113875 (4)

where m is the mass of the microelement g is the accel-eration of gravity τ is shear stress a is the horizontal ac-celeration of the microelement and φ is the friction angle ofsand material

Eventually under the balance of equations (3) and (4)the sliding body that is still moving at the trailing edge is verysmall as the speed so the shear force provided can be re-duced accordingly At this time if the internal shearing forceof the sliding body at the trailing edge is greater than theshearing strength of the sliding body the subsequent slidingbody is deposited at the foot of the slope to form the backupheaval and conducts a localized shear motion based on theshear force similar to the microelement ① however if theshear force is less than the shear strength of the sliding bodythen the sliding body deposits at the slope break to form theback upheaval In general the order of the formation processof the double upheaval is front upheaval wash and backupheaval

e formation of double upheaval morphology is alsoowing to properties of the granular material In this studythere is a double upheaval morphology nevertheless this isnot in the study by Hu [14] e range of the granularparticle from 5times10minus 6m to 4times10minus 4m in this article is ob-viously less than the range of the granular particle from10minus 3m to 16times10minus 2m in the research by Hu [14] ereforethe internal friction angle and cohesion of the experimentalmaterial in this study are larger than those in the research byHu [14]emore the internal friction angle and cohesion ofgranular materials are the more the difficult to disturb thestate of the deposit morphology [19] is When the frontupheaval is formed the granular material with a particle sizeof 5times10minus 6m to 4times10minus 4m is more likely to keep its mor-phology of the front upheaval in the disturbance of the backsliding body compared with a particle size of 10minus 3m to16times10minus 2m If the backsliding body can not disturb theformed front upheaval it will accumulate in the slope breakforming the back upheaval meanwhile whereupon thedouble upheaval is formed

For these change laws as the slope angle increases themaximum width maximum length and area of the depositfirst increase and then decrease similar to the results by Hu[14] at the condition of the mass of fine particles changes

Phase 1Phase 2Phase 3

m1 m2

σ1

σ1

σ2

σ2

α

ατ1 τ1

12

Figure 13 Schematic illustration of shear force of flowslide

10 Advances in Civil Engineering

Based on those results a conclusion was proposed that there isan optimum proportion of fine and coarse particles in thegranular materials because the shear strength and frictioncoefficient are increasing with the mass of fine particles in-creases [41 42] In the condition that the internal frictionangle and cohesion are the smallest the maximum widthmaximum length and area are maximum Hence the de-duction mentioned before is reasonable that there should be acritical slope angle in the changes of the deposit morphology

36 Apparent Friction Coefficient To further understand therules that govern changes in the apparent friction coefficientwith slope angle and the distribution characteristics of theapparent friction coefficient in different lithologies and dif-ferent regions the apparent friction coefficients and corre-sponding slope angle data were collected for research fromthree places ere are 28 soil flowslides in the Jingyang Southplatform (34deg29prime32PrimeN 108deg50prime13Prime E) Shaanxi province China[30] consisting of loesse accumulation area is the even riverterrace comprised of the loess deposit underlain by the siltyclay deposit with interbedded sand layers occasionally Alongthe coast of northern Europe(51deg05prime03Prime N 1deg20prime36Prime E) thereare 27 chalk flows [22] with a porosity ranging from 9 to 52and dry density ranging from 2413 to 1265 kNm3 InWenchuan (31deg00prime00Prime N 103deg40prime00Prime E) Sichuan provinceChina [2] there are 44 rock flowslides that occurred in themountainous area and consisted of the sand-mudstonemainlye accumulation area is relatively flat though it is hard tocompare with flowslides that occurred in Jingyang and coast ofnorthern Europe An important reason why the above threeflowslide groups are taken as research objects is that they havesimilar flow characteristic and long runout distance In ad-dition the topographical condition of the accumulation area issimilar to the model test conditions It can be seen fromFigure 14 that the overall relationship between the apparentfriction coefficient of flowslide and the slope angles shows anonlinear growth relationship is conclusion is consistentwith Crosta [22] e growth relationship is as follows

H

L

tan θ1 + (cos θ + ε sin θ)2 middot ((tan θμ) minus 1)1113960 1113961

(5)

where θ represents the slope angle H represents the slopeheight L represents the moving distance of the flowslide ε isthe coefficient of velocity restitution when the sliding bodyreaches the slope break to change its moving direction tohorizontal μ is the friction coefficient

e difference in regional lithology determines thedifference between the friction coefficient μ and the co-efficient of velocity restitution ε so two different μ valuesand three different ε values are set in this paper When μ is02 and ε is 0 02 or 04 the curve better reflects theoverall trend of the three natural flowslide groups Whenthe friction coefficient μ between the inclined plate andthe experimental material is 055 and ε is 04 the curveshows good correlation with the experimental points Itcan be said that the friction coefficient determines theheight of the initial benchmark of the curve e larger thevalue of the friction coefficient the higher the initial

benchmark of the curve e coefficient of velocity res-titution determines the change of the inflection pointwhen the curve changes from flat to rapid e larger thevalue of the coefficient of velocity restitution the largerthe horizontal coordinates of the inflection point

From the point of energy conversion energy loss at theslope break is closely related to terrain and lithology ekinetic energy loss is small when the junction of slope and theaccumulation area is a curved slope break when the junctionhas sharp edges the kinetic energy loss is larger [13 22]Under the experimental conditions in which slope height Hand volume are constant as the slope angle increases energyloss also increases due to the collision of the flowslide and thehorizontal plate at the slope break [43 44] which results inshorter distance of movement of the sliding body e slopeangle in the chalk flow area is too large and the energy loss isexcessive though the sliding path is lubricated by water be-neath the accumulation area 1-2m e flowslides in Wen-chuan affected by lubrication due to water are few Besidesthere are many obstacles in the movement path because theseflowslides are located in a mountainous area though theaccumulation area is relatively flat so energy loss is largeHowever for the flowslides in Jingyang the path is flat andstraight as well as lubricated by the liquefied effect [31] andthe energy loss is small e flowslides in the model exper-iment fall between the three natural flowslide groups to theleft of chalk flows above loess flowslides and intersectingwith rock flowslides e apparent friction coefficient in thecondition of this model experiment is not the smallest al-though the condition is ideal flat for the reason that there areno lubrications erefore the lubrication effect plays a keyrole in affecting the apparent friction coefficient

4 Discussion

In this paper the Taguchi method was employed inno-vatively to study quantitatively the order of significance

0 20 40 60 80 10000

05

10

15

20

Wenchuan rock flowslidesJingyang loess flowslides

Chalk flowsExperimental flowslides

HL

Slope angle (deg)

ε = 0

ε = 02

ε = 0

2

ε = 04

ε = 0

4

ε = 0μ = 055

μ = 02

tan θ1 + [(cos θ + εsin θ)2 middot (tan θμ ndash1)]

ndashHL =

Figure 14 Comprehensive relationship between slope angle andapparent friction coefficient Formula comes from Crosta [22]

Advances in Civil Engineering 11

and the maximal responses of combinations of factorsaffecting flowslide deposit morphology e calculationresults in Figure 5 confirm that slope angle is the mostinfluential factor e maximum influencing combinationis a volume of 54 times10minus 3 m3 at a slope angle of 70deg and aslope height of 090m which will be useful to evaluate thehazards of potential flowslides Based on the acceptedknowledge that slope angle is the largest influencingfactor further experimental research was performed econditions of small slope angles and large slope anglesshow interesting phenomena of single upheaval anddouble upheaval respectively the Mohr-Coulomb crite-rion is used to explain the action of the shear motionduring the process e comprehensive study of apparentfriction coefficient shows a nonlinear increase with slopeangle as well as spatial and lithological differences indistribution which are affected by the friction coefficientand the coefficient of velocity restitution

Similar model experiments have been carried out toanalyze the movement rule of flowslide [4 12 21] anddeposit morphology [3 16 18 22 25] Previous studiesfocused on the influence of a single factor and the researchcontent is more in-depth [8 11 45 46] but the researchprocess is not systematic enough as flowslide movement anddeposit morphology are determined by multiple factorsincluding volume slope angle slope height sliding bodymaterial and slip bed material Of course some scholarshave conducted multifactor experiments for exampleManzella [13] studied the influence of flowslide volumeinitial height slope of slip bed and the slope break onmovement characteristics but the relative significance ofvarious factors governing the movement of flowslide has notbeen revealed although the factors considered are com-prehensive An orthogonal experiment was used to study themotion characteristics (velocity acceleration) of flowslideunder the influence of various factors and found that slopeangle was the largest influencing factor [33] which isconsistent with the findings of this article though the re-search content was different

e morphology of the double upheaval of flowslide wasalso observed in an experiment similar to the current study[1] which examined wave propagation in the macroscopicview at experiment was also carried out with bulk ma-terials A wave would be formed when the experimentalmaterials slid or fell into the accumulation area the initialwave would accelerate suddenly as it was pushed by asubsequent wave propagating forward e wave crestformed when motion stopped due to resistance may be oneof the reasons for the formation of upheaval and the wavetrough can be regarded as wash [1] in this paper Accordingto the motion rule of the object the flowslide is partiallyreflected at the slope break [22] and then propagates in theform of waves which were observed somewhat during theexperiment of larger slope angles Macroscopic wavepropagation may be the cause of the formation of upheavaland wash on the surface of flowslide but if the phenomenonis deeply understood from the interior of sliding body theinternal shear behavior in the movement of flowslide may bereasonable In addition the morphology of actual flowslides

is also affected by properties of sliding materials [14 19 20]With an increase of the mass of fine particles the length andwidth of the flowslide deposit increase first and then de-crease [14] erefore the difference of properties ofgranular materials is contributing to the formation of thedouble upheaval

e apparent friction coefficient is often involved in thestudy of flowslide but the focus here is on its relationshipwith volume [5 7 9 21 38] ere are few studies that useslope angle as a variable including Lied and Bakkehoi [47]and Okura [3] both of which proposed that the apparentfriction coefficient of flowslide was directly proportional tothe slope angle but the results were based on the analysis ofonly about 20 flowslides so the conclusion was limited eanalysis in this paper involved 99 landslides in three areasand the conclusion is consistent with the viewpoint of Crosta[22] in finding that the apparent friction coefficient increasesnonlinearly as the slope angle increases In the comparisonmade in this study spatial and lithological differences in theapparent friction coefficient are also found

In this paper an experimental idea of constant height ofthe center of gravity is provided but when the angle changesthe sliding distance of the flowslide on the inclined plate(Figure 4(b)) will inevitably increase and the loss of kineticenergy due to friction will increase accordingly ereforefurther improvement of the experimental device is needed toreduce the kinetic energy loss due to the increase in thesliding distance on the inclined plate Modeling should alsoinclude more influencing factors

5 Conclusions

(1) e combined influence factor obtained by theTaguchi method demonstrates that when flowslidevolume is 54times10minus 3m3 the slope angle is 70deg theslope height is 090m and the changes of depositmorphology of the flowslide is the largest e ex-treme differences of maximum length (Ld) maxi-mum thickness (Dd) area (A) and the length-widthratio (LdWd) at three different levels under variousfactors are 277 551 454 and 17 respectively ecalculation results indicate that the factor that has thegreatest impact on the deposit morphology offlowslide is slope angle followed by volume andslope height

(2) With an increase in slope angle the width lengtharea and length-width ratio of flowslide deposits firstincrease and then decrease and the maximum ac-cumulation thickness first decreases and then in-creases And there should be a critical angle in thechange laws of the flowslide deposit morphologythat is between 60deg and 70deg under the experimentalconditions When the volume is 36times10minus 3m3 andthe slope height is 090m the slope angles at 40deg and50deg both result in a single upheaval while slopeangles of 60deg and 70deg both result in a double upheavalwhich has been explained by shear behavior in thesliding body based on the Mohr-Coulomb criterion

12 Advances in Civil Engineering

(3) e apparent friction coefficient of flowslide in-creases nonlinearly as the slope angle increases andshows spatial and lithological difference e greaterthe friction coefficient of the sliding material thehigher the initial benchmark of the curve the largerthe coefficient of velocity restitution the larger theabscissa corresponding to the position of inflectionpoint when the curve changes from flat to rapid

Data Availability

e data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

e authors declare that they have no conflicts of interest

Acknowledgments

is study would not have been possible without financialsupport from the Special Fund for the National NaturalScience Foundation of China under Grant nos 4170229841790442 and 41602359 as well as the Project Supported byNatural Science Basic Research Plan in Shaanxi Province ofChina under Grant no 2017JQ4020

Supplementary Materials

In this section the data on the slope angles and the apparentfriction coefficients are provided as a formation of tableese data are related to 44 rock flowslides in theWenchuanSichuan province China 28 soil flowslides in in the JingyangSouth platform Shaanxi province China and 27 chalkflowslides along the coast of northern Europe (Supple-mentary Materials)

References

[1] O Roche M Attali A Mangeney and A Lucas ldquoOn the run-out distance of geophysical gravitational flows insight fromfluidized granular collapse experimentsrdquo Earth PlanetaryScience Letters vol 311 no 3-4 pp 375ndash385 2011

[2] D Guo M Hamada C He Y Wang and Y Zou ldquoAnempirical model for landslide travel distance prediction inWenchuan earthquake areardquo Landslides vol 11 no 2pp 281ndash291 2014

[3] Y Okura H Kitahara A Kawanami and U KurokawaldquoTopography and volume effects on travel distance of surfacefailurerdquo Engineering Geology vol 67 no 3-4 pp 243ndash2542003

[4] Z Han Y Li J Huang et al ldquoNumerical simulation for run-out extent of debris flows using an improved cellular au-tomaton modelrdquo Bulletin of Engineering Geology and theEnvironment vol 76 no 3 pp 961ndash974 2017

[5] F Guzzeti F Ardizzone M Cardinali M Rossi andD Valigi ldquoLandslide volumes and landslide mobilizationrates in Umbria Central Italyrdquo Earth Planetary Science Let-ters vol 279 no 3-4 pp 222ndash229 2009

[6] J D Chaytor S Uri A R Solow and B Andrews ldquoSizedistribution of submarine landslides along the US Atlanticmarginrdquo Marine Geology vol 264 no 1-2 pp 16ndash27 2009

[7] H Qiu P Cui A D Regmi et al ldquoInfluence of topographyand volume on mobility of loess slides within different slipsurfacesrdquo Catena vol 157 pp 180ndash188 2017

[8] J Hu J Peng G Wang I Javed Y Wang and W LildquoDistribution and characteristics of landslide in LoessPlateau a case study in Shaanxi provincerdquo EngineeringGeology vol 236 pp 89ndash96 2018

[9] L Xu F C Dai Q M Gong L G am and H MinldquoIrrigation-induced loess flow failure in Heifangtai Platformnorth-west Chinardquo Environmental Earth Sciences vol 66no 6 pp 1707ndash1713 2012

[10] R Q Huang W H Liu J P Zhou and X J Pei ldquoRolling testson movement characteristics of rock blocksrdquo Journal ofGeotechnical Engineering vol 29 no 9 pp 1296ndash1302 2007in Chinese

[11] I Manzella and V Labiouse ldquoPhysical modelling to betterunderstand rock avalanchesrdquo in Physical Modelling in Geo-technics Two Volume Set Proceedings of the 7th InternationalConference on Physical Modelling in Geotechnics (ICPMG2010) vol 2 pp 1259ndash1265 Taylor amp Francis Group LondonUK 2010

[12] G B Crosta F V De Blasio M De Caro G VolpiS Imposimato and D Roddeman ldquoModes of propagationand deposition of granular flows onto an erodible substrateexperimental analytical and numerical studyrdquo Landslidesvol 14 no 1 pp 47ndash68 2017

[13] I Manzella and V Labiouse ldquoEmpirical and analytical ana-lyses of laboratory granular flows to investigate rock ava-lanche propagationrdquo Landslides vol 10 no 1 pp 23ndash362013

[14] Y-X Hu H-B Li S-C Qi G Fan and J-W ZhouldquoGranular effects on depositional processes of debris ava-lanchesrdquo KSCE Journal of Civil Engineering vol 24 no 4pp 1116ndash1127 2020

[15] I Goktepeli U Atmaca and A Cakan ldquoInvestigation of heattransfer augmentation between the ribbed plates via Taguchiapproach and computational fluid dynamicsrdquo Journal ofEermal Science vol 29 no 3 pp 647ndash666 2020

[16] M Farin A Mangeney and O Roche ldquoFundamental changesof granular flow dynamics deposition and erosion processesat high slope angles insights from laboratory experimentsrdquoJournal of Geophysical Research Earth Surface vol 119 no 3pp 504ndash532 2014

[17] M Kesseler V Heller and B Turnbull ldquoA laboratory-nu-merical approach for modelling scale effects in dry granularslidesrdquo Landslides vol 15 no 11 pp 2145ndash2159 2018

[18] A Dufresne and T R Davies ldquoLongitudinal ridges in massmovement depositsrdquo Geomorphology vol 105 no 3-4pp 171ndash181 2009

[19] T De Haas L Braat J R F W Leuven I R Lokhorst andM G Kleinhans ldquoEffects of debris flow composition onrunout depositional mechanisms and deposit morphology inlaboratory experimentsrdquo Journal of Geophysical ResearchEarth Surface vol 120 no 9 pp 1949ndash1972 2015

[20] C Goujon N omas and B Dalloz-Dubrujeaud ldquoMono-disperse dry granular flows on inclined planes role ofroughnessrdquo Ee European Physical Journal E vol 11 no 2pp 147ndash157 2003

[21] X-Y Fan S-J Tian and Y-Y Zhang ldquoMass-front velocity ofdry granular flows influenced by the angle of the slope to therunout plane and particle size gradationrdquo Journal of MountainScience vol 13 no 2 pp 234ndash245 2016

[22] G B Crosta F V D Blasio M Locatelli S Imposimato andD Roddeman ldquoLandslides falling onto a shallow erodible

Advances in Civil Engineering 13

tongue-shaped deposit form at low angles while alternatingupheaval and wash on the deposit surface and a generallycircle-shaped deposit form at large angles

34 Profile Morphology at Different AnglesMultidimensional analysis of the morphology of flowslidedeposit can help researchers gain a deeper understanding ofmorphology characteristics Figure 10 shows the profiles ofthe four flowslides depicted in Figure 6 Combining theprofile (Figure 10) with planar morphology (Figure 6) it canbe seen that at smaller slope angles of 40deg and 50deg a generallysingle upheaval with heights of 0027m and 0018m can berecognized with corresponding peak positions of 028m and050m respectively although the surface of each deposit isslightly undulated e position of the center of gravity isconsistent with the peak position of the upheaval At largerslope angles of 60deg and 70deg the surfaces of the deposits areundulated and there is double upheaval e maximumheight differences between adjacent upheaval and wash are0005m and 0023m respectively e heights of the frontupheaval are 0009m and 0026m with corresponding peakpositions of 067m and 040m respectively e heights ofthe back upheaval are 0012m and 0035m with corre-sponding peak positions of 010m and 006m respectivelye heights of the middle wash are 0006m and 0011m withcorresponding positions of 041m and 022m respectivelye positions of the center of gravity are relatively consistentwith the position of the middle wash

Comparing the four profiles shows a single upheavalshape at smaller angles of 40deg and 50deg and a double upheavalshape at larger angles of 60deg and 70deg In the flowslides withdouble upheaval the height of the front upheaval is smallerthan that of the back upheaval As the slope angle increasesthe average thickness of the deposit first decreases and thenincreases and the position of the center of gravity firstadvances and then retreats Based on the above comparisonit can be considered that there should be a key angle in thechange laws of flowslide deposit morphology If the slopeangle is less than or larger than the key angle under the sameconditions the length and width of the deposit will decreasee key angle is between 60deg and 70deg under the experimentalconditions in the paper However actually the key angle offlowslides should be different values depending on thematerial of the sliding body and the terrain conditions

e profile figures of chalk flows [36] on the steep cliffs ofthe European coast show similarities with the profilemorphology of the sandbox experiment for the larger anglesespecially the feature of double upheaval (Figures 11 and 12)ese two deposits shown in Figures 11 and 12 are 320mand 200m in length respectively It can be seen from thepicture that the leading edge of the flowslide deposit is flatand the surface shows double upheaval e front upheavalis lower than the back upheaval e height difference be-tween the middle wash and the back upheaval is greater thanthat of the front upheaval e center of gravity is locatedslightly in front of the middle wash It is noticed that there isa difference although the morphology of the double

Table 5 Variance analysis statistics

Morphological elements of deposit Factors Quadratic sum DOF Mean square error Partial η2

Ld (m)

A 68628 2 34314 055B 98732 2 49366 064C 597 2 2985 01

Error 56458 2 28229Total 229787 8

Wd (m)

A 108303 2 54151 087B 22448 2 11224 059C 4897 2 2449 024

Error 15835 2 7917Total 151483 8

Dd (m)

A 289 2 144 076B 342 2 171 079C 073 2 037 045

Error 091 2 045Total 795 8

A (m2)

A 011 2 006 052B 013 2 007 056C 001 2 0 006

Error 011 2 005Total 036 8

(LdWd) (1)

A 004 2 002 069B 005 2 003 073C 0 2 0 005

Error 002 2 001Total 011 8

Note Ld maximum sliding length of flowslide deposit on the surface Wd maximum width of flowslide deposit Dd maximum depth of flowslide deposit Aarea of flowslide deposit projected on horizontal plane θ slope angle (LdWd) length-width ratio

Advances in Civil Engineering 7

upheaval is the same e center of gravity of flowslidedeposit is located in the middle wash in experimentalconditions However the center of gravity is located in frontof the middle wash in actual flowslides e possible reasonmainly is the difference of properties of granular materialsbecause the movement and accumulation process of flow-slides are also affected by internal factors such as the particlesize and roundness of sliding body materials [14]

35 Mechanical Analysis of the Deposit MorphologyBefore the mechanical analysis the properties of themedium-fine sand and plexiglass should be noticed as abasis for this analysis e friction coefficient 0727 of themedium-fine sand is larger than 0538 of the interface of

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

aprime

0

6

0

6121812

918120 6

(a)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

bprime0 4

8

80

412

0

04812

4

(b)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

cprime

0 46

28

8

9

9

68

4

2

02

6

84

0

20 468

8

10

8

(c)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

dprime

04

1216

20

242016128

24

16 12

048

04816

12

08

4

(d)

Figure 6 Contour map of flowslides topography at different slope angles (a) (b) (c) and (d) corresponding to 40deg 50deg 60deg and 70deg respectively

N

Figure 7 Xi Hetan flowslide in China (34deg29prime32PrimeN 108deg50prime13Prime E)

N

Figure 8 Shenzhen flowslide Source Aerial image form Sou-hucom in China (22deg43prime04Prime N 113deg56prime31Prime E)

8 Advances in Civil Engineering

the plexiglass with the medium-fine sand e differenceof the friction coefficient between the interface and thesliding body itself also existed in actual flowslides becausethe interface will be lubricated by water or the liquefactioneffect [30 38] erefore the sliding body tends to flowalong the interface of the plexiglass and the sand forms athin sand layer like an erodible bed before the main slidingbody arrived on both inclined and horizontal plates[39 40] It provides an essential condition of the shearaction Figure 13 shows the cross-sectional state of thesliding body at three phases in which phase 3 indicatedby a solid line represents the final deposit state of the

N

(a)

N

(b)

N

(c)

N

(d)

Figure 9 Topography of typical flowslides (a) and (b) located in Folkestone Warren (51deg05prime03Prime N 1deg20prime36Prime E) (c) and (d) located DoverKent (51deg07prime53Prime N 1deg20prime36Prime E) Source from Hutchinson [36]

0 200 400 600 800 10000

10

20

30

40

Dep

th o

f dep

osit

(mm

)

Length of deposit (mm)

40degprofile50degprofile

60degprofile70degprofile

Figure 10 e deposit profiles of flowslides at different angles

N

Front upheavalFront upheavalBack upheavalWash

Figure 11 Flowslide occurred in May 1981 at Le Chien Neuf(49deg46prime59Prime N 0deg25prime19Prime E) Source from Hutchinson [36]

N

Front upheavalBack upheaval

Wash

Figure 12 Flowslide occurred in May 1958 at Wissower KlinkenRugen Germany (54deg32prime09Prime N 13deg40prime42Prime E) Source fromHutchinson [36]

Advances in Civil Engineering 9

sliding body In this paper the Mohr-Coulomb criterion isused in the analysis of the formation of single and doubleupheaval as follows

τ c + σ tanφ (2)Taking the numbers① and② as two microelements in

the sliding body under a stress condition whereby themaximum principal stress σ1 of the microelement at theslope break is parallel to the direction of the inclined plateand the confining pressure σ2 σ3 is perpendicular to thedirection of the inclined plate (the microelements are smallenough so it is considered that σ2 σ3) shear behavioroccurs [22 30 31]

e angle between the shear plane and the maximumprincipal stress is defined as 45degminus (φ2) and the friction angleφ in the sand material is 36deg For a microelement ① atsmaller slope angles of 40deg and 50deg the counterclockwiseangles between the shear action direction and the overallmovement direction of the flowslide are 167deg and 157degrespectively tending to horizontal which is approximatelythe same as the direction of the maximum principal stress inthe state of the microelement ② which will be a source ofpower for the morphology of single upheaval Afterwardsthe microelements located on the horizontal plate movealong the shear plane at an angle of 27deg counterclockwisewith the movement direction of the sliding body on thehorizontal plate under the obstruction of the other particlesin the front and the push of rear particles similar to a thrustnappe in macroscopic performance resulting in depositmorphology as a single upheaval

e formation of double upheaval is more complicatedHere it will be described according to the order of the frontupheaval the wash and the back upheaval First based on thesame calculation method as for the smaller angles thecounterclockwise angles between the shear-action direction ofthe microelement ① and the movement direction of thesliding body on the horizontal plate are 93deg and 83deg at thelarger slope angles of 60deg and 70deg respectively At this pointthe sliding body deposits part of its load under the actions ofshear force and thrust force to form a rudiment of the frontupheaval while the part of the sliding body on the horizontalplate keeps moving under the inertial force and the thrust

force of the back part sliding body (the force state is similar tothe microelement②) and continues development of the frontupheaval gradually Next themicroelements at the back of thefront upheaval stop their shearing motion when the equi-librium condition of equation (3) is satisfied climbing isstopped and the wash is formed However mechanicalconditions still allow the microelements to move horizontallyWhen the equilibrium condition of equation (4) is satisfiedthe horizontal movement stops Assuming that the micro-elements are small enough the equilibrium condition can belisted as follows

m middot g τ middot sin 45deg minusφ2

1113874 1113875 (3)

m middot a τ middot cos 45deg minusφ2

1113874 1113875 (4)

where m is the mass of the microelement g is the accel-eration of gravity τ is shear stress a is the horizontal ac-celeration of the microelement and φ is the friction angle ofsand material

Eventually under the balance of equations (3) and (4)the sliding body that is still moving at the trailing edge is verysmall as the speed so the shear force provided can be re-duced accordingly At this time if the internal shearing forceof the sliding body at the trailing edge is greater than theshearing strength of the sliding body the subsequent slidingbody is deposited at the foot of the slope to form the backupheaval and conducts a localized shear motion based on theshear force similar to the microelement ① however if theshear force is less than the shear strength of the sliding bodythen the sliding body deposits at the slope break to form theback upheaval In general the order of the formation processof the double upheaval is front upheaval wash and backupheaval

e formation of double upheaval morphology is alsoowing to properties of the granular material In this studythere is a double upheaval morphology nevertheless this isnot in the study by Hu [14] e range of the granularparticle from 5times10minus 6m to 4times10minus 4m in this article is ob-viously less than the range of the granular particle from10minus 3m to 16times10minus 2m in the research by Hu [14] ereforethe internal friction angle and cohesion of the experimentalmaterial in this study are larger than those in the research byHu [14]emore the internal friction angle and cohesion ofgranular materials are the more the difficult to disturb thestate of the deposit morphology [19] is When the frontupheaval is formed the granular material with a particle sizeof 5times10minus 6m to 4times10minus 4m is more likely to keep its mor-phology of the front upheaval in the disturbance of the backsliding body compared with a particle size of 10minus 3m to16times10minus 2m If the backsliding body can not disturb theformed front upheaval it will accumulate in the slope breakforming the back upheaval meanwhile whereupon thedouble upheaval is formed

For these change laws as the slope angle increases themaximum width maximum length and area of the depositfirst increase and then decrease similar to the results by Hu[14] at the condition of the mass of fine particles changes

Phase 1Phase 2Phase 3

m1 m2

σ1

σ1

σ2

σ2

α

ατ1 τ1

12

Figure 13 Schematic illustration of shear force of flowslide

10 Advances in Civil Engineering

Based on those results a conclusion was proposed that there isan optimum proportion of fine and coarse particles in thegranular materials because the shear strength and frictioncoefficient are increasing with the mass of fine particles in-creases [41 42] In the condition that the internal frictionangle and cohesion are the smallest the maximum widthmaximum length and area are maximum Hence the de-duction mentioned before is reasonable that there should be acritical slope angle in the changes of the deposit morphology

36 Apparent Friction Coefficient To further understand therules that govern changes in the apparent friction coefficientwith slope angle and the distribution characteristics of theapparent friction coefficient in different lithologies and dif-ferent regions the apparent friction coefficients and corre-sponding slope angle data were collected for research fromthree places ere are 28 soil flowslides in the Jingyang Southplatform (34deg29prime32PrimeN 108deg50prime13Prime E) Shaanxi province China[30] consisting of loesse accumulation area is the even riverterrace comprised of the loess deposit underlain by the siltyclay deposit with interbedded sand layers occasionally Alongthe coast of northern Europe(51deg05prime03Prime N 1deg20prime36Prime E) thereare 27 chalk flows [22] with a porosity ranging from 9 to 52and dry density ranging from 2413 to 1265 kNm3 InWenchuan (31deg00prime00Prime N 103deg40prime00Prime E) Sichuan provinceChina [2] there are 44 rock flowslides that occurred in themountainous area and consisted of the sand-mudstonemainlye accumulation area is relatively flat though it is hard tocompare with flowslides that occurred in Jingyang and coast ofnorthern Europe An important reason why the above threeflowslide groups are taken as research objects is that they havesimilar flow characteristic and long runout distance In ad-dition the topographical condition of the accumulation area issimilar to the model test conditions It can be seen fromFigure 14 that the overall relationship between the apparentfriction coefficient of flowslide and the slope angles shows anonlinear growth relationship is conclusion is consistentwith Crosta [22] e growth relationship is as follows

H

L

tan θ1 + (cos θ + ε sin θ)2 middot ((tan θμ) minus 1)1113960 1113961

(5)

where θ represents the slope angle H represents the slopeheight L represents the moving distance of the flowslide ε isthe coefficient of velocity restitution when the sliding bodyreaches the slope break to change its moving direction tohorizontal μ is the friction coefficient

e difference in regional lithology determines thedifference between the friction coefficient μ and the co-efficient of velocity restitution ε so two different μ valuesand three different ε values are set in this paper When μ is02 and ε is 0 02 or 04 the curve better reflects theoverall trend of the three natural flowslide groups Whenthe friction coefficient μ between the inclined plate andthe experimental material is 055 and ε is 04 the curveshows good correlation with the experimental points Itcan be said that the friction coefficient determines theheight of the initial benchmark of the curve e larger thevalue of the friction coefficient the higher the initial

benchmark of the curve e coefficient of velocity res-titution determines the change of the inflection pointwhen the curve changes from flat to rapid e larger thevalue of the coefficient of velocity restitution the largerthe horizontal coordinates of the inflection point

From the point of energy conversion energy loss at theslope break is closely related to terrain and lithology ekinetic energy loss is small when the junction of slope and theaccumulation area is a curved slope break when the junctionhas sharp edges the kinetic energy loss is larger [13 22]Under the experimental conditions in which slope height Hand volume are constant as the slope angle increases energyloss also increases due to the collision of the flowslide and thehorizontal plate at the slope break [43 44] which results inshorter distance of movement of the sliding body e slopeangle in the chalk flow area is too large and the energy loss isexcessive though the sliding path is lubricated by water be-neath the accumulation area 1-2m e flowslides in Wen-chuan affected by lubrication due to water are few Besidesthere are many obstacles in the movement path because theseflowslides are located in a mountainous area though theaccumulation area is relatively flat so energy loss is largeHowever for the flowslides in Jingyang the path is flat andstraight as well as lubricated by the liquefied effect [31] andthe energy loss is small e flowslides in the model exper-iment fall between the three natural flowslide groups to theleft of chalk flows above loess flowslides and intersectingwith rock flowslides e apparent friction coefficient in thecondition of this model experiment is not the smallest al-though the condition is ideal flat for the reason that there areno lubrications erefore the lubrication effect plays a keyrole in affecting the apparent friction coefficient

4 Discussion

In this paper the Taguchi method was employed inno-vatively to study quantitatively the order of significance

0 20 40 60 80 10000

05

10

15

20

Wenchuan rock flowslidesJingyang loess flowslides

Chalk flowsExperimental flowslides

HL

Slope angle (deg)

ε = 0

ε = 02

ε = 0

2

ε = 04

ε = 0

4

ε = 0μ = 055

μ = 02

tan θ1 + [(cos θ + εsin θ)2 middot (tan θμ ndash1)]

ndashHL =

Figure 14 Comprehensive relationship between slope angle andapparent friction coefficient Formula comes from Crosta [22]

Advances in Civil Engineering 11

and the maximal responses of combinations of factorsaffecting flowslide deposit morphology e calculationresults in Figure 5 confirm that slope angle is the mostinfluential factor e maximum influencing combinationis a volume of 54 times10minus 3 m3 at a slope angle of 70deg and aslope height of 090m which will be useful to evaluate thehazards of potential flowslides Based on the acceptedknowledge that slope angle is the largest influencingfactor further experimental research was performed econditions of small slope angles and large slope anglesshow interesting phenomena of single upheaval anddouble upheaval respectively the Mohr-Coulomb crite-rion is used to explain the action of the shear motionduring the process e comprehensive study of apparentfriction coefficient shows a nonlinear increase with slopeangle as well as spatial and lithological differences indistribution which are affected by the friction coefficientand the coefficient of velocity restitution

Similar model experiments have been carried out toanalyze the movement rule of flowslide [4 12 21] anddeposit morphology [3 16 18 22 25] Previous studiesfocused on the influence of a single factor and the researchcontent is more in-depth [8 11 45 46] but the researchprocess is not systematic enough as flowslide movement anddeposit morphology are determined by multiple factorsincluding volume slope angle slope height sliding bodymaterial and slip bed material Of course some scholarshave conducted multifactor experiments for exampleManzella [13] studied the influence of flowslide volumeinitial height slope of slip bed and the slope break onmovement characteristics but the relative significance ofvarious factors governing the movement of flowslide has notbeen revealed although the factors considered are com-prehensive An orthogonal experiment was used to study themotion characteristics (velocity acceleration) of flowslideunder the influence of various factors and found that slopeangle was the largest influencing factor [33] which isconsistent with the findings of this article though the re-search content was different

e morphology of the double upheaval of flowslide wasalso observed in an experiment similar to the current study[1] which examined wave propagation in the macroscopicview at experiment was also carried out with bulk ma-terials A wave would be formed when the experimentalmaterials slid or fell into the accumulation area the initialwave would accelerate suddenly as it was pushed by asubsequent wave propagating forward e wave crestformed when motion stopped due to resistance may be oneof the reasons for the formation of upheaval and the wavetrough can be regarded as wash [1] in this paper Accordingto the motion rule of the object the flowslide is partiallyreflected at the slope break [22] and then propagates in theform of waves which were observed somewhat during theexperiment of larger slope angles Macroscopic wavepropagation may be the cause of the formation of upheavaland wash on the surface of flowslide but if the phenomenonis deeply understood from the interior of sliding body theinternal shear behavior in the movement of flowslide may bereasonable In addition the morphology of actual flowslides

is also affected by properties of sliding materials [14 19 20]With an increase of the mass of fine particles the length andwidth of the flowslide deposit increase first and then de-crease [14] erefore the difference of properties ofgranular materials is contributing to the formation of thedouble upheaval

e apparent friction coefficient is often involved in thestudy of flowslide but the focus here is on its relationshipwith volume [5 7 9 21 38] ere are few studies that useslope angle as a variable including Lied and Bakkehoi [47]and Okura [3] both of which proposed that the apparentfriction coefficient of flowslide was directly proportional tothe slope angle but the results were based on the analysis ofonly about 20 flowslides so the conclusion was limited eanalysis in this paper involved 99 landslides in three areasand the conclusion is consistent with the viewpoint of Crosta[22] in finding that the apparent friction coefficient increasesnonlinearly as the slope angle increases In the comparisonmade in this study spatial and lithological differences in theapparent friction coefficient are also found

In this paper an experimental idea of constant height ofthe center of gravity is provided but when the angle changesthe sliding distance of the flowslide on the inclined plate(Figure 4(b)) will inevitably increase and the loss of kineticenergy due to friction will increase accordingly ereforefurther improvement of the experimental device is needed toreduce the kinetic energy loss due to the increase in thesliding distance on the inclined plate Modeling should alsoinclude more influencing factors

5 Conclusions

(1) e combined influence factor obtained by theTaguchi method demonstrates that when flowslidevolume is 54times10minus 3m3 the slope angle is 70deg theslope height is 090m and the changes of depositmorphology of the flowslide is the largest e ex-treme differences of maximum length (Ld) maxi-mum thickness (Dd) area (A) and the length-widthratio (LdWd) at three different levels under variousfactors are 277 551 454 and 17 respectively ecalculation results indicate that the factor that has thegreatest impact on the deposit morphology offlowslide is slope angle followed by volume andslope height

(2) With an increase in slope angle the width lengtharea and length-width ratio of flowslide deposits firstincrease and then decrease and the maximum ac-cumulation thickness first decreases and then in-creases And there should be a critical angle in thechange laws of the flowslide deposit morphologythat is between 60deg and 70deg under the experimentalconditions When the volume is 36times10minus 3m3 andthe slope height is 090m the slope angles at 40deg and50deg both result in a single upheaval while slopeangles of 60deg and 70deg both result in a double upheavalwhich has been explained by shear behavior in thesliding body based on the Mohr-Coulomb criterion

12 Advances in Civil Engineering

(3) e apparent friction coefficient of flowslide in-creases nonlinearly as the slope angle increases andshows spatial and lithological difference e greaterthe friction coefficient of the sliding material thehigher the initial benchmark of the curve the largerthe coefficient of velocity restitution the larger theabscissa corresponding to the position of inflectionpoint when the curve changes from flat to rapid

Data Availability

e data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

e authors declare that they have no conflicts of interest

Acknowledgments

is study would not have been possible without financialsupport from the Special Fund for the National NaturalScience Foundation of China under Grant nos 4170229841790442 and 41602359 as well as the Project Supported byNatural Science Basic Research Plan in Shaanxi Province ofChina under Grant no 2017JQ4020

Supplementary Materials

In this section the data on the slope angles and the apparentfriction coefficients are provided as a formation of tableese data are related to 44 rock flowslides in theWenchuanSichuan province China 28 soil flowslides in in the JingyangSouth platform Shaanxi province China and 27 chalkflowslides along the coast of northern Europe (Supple-mentary Materials)

References

[1] O Roche M Attali A Mangeney and A Lucas ldquoOn the run-out distance of geophysical gravitational flows insight fromfluidized granular collapse experimentsrdquo Earth PlanetaryScience Letters vol 311 no 3-4 pp 375ndash385 2011

[2] D Guo M Hamada C He Y Wang and Y Zou ldquoAnempirical model for landslide travel distance prediction inWenchuan earthquake areardquo Landslides vol 11 no 2pp 281ndash291 2014

[3] Y Okura H Kitahara A Kawanami and U KurokawaldquoTopography and volume effects on travel distance of surfacefailurerdquo Engineering Geology vol 67 no 3-4 pp 243ndash2542003

[4] Z Han Y Li J Huang et al ldquoNumerical simulation for run-out extent of debris flows using an improved cellular au-tomaton modelrdquo Bulletin of Engineering Geology and theEnvironment vol 76 no 3 pp 961ndash974 2017

[5] F Guzzeti F Ardizzone M Cardinali M Rossi andD Valigi ldquoLandslide volumes and landslide mobilizationrates in Umbria Central Italyrdquo Earth Planetary Science Let-ters vol 279 no 3-4 pp 222ndash229 2009

[6] J D Chaytor S Uri A R Solow and B Andrews ldquoSizedistribution of submarine landslides along the US Atlanticmarginrdquo Marine Geology vol 264 no 1-2 pp 16ndash27 2009

[7] H Qiu P Cui A D Regmi et al ldquoInfluence of topographyand volume on mobility of loess slides within different slipsurfacesrdquo Catena vol 157 pp 180ndash188 2017

[8] J Hu J Peng G Wang I Javed Y Wang and W LildquoDistribution and characteristics of landslide in LoessPlateau a case study in Shaanxi provincerdquo EngineeringGeology vol 236 pp 89ndash96 2018

[9] L Xu F C Dai Q M Gong L G am and H MinldquoIrrigation-induced loess flow failure in Heifangtai Platformnorth-west Chinardquo Environmental Earth Sciences vol 66no 6 pp 1707ndash1713 2012

[10] R Q Huang W H Liu J P Zhou and X J Pei ldquoRolling testson movement characteristics of rock blocksrdquo Journal ofGeotechnical Engineering vol 29 no 9 pp 1296ndash1302 2007in Chinese

[11] I Manzella and V Labiouse ldquoPhysical modelling to betterunderstand rock avalanchesrdquo in Physical Modelling in Geo-technics Two Volume Set Proceedings of the 7th InternationalConference on Physical Modelling in Geotechnics (ICPMG2010) vol 2 pp 1259ndash1265 Taylor amp Francis Group LondonUK 2010

[12] G B Crosta F V De Blasio M De Caro G VolpiS Imposimato and D Roddeman ldquoModes of propagationand deposition of granular flows onto an erodible substrateexperimental analytical and numerical studyrdquo Landslidesvol 14 no 1 pp 47ndash68 2017

[13] I Manzella and V Labiouse ldquoEmpirical and analytical ana-lyses of laboratory granular flows to investigate rock ava-lanche propagationrdquo Landslides vol 10 no 1 pp 23ndash362013

[14] Y-X Hu H-B Li S-C Qi G Fan and J-W ZhouldquoGranular effects on depositional processes of debris ava-lanchesrdquo KSCE Journal of Civil Engineering vol 24 no 4pp 1116ndash1127 2020

[15] I Goktepeli U Atmaca and A Cakan ldquoInvestigation of heattransfer augmentation between the ribbed plates via Taguchiapproach and computational fluid dynamicsrdquo Journal ofEermal Science vol 29 no 3 pp 647ndash666 2020

[16] M Farin A Mangeney and O Roche ldquoFundamental changesof granular flow dynamics deposition and erosion processesat high slope angles insights from laboratory experimentsrdquoJournal of Geophysical Research Earth Surface vol 119 no 3pp 504ndash532 2014

[17] M Kesseler V Heller and B Turnbull ldquoA laboratory-nu-merical approach for modelling scale effects in dry granularslidesrdquo Landslides vol 15 no 11 pp 2145ndash2159 2018

[18] A Dufresne and T R Davies ldquoLongitudinal ridges in massmovement depositsrdquo Geomorphology vol 105 no 3-4pp 171ndash181 2009

[19] T De Haas L Braat J R F W Leuven I R Lokhorst andM G Kleinhans ldquoEffects of debris flow composition onrunout depositional mechanisms and deposit morphology inlaboratory experimentsrdquo Journal of Geophysical ResearchEarth Surface vol 120 no 9 pp 1949ndash1972 2015

[20] C Goujon N omas and B Dalloz-Dubrujeaud ldquoMono-disperse dry granular flows on inclined planes role ofroughnessrdquo Ee European Physical Journal E vol 11 no 2pp 147ndash157 2003

[21] X-Y Fan S-J Tian and Y-Y Zhang ldquoMass-front velocity ofdry granular flows influenced by the angle of the slope to therunout plane and particle size gradationrdquo Journal of MountainScience vol 13 no 2 pp 234ndash245 2016

[22] G B Crosta F V D Blasio M Locatelli S Imposimato andD Roddeman ldquoLandslides falling onto a shallow erodible

Advances in Civil Engineering 13

upheaval is the same e center of gravity of flowslidedeposit is located in the middle wash in experimentalconditions However the center of gravity is located in frontof the middle wash in actual flowslides e possible reasonmainly is the difference of properties of granular materialsbecause the movement and accumulation process of flow-slides are also affected by internal factors such as the particlesize and roundness of sliding body materials [14]

35 Mechanical Analysis of the Deposit MorphologyBefore the mechanical analysis the properties of themedium-fine sand and plexiglass should be noticed as abasis for this analysis e friction coefficient 0727 of themedium-fine sand is larger than 0538 of the interface of

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

aprime

0

6

0

6121812

918120 6

(a)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

bprime0 4

8

80

412

0

04812

4

(b)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

cprime

0 46

28

8

9

9

68

4

2

02

6

84

0

20 468

8

10

8

(c)

0

ndash200

ndash400

ndash600

ndash800

ndash1000ndash600 ndash400 ndash200 0 200 400 600

dprime

04

1216

20

242016128

24

16 12

048

04816

12

08

4

(d)

Figure 6 Contour map of flowslides topography at different slope angles (a) (b) (c) and (d) corresponding to 40deg 50deg 60deg and 70deg respectively

N

Figure 7 Xi Hetan flowslide in China (34deg29prime32PrimeN 108deg50prime13Prime E)

N

Figure 8 Shenzhen flowslide Source Aerial image form Sou-hucom in China (22deg43prime04Prime N 113deg56prime31Prime E)

8 Advances in Civil Engineering

the plexiglass with the medium-fine sand e differenceof the friction coefficient between the interface and thesliding body itself also existed in actual flowslides becausethe interface will be lubricated by water or the liquefactioneffect [30 38] erefore the sliding body tends to flowalong the interface of the plexiglass and the sand forms athin sand layer like an erodible bed before the main slidingbody arrived on both inclined and horizontal plates[39 40] It provides an essential condition of the shearaction Figure 13 shows the cross-sectional state of thesliding body at three phases in which phase 3 indicatedby a solid line represents the final deposit state of the

N

(a)

N

(b)

N

(c)

N

(d)

Figure 9 Topography of typical flowslides (a) and (b) located in Folkestone Warren (51deg05prime03Prime N 1deg20prime36Prime E) (c) and (d) located DoverKent (51deg07prime53Prime N 1deg20prime36Prime E) Source from Hutchinson [36]

0 200 400 600 800 10000

10

20

30

40

Dep

th o

f dep

osit

(mm

)

Length of deposit (mm)

40degprofile50degprofile

60degprofile70degprofile

Figure 10 e deposit profiles of flowslides at different angles

N

Front upheavalFront upheavalBack upheavalWash

Figure 11 Flowslide occurred in May 1981 at Le Chien Neuf(49deg46prime59Prime N 0deg25prime19Prime E) Source from Hutchinson [36]

N

Front upheavalBack upheaval

Wash

Figure 12 Flowslide occurred in May 1958 at Wissower KlinkenRugen Germany (54deg32prime09Prime N 13deg40prime42Prime E) Source fromHutchinson [36]

Advances in Civil Engineering 9

sliding body In this paper the Mohr-Coulomb criterion isused in the analysis of the formation of single and doubleupheaval as follows

τ c + σ tanφ (2)Taking the numbers① and② as two microelements in

the sliding body under a stress condition whereby themaximum principal stress σ1 of the microelement at theslope break is parallel to the direction of the inclined plateand the confining pressure σ2 σ3 is perpendicular to thedirection of the inclined plate (the microelements are smallenough so it is considered that σ2 σ3) shear behavioroccurs [22 30 31]

e angle between the shear plane and the maximumprincipal stress is defined as 45degminus (φ2) and the friction angleφ in the sand material is 36deg For a microelement ① atsmaller slope angles of 40deg and 50deg the counterclockwiseangles between the shear action direction and the overallmovement direction of the flowslide are 167deg and 157degrespectively tending to horizontal which is approximatelythe same as the direction of the maximum principal stress inthe state of the microelement ② which will be a source ofpower for the morphology of single upheaval Afterwardsthe microelements located on the horizontal plate movealong the shear plane at an angle of 27deg counterclockwisewith the movement direction of the sliding body on thehorizontal plate under the obstruction of the other particlesin the front and the push of rear particles similar to a thrustnappe in macroscopic performance resulting in depositmorphology as a single upheaval

e formation of double upheaval is more complicatedHere it will be described according to the order of the frontupheaval the wash and the back upheaval First based on thesame calculation method as for the smaller angles thecounterclockwise angles between the shear-action direction ofthe microelement ① and the movement direction of thesliding body on the horizontal plate are 93deg and 83deg at thelarger slope angles of 60deg and 70deg respectively At this pointthe sliding body deposits part of its load under the actions ofshear force and thrust force to form a rudiment of the frontupheaval while the part of the sliding body on the horizontalplate keeps moving under the inertial force and the thrust

force of the back part sliding body (the force state is similar tothe microelement②) and continues development of the frontupheaval gradually Next themicroelements at the back of thefront upheaval stop their shearing motion when the equi-librium condition of equation (3) is satisfied climbing isstopped and the wash is formed However mechanicalconditions still allow the microelements to move horizontallyWhen the equilibrium condition of equation (4) is satisfiedthe horizontal movement stops Assuming that the micro-elements are small enough the equilibrium condition can belisted as follows

m middot g τ middot sin 45deg minusφ2

1113874 1113875 (3)

m middot a τ middot cos 45deg minusφ2

1113874 1113875 (4)

where m is the mass of the microelement g is the accel-eration of gravity τ is shear stress a is the horizontal ac-celeration of the microelement and φ is the friction angle ofsand material

Eventually under the balance of equations (3) and (4)the sliding body that is still moving at the trailing edge is verysmall as the speed so the shear force provided can be re-duced accordingly At this time if the internal shearing forceof the sliding body at the trailing edge is greater than theshearing strength of the sliding body the subsequent slidingbody is deposited at the foot of the slope to form the backupheaval and conducts a localized shear motion based on theshear force similar to the microelement ① however if theshear force is less than the shear strength of the sliding bodythen the sliding body deposits at the slope break to form theback upheaval In general the order of the formation processof the double upheaval is front upheaval wash and backupheaval

e formation of double upheaval morphology is alsoowing to properties of the granular material In this studythere is a double upheaval morphology nevertheless this isnot in the study by Hu [14] e range of the granularparticle from 5times10minus 6m to 4times10minus 4m in this article is ob-viously less than the range of the granular particle from10minus 3m to 16times10minus 2m in the research by Hu [14] ereforethe internal friction angle and cohesion of the experimentalmaterial in this study are larger than those in the research byHu [14]emore the internal friction angle and cohesion ofgranular materials are the more the difficult to disturb thestate of the deposit morphology [19] is When the frontupheaval is formed the granular material with a particle sizeof 5times10minus 6m to 4times10minus 4m is more likely to keep its mor-phology of the front upheaval in the disturbance of the backsliding body compared with a particle size of 10minus 3m to16times10minus 2m If the backsliding body can not disturb theformed front upheaval it will accumulate in the slope breakforming the back upheaval meanwhile whereupon thedouble upheaval is formed

For these change laws as the slope angle increases themaximum width maximum length and area of the depositfirst increase and then decrease similar to the results by Hu[14] at the condition of the mass of fine particles changes

Phase 1Phase 2Phase 3

m1 m2

σ1

σ1

σ2

σ2

α

ατ1 τ1

12

Figure 13 Schematic illustration of shear force of flowslide

10 Advances in Civil Engineering

Based on those results a conclusion was proposed that there isan optimum proportion of fine and coarse particles in thegranular materials because the shear strength and frictioncoefficient are increasing with the mass of fine particles in-creases [41 42] In the condition that the internal frictionangle and cohesion are the smallest the maximum widthmaximum length and area are maximum Hence the de-duction mentioned before is reasonable that there should be acritical slope angle in the changes of the deposit morphology

36 Apparent Friction Coefficient To further understand therules that govern changes in the apparent friction coefficientwith slope angle and the distribution characteristics of theapparent friction coefficient in different lithologies and dif-ferent regions the apparent friction coefficients and corre-sponding slope angle data were collected for research fromthree places ere are 28 soil flowslides in the Jingyang Southplatform (34deg29prime32PrimeN 108deg50prime13Prime E) Shaanxi province China[30] consisting of loesse accumulation area is the even riverterrace comprised of the loess deposit underlain by the siltyclay deposit with interbedded sand layers occasionally Alongthe coast of northern Europe(51deg05prime03Prime N 1deg20prime36Prime E) thereare 27 chalk flows [22] with a porosity ranging from 9 to 52and dry density ranging from 2413 to 1265 kNm3 InWenchuan (31deg00prime00Prime N 103deg40prime00Prime E) Sichuan provinceChina [2] there are 44 rock flowslides that occurred in themountainous area and consisted of the sand-mudstonemainlye accumulation area is relatively flat though it is hard tocompare with flowslides that occurred in Jingyang and coast ofnorthern Europe An important reason why the above threeflowslide groups are taken as research objects is that they havesimilar flow characteristic and long runout distance In ad-dition the topographical condition of the accumulation area issimilar to the model test conditions It can be seen fromFigure 14 that the overall relationship between the apparentfriction coefficient of flowslide and the slope angles shows anonlinear growth relationship is conclusion is consistentwith Crosta [22] e growth relationship is as follows

H

L

tan θ1 + (cos θ + ε sin θ)2 middot ((tan θμ) minus 1)1113960 1113961

(5)

where θ represents the slope angle H represents the slopeheight L represents the moving distance of the flowslide ε isthe coefficient of velocity restitution when the sliding bodyreaches the slope break to change its moving direction tohorizontal μ is the friction coefficient

e difference in regional lithology determines thedifference between the friction coefficient μ and the co-efficient of velocity restitution ε so two different μ valuesand three different ε values are set in this paper When μ is02 and ε is 0 02 or 04 the curve better reflects theoverall trend of the three natural flowslide groups Whenthe friction coefficient μ between the inclined plate andthe experimental material is 055 and ε is 04 the curveshows good correlation with the experimental points Itcan be said that the friction coefficient determines theheight of the initial benchmark of the curve e larger thevalue of the friction coefficient the higher the initial

benchmark of the curve e coefficient of velocity res-titution determines the change of the inflection pointwhen the curve changes from flat to rapid e larger thevalue of the coefficient of velocity restitution the largerthe horizontal coordinates of the inflection point

From the point of energy conversion energy loss at theslope break is closely related to terrain and lithology ekinetic energy loss is small when the junction of slope and theaccumulation area is a curved slope break when the junctionhas sharp edges the kinetic energy loss is larger [13 22]Under the experimental conditions in which slope height Hand volume are constant as the slope angle increases energyloss also increases due to the collision of the flowslide and thehorizontal plate at the slope break [43 44] which results inshorter distance of movement of the sliding body e slopeangle in the chalk flow area is too large and the energy loss isexcessive though the sliding path is lubricated by water be-neath the accumulation area 1-2m e flowslides in Wen-chuan affected by lubrication due to water are few Besidesthere are many obstacles in the movement path because theseflowslides are located in a mountainous area though theaccumulation area is relatively flat so energy loss is largeHowever for the flowslides in Jingyang the path is flat andstraight as well as lubricated by the liquefied effect [31] andthe energy loss is small e flowslides in the model exper-iment fall between the three natural flowslide groups to theleft of chalk flows above loess flowslides and intersectingwith rock flowslides e apparent friction coefficient in thecondition of this model experiment is not the smallest al-though the condition is ideal flat for the reason that there areno lubrications erefore the lubrication effect plays a keyrole in affecting the apparent friction coefficient

4 Discussion

In this paper the Taguchi method was employed inno-vatively to study quantitatively the order of significance

0 20 40 60 80 10000

05

10

15

20

Wenchuan rock flowslidesJingyang loess flowslides

Chalk flowsExperimental flowslides

HL

Slope angle (deg)

ε = 0

ε = 02

ε = 0

2

ε = 04

ε = 0

4

ε = 0μ = 055

μ = 02

tan θ1 + [(cos θ + εsin θ)2 middot (tan θμ ndash1)]

ndashHL =

Figure 14 Comprehensive relationship between slope angle andapparent friction coefficient Formula comes from Crosta [22]

Advances in Civil Engineering 11

and the maximal responses of combinations of factorsaffecting flowslide deposit morphology e calculationresults in Figure 5 confirm that slope angle is the mostinfluential factor e maximum influencing combinationis a volume of 54 times10minus 3 m3 at a slope angle of 70deg and aslope height of 090m which will be useful to evaluate thehazards of potential flowslides Based on the acceptedknowledge that slope angle is the largest influencingfactor further experimental research was performed econditions of small slope angles and large slope anglesshow interesting phenomena of single upheaval anddouble upheaval respectively the Mohr-Coulomb crite-rion is used to explain the action of the shear motionduring the process e comprehensive study of apparentfriction coefficient shows a nonlinear increase with slopeangle as well as spatial and lithological differences indistribution which are affected by the friction coefficientand the coefficient of velocity restitution

Similar model experiments have been carried out toanalyze the movement rule of flowslide [4 12 21] anddeposit morphology [3 16 18 22 25] Previous studiesfocused on the influence of a single factor and the researchcontent is more in-depth [8 11 45 46] but the researchprocess is not systematic enough as flowslide movement anddeposit morphology are determined by multiple factorsincluding volume slope angle slope height sliding bodymaterial and slip bed material Of course some scholarshave conducted multifactor experiments for exampleManzella [13] studied the influence of flowslide volumeinitial height slope of slip bed and the slope break onmovement characteristics but the relative significance ofvarious factors governing the movement of flowslide has notbeen revealed although the factors considered are com-prehensive An orthogonal experiment was used to study themotion characteristics (velocity acceleration) of flowslideunder the influence of various factors and found that slopeangle was the largest influencing factor [33] which isconsistent with the findings of this article though the re-search content was different

e morphology of the double upheaval of flowslide wasalso observed in an experiment similar to the current study[1] which examined wave propagation in the macroscopicview at experiment was also carried out with bulk ma-terials A wave would be formed when the experimentalmaterials slid or fell into the accumulation area the initialwave would accelerate suddenly as it was pushed by asubsequent wave propagating forward e wave crestformed when motion stopped due to resistance may be oneof the reasons for the formation of upheaval and the wavetrough can be regarded as wash [1] in this paper Accordingto the motion rule of the object the flowslide is partiallyreflected at the slope break [22] and then propagates in theform of waves which were observed somewhat during theexperiment of larger slope angles Macroscopic wavepropagation may be the cause of the formation of upheavaland wash on the surface of flowslide but if the phenomenonis deeply understood from the interior of sliding body theinternal shear behavior in the movement of flowslide may bereasonable In addition the morphology of actual flowslides

is also affected by properties of sliding materials [14 19 20]With an increase of the mass of fine particles the length andwidth of the flowslide deposit increase first and then de-crease [14] erefore the difference of properties ofgranular materials is contributing to the formation of thedouble upheaval

e apparent friction coefficient is often involved in thestudy of flowslide but the focus here is on its relationshipwith volume [5 7 9 21 38] ere are few studies that useslope angle as a variable including Lied and Bakkehoi [47]and Okura [3] both of which proposed that the apparentfriction coefficient of flowslide was directly proportional tothe slope angle but the results were based on the analysis ofonly about 20 flowslides so the conclusion was limited eanalysis in this paper involved 99 landslides in three areasand the conclusion is consistent with the viewpoint of Crosta[22] in finding that the apparent friction coefficient increasesnonlinearly as the slope angle increases In the comparisonmade in this study spatial and lithological differences in theapparent friction coefficient are also found

In this paper an experimental idea of constant height ofthe center of gravity is provided but when the angle changesthe sliding distance of the flowslide on the inclined plate(Figure 4(b)) will inevitably increase and the loss of kineticenergy due to friction will increase accordingly ereforefurther improvement of the experimental device is needed toreduce the kinetic energy loss due to the increase in thesliding distance on the inclined plate Modeling should alsoinclude more influencing factors

5 Conclusions

(1) e combined influence factor obtained by theTaguchi method demonstrates that when flowslidevolume is 54times10minus 3m3 the slope angle is 70deg theslope height is 090m and the changes of depositmorphology of the flowslide is the largest e ex-treme differences of maximum length (Ld) maxi-mum thickness (Dd) area (A) and the length-widthratio (LdWd) at three different levels under variousfactors are 277 551 454 and 17 respectively ecalculation results indicate that the factor that has thegreatest impact on the deposit morphology offlowslide is slope angle followed by volume andslope height

(2) With an increase in slope angle the width lengtharea and length-width ratio of flowslide deposits firstincrease and then decrease and the maximum ac-cumulation thickness first decreases and then in-creases And there should be a critical angle in thechange laws of the flowslide deposit morphologythat is between 60deg and 70deg under the experimentalconditions When the volume is 36times10minus 3m3 andthe slope height is 090m the slope angles at 40deg and50deg both result in a single upheaval while slopeangles of 60deg and 70deg both result in a double upheavalwhich has been explained by shear behavior in thesliding body based on the Mohr-Coulomb criterion

12 Advances in Civil Engineering

(3) e apparent friction coefficient of flowslide in-creases nonlinearly as the slope angle increases andshows spatial and lithological difference e greaterthe friction coefficient of the sliding material thehigher the initial benchmark of the curve the largerthe coefficient of velocity restitution the larger theabscissa corresponding to the position of inflectionpoint when the curve changes from flat to rapid

Data Availability

e data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

e authors declare that they have no conflicts of interest

Acknowledgments

is study would not have been possible without financialsupport from the Special Fund for the National NaturalScience Foundation of China under Grant nos 4170229841790442 and 41602359 as well as the Project Supported byNatural Science Basic Research Plan in Shaanxi Province ofChina under Grant no 2017JQ4020

Supplementary Materials

In this section the data on the slope angles and the apparentfriction coefficients are provided as a formation of tableese data are related to 44 rock flowslides in theWenchuanSichuan province China 28 soil flowslides in in the JingyangSouth platform Shaanxi province China and 27 chalkflowslides along the coast of northern Europe (Supple-mentary Materials)

References

[1] O Roche M Attali A Mangeney and A Lucas ldquoOn the run-out distance of geophysical gravitational flows insight fromfluidized granular collapse experimentsrdquo Earth PlanetaryScience Letters vol 311 no 3-4 pp 375ndash385 2011

[2] D Guo M Hamada C He Y Wang and Y Zou ldquoAnempirical model for landslide travel distance prediction inWenchuan earthquake areardquo Landslides vol 11 no 2pp 281ndash291 2014

[3] Y Okura H Kitahara A Kawanami and U KurokawaldquoTopography and volume effects on travel distance of surfacefailurerdquo Engineering Geology vol 67 no 3-4 pp 243ndash2542003

[4] Z Han Y Li J Huang et al ldquoNumerical simulation for run-out extent of debris flows using an improved cellular au-tomaton modelrdquo Bulletin of Engineering Geology and theEnvironment vol 76 no 3 pp 961ndash974 2017

[5] F Guzzeti F Ardizzone M Cardinali M Rossi andD Valigi ldquoLandslide volumes and landslide mobilizationrates in Umbria Central Italyrdquo Earth Planetary Science Let-ters vol 279 no 3-4 pp 222ndash229 2009

[6] J D Chaytor S Uri A R Solow and B Andrews ldquoSizedistribution of submarine landslides along the US Atlanticmarginrdquo Marine Geology vol 264 no 1-2 pp 16ndash27 2009

[7] H Qiu P Cui A D Regmi et al ldquoInfluence of topographyand volume on mobility of loess slides within different slipsurfacesrdquo Catena vol 157 pp 180ndash188 2017

[8] J Hu J Peng G Wang I Javed Y Wang and W LildquoDistribution and characteristics of landslide in LoessPlateau a case study in Shaanxi provincerdquo EngineeringGeology vol 236 pp 89ndash96 2018

[9] L Xu F C Dai Q M Gong L G am and H MinldquoIrrigation-induced loess flow failure in Heifangtai Platformnorth-west Chinardquo Environmental Earth Sciences vol 66no 6 pp 1707ndash1713 2012

[10] R Q Huang W H Liu J P Zhou and X J Pei ldquoRolling testson movement characteristics of rock blocksrdquo Journal ofGeotechnical Engineering vol 29 no 9 pp 1296ndash1302 2007in Chinese

[11] I Manzella and V Labiouse ldquoPhysical modelling to betterunderstand rock avalanchesrdquo in Physical Modelling in Geo-technics Two Volume Set Proceedings of the 7th InternationalConference on Physical Modelling in Geotechnics (ICPMG2010) vol 2 pp 1259ndash1265 Taylor amp Francis Group LondonUK 2010

[12] G B Crosta F V De Blasio M De Caro G VolpiS Imposimato and D Roddeman ldquoModes of propagationand deposition of granular flows onto an erodible substrateexperimental analytical and numerical studyrdquo Landslidesvol 14 no 1 pp 47ndash68 2017

[13] I Manzella and V Labiouse ldquoEmpirical and analytical ana-lyses of laboratory granular flows to investigate rock ava-lanche propagationrdquo Landslides vol 10 no 1 pp 23ndash362013

[14] Y-X Hu H-B Li S-C Qi G Fan and J-W ZhouldquoGranular effects on depositional processes of debris ava-lanchesrdquo KSCE Journal of Civil Engineering vol 24 no 4pp 1116ndash1127 2020

[15] I Goktepeli U Atmaca and A Cakan ldquoInvestigation of heattransfer augmentation between the ribbed plates via Taguchiapproach and computational fluid dynamicsrdquo Journal ofEermal Science vol 29 no 3 pp 647ndash666 2020

[16] M Farin A Mangeney and O Roche ldquoFundamental changesof granular flow dynamics deposition and erosion processesat high slope angles insights from laboratory experimentsrdquoJournal of Geophysical Research Earth Surface vol 119 no 3pp 504ndash532 2014

[17] M Kesseler V Heller and B Turnbull ldquoA laboratory-nu-merical approach for modelling scale effects in dry granularslidesrdquo Landslides vol 15 no 11 pp 2145ndash2159 2018

[18] A Dufresne and T R Davies ldquoLongitudinal ridges in massmovement depositsrdquo Geomorphology vol 105 no 3-4pp 171ndash181 2009

[19] T De Haas L Braat J R F W Leuven I R Lokhorst andM G Kleinhans ldquoEffects of debris flow composition onrunout depositional mechanisms and deposit morphology inlaboratory experimentsrdquo Journal of Geophysical ResearchEarth Surface vol 120 no 9 pp 1949ndash1972 2015

[20] C Goujon N omas and B Dalloz-Dubrujeaud ldquoMono-disperse dry granular flows on inclined planes role ofroughnessrdquo Ee European Physical Journal E vol 11 no 2pp 147ndash157 2003

[21] X-Y Fan S-J Tian and Y-Y Zhang ldquoMass-front velocity ofdry granular flows influenced by the angle of the slope to therunout plane and particle size gradationrdquo Journal of MountainScience vol 13 no 2 pp 234ndash245 2016

[22] G B Crosta F V D Blasio M Locatelli S Imposimato andD Roddeman ldquoLandslides falling onto a shallow erodible

Advances in Civil Engineering 13

the plexiglass with the medium-fine sand e differenceof the friction coefficient between the interface and thesliding body itself also existed in actual flowslides becausethe interface will be lubricated by water or the liquefactioneffect [30 38] erefore the sliding body tends to flowalong the interface of the plexiglass and the sand forms athin sand layer like an erodible bed before the main slidingbody arrived on both inclined and horizontal plates[39 40] It provides an essential condition of the shearaction Figure 13 shows the cross-sectional state of thesliding body at three phases in which phase 3 indicatedby a solid line represents the final deposit state of the

N

(a)

N

(b)

N

(c)

N

(d)

Figure 9 Topography of typical flowslides (a) and (b) located in Folkestone Warren (51deg05prime03Prime N 1deg20prime36Prime E) (c) and (d) located DoverKent (51deg07prime53Prime N 1deg20prime36Prime E) Source from Hutchinson [36]

0 200 400 600 800 10000

10

20

30

40

Dep

th o

f dep

osit

(mm

)

Length of deposit (mm)

40degprofile50degprofile

60degprofile70degprofile

Figure 10 e deposit profiles of flowslides at different angles

N

Front upheavalFront upheavalBack upheavalWash

Figure 11 Flowslide occurred in May 1981 at Le Chien Neuf(49deg46prime59Prime N 0deg25prime19Prime E) Source from Hutchinson [36]

N

Front upheavalBack upheaval

Wash

Figure 12 Flowslide occurred in May 1958 at Wissower KlinkenRugen Germany (54deg32prime09Prime N 13deg40prime42Prime E) Source fromHutchinson [36]

Advances in Civil Engineering 9

sliding body In this paper the Mohr-Coulomb criterion isused in the analysis of the formation of single and doubleupheaval as follows

τ c + σ tanφ (2)Taking the numbers① and② as two microelements in

the sliding body under a stress condition whereby themaximum principal stress σ1 of the microelement at theslope break is parallel to the direction of the inclined plateand the confining pressure σ2 σ3 is perpendicular to thedirection of the inclined plate (the microelements are smallenough so it is considered that σ2 σ3) shear behavioroccurs [22 30 31]

e angle between the shear plane and the maximumprincipal stress is defined as 45degminus (φ2) and the friction angleφ in the sand material is 36deg For a microelement ① atsmaller slope angles of 40deg and 50deg the counterclockwiseangles between the shear action direction and the overallmovement direction of the flowslide are 167deg and 157degrespectively tending to horizontal which is approximatelythe same as the direction of the maximum principal stress inthe state of the microelement ② which will be a source ofpower for the morphology of single upheaval Afterwardsthe microelements located on the horizontal plate movealong the shear plane at an angle of 27deg counterclockwisewith the movement direction of the sliding body on thehorizontal plate under the obstruction of the other particlesin the front and the push of rear particles similar to a thrustnappe in macroscopic performance resulting in depositmorphology as a single upheaval

e formation of double upheaval is more complicatedHere it will be described according to the order of the frontupheaval the wash and the back upheaval First based on thesame calculation method as for the smaller angles thecounterclockwise angles between the shear-action direction ofthe microelement ① and the movement direction of thesliding body on the horizontal plate are 93deg and 83deg at thelarger slope angles of 60deg and 70deg respectively At this pointthe sliding body deposits part of its load under the actions ofshear force and thrust force to form a rudiment of the frontupheaval while the part of the sliding body on the horizontalplate keeps moving under the inertial force and the thrust

force of the back part sliding body (the force state is similar tothe microelement②) and continues development of the frontupheaval gradually Next themicroelements at the back of thefront upheaval stop their shearing motion when the equi-librium condition of equation (3) is satisfied climbing isstopped and the wash is formed However mechanicalconditions still allow the microelements to move horizontallyWhen the equilibrium condition of equation (4) is satisfiedthe horizontal movement stops Assuming that the micro-elements are small enough the equilibrium condition can belisted as follows

m middot g τ middot sin 45deg minusφ2

1113874 1113875 (3)

m middot a τ middot cos 45deg minusφ2

1113874 1113875 (4)

where m is the mass of the microelement g is the accel-eration of gravity τ is shear stress a is the horizontal ac-celeration of the microelement and φ is the friction angle ofsand material

Eventually under the balance of equations (3) and (4)the sliding body that is still moving at the trailing edge is verysmall as the speed so the shear force provided can be re-duced accordingly At this time if the internal shearing forceof the sliding body at the trailing edge is greater than theshearing strength of the sliding body the subsequent slidingbody is deposited at the foot of the slope to form the backupheaval and conducts a localized shear motion based on theshear force similar to the microelement ① however if theshear force is less than the shear strength of the sliding bodythen the sliding body deposits at the slope break to form theback upheaval In general the order of the formation processof the double upheaval is front upheaval wash and backupheaval

e formation of double upheaval morphology is alsoowing to properties of the granular material In this studythere is a double upheaval morphology nevertheless this isnot in the study by Hu [14] e range of the granularparticle from 5times10minus 6m to 4times10minus 4m in this article is ob-viously less than the range of the granular particle from10minus 3m to 16times10minus 2m in the research by Hu [14] ereforethe internal friction angle and cohesion of the experimentalmaterial in this study are larger than those in the research byHu [14]emore the internal friction angle and cohesion ofgranular materials are the more the difficult to disturb thestate of the deposit morphology [19] is When the frontupheaval is formed the granular material with a particle sizeof 5times10minus 6m to 4times10minus 4m is more likely to keep its mor-phology of the front upheaval in the disturbance of the backsliding body compared with a particle size of 10minus 3m to16times10minus 2m If the backsliding body can not disturb theformed front upheaval it will accumulate in the slope breakforming the back upheaval meanwhile whereupon thedouble upheaval is formed

For these change laws as the slope angle increases themaximum width maximum length and area of the depositfirst increase and then decrease similar to the results by Hu[14] at the condition of the mass of fine particles changes

Phase 1Phase 2Phase 3

m1 m2

σ1

σ1

σ2

σ2

α

ατ1 τ1

12

Figure 13 Schematic illustration of shear force of flowslide

10 Advances in Civil Engineering

Based on those results a conclusion was proposed that there isan optimum proportion of fine and coarse particles in thegranular materials because the shear strength and frictioncoefficient are increasing with the mass of fine particles in-creases [41 42] In the condition that the internal frictionangle and cohesion are the smallest the maximum widthmaximum length and area are maximum Hence the de-duction mentioned before is reasonable that there should be acritical slope angle in the changes of the deposit morphology

36 Apparent Friction Coefficient To further understand therules that govern changes in the apparent friction coefficientwith slope angle and the distribution characteristics of theapparent friction coefficient in different lithologies and dif-ferent regions the apparent friction coefficients and corre-sponding slope angle data were collected for research fromthree places ere are 28 soil flowslides in the Jingyang Southplatform (34deg29prime32PrimeN 108deg50prime13Prime E) Shaanxi province China[30] consisting of loesse accumulation area is the even riverterrace comprised of the loess deposit underlain by the siltyclay deposit with interbedded sand layers occasionally Alongthe coast of northern Europe(51deg05prime03Prime N 1deg20prime36Prime E) thereare 27 chalk flows [22] with a porosity ranging from 9 to 52and dry density ranging from 2413 to 1265 kNm3 InWenchuan (31deg00prime00Prime N 103deg40prime00Prime E) Sichuan provinceChina [2] there are 44 rock flowslides that occurred in themountainous area and consisted of the sand-mudstonemainlye accumulation area is relatively flat though it is hard tocompare with flowslides that occurred in Jingyang and coast ofnorthern Europe An important reason why the above threeflowslide groups are taken as research objects is that they havesimilar flow characteristic and long runout distance In ad-dition the topographical condition of the accumulation area issimilar to the model test conditions It can be seen fromFigure 14 that the overall relationship between the apparentfriction coefficient of flowslide and the slope angles shows anonlinear growth relationship is conclusion is consistentwith Crosta [22] e growth relationship is as follows

H

L

tan θ1 + (cos θ + ε sin θ)2 middot ((tan θμ) minus 1)1113960 1113961

(5)

where θ represents the slope angle H represents the slopeheight L represents the moving distance of the flowslide ε isthe coefficient of velocity restitution when the sliding bodyreaches the slope break to change its moving direction tohorizontal μ is the friction coefficient

e difference in regional lithology determines thedifference between the friction coefficient μ and the co-efficient of velocity restitution ε so two different μ valuesand three different ε values are set in this paper When μ is02 and ε is 0 02 or 04 the curve better reflects theoverall trend of the three natural flowslide groups Whenthe friction coefficient μ between the inclined plate andthe experimental material is 055 and ε is 04 the curveshows good correlation with the experimental points Itcan be said that the friction coefficient determines theheight of the initial benchmark of the curve e larger thevalue of the friction coefficient the higher the initial

benchmark of the curve e coefficient of velocity res-titution determines the change of the inflection pointwhen the curve changes from flat to rapid e larger thevalue of the coefficient of velocity restitution the largerthe horizontal coordinates of the inflection point

From the point of energy conversion energy loss at theslope break is closely related to terrain and lithology ekinetic energy loss is small when the junction of slope and theaccumulation area is a curved slope break when the junctionhas sharp edges the kinetic energy loss is larger [13 22]Under the experimental conditions in which slope height Hand volume are constant as the slope angle increases energyloss also increases due to the collision of the flowslide and thehorizontal plate at the slope break [43 44] which results inshorter distance of movement of the sliding body e slopeangle in the chalk flow area is too large and the energy loss isexcessive though the sliding path is lubricated by water be-neath the accumulation area 1-2m e flowslides in Wen-chuan affected by lubrication due to water are few Besidesthere are many obstacles in the movement path because theseflowslides are located in a mountainous area though theaccumulation area is relatively flat so energy loss is largeHowever for the flowslides in Jingyang the path is flat andstraight as well as lubricated by the liquefied effect [31] andthe energy loss is small e flowslides in the model exper-iment fall between the three natural flowslide groups to theleft of chalk flows above loess flowslides and intersectingwith rock flowslides e apparent friction coefficient in thecondition of this model experiment is not the smallest al-though the condition is ideal flat for the reason that there areno lubrications erefore the lubrication effect plays a keyrole in affecting the apparent friction coefficient

4 Discussion

In this paper the Taguchi method was employed inno-vatively to study quantitatively the order of significance

0 20 40 60 80 10000

05

10

15

20

Wenchuan rock flowslidesJingyang loess flowslides

Chalk flowsExperimental flowslides

HL

Slope angle (deg)

ε = 0

ε = 02

ε = 0

2

ε = 04

ε = 0

4

ε = 0μ = 055

μ = 02

tan θ1 + [(cos θ + εsin θ)2 middot (tan θμ ndash1)]

ndashHL =

Figure 14 Comprehensive relationship between slope angle andapparent friction coefficient Formula comes from Crosta [22]

Advances in Civil Engineering 11

and the maximal responses of combinations of factorsaffecting flowslide deposit morphology e calculationresults in Figure 5 confirm that slope angle is the mostinfluential factor e maximum influencing combinationis a volume of 54 times10minus 3 m3 at a slope angle of 70deg and aslope height of 090m which will be useful to evaluate thehazards of potential flowslides Based on the acceptedknowledge that slope angle is the largest influencingfactor further experimental research was performed econditions of small slope angles and large slope anglesshow interesting phenomena of single upheaval anddouble upheaval respectively the Mohr-Coulomb crite-rion is used to explain the action of the shear motionduring the process e comprehensive study of apparentfriction coefficient shows a nonlinear increase with slopeangle as well as spatial and lithological differences indistribution which are affected by the friction coefficientand the coefficient of velocity restitution

Similar model experiments have been carried out toanalyze the movement rule of flowslide [4 12 21] anddeposit morphology [3 16 18 22 25] Previous studiesfocused on the influence of a single factor and the researchcontent is more in-depth [8 11 45 46] but the researchprocess is not systematic enough as flowslide movement anddeposit morphology are determined by multiple factorsincluding volume slope angle slope height sliding bodymaterial and slip bed material Of course some scholarshave conducted multifactor experiments for exampleManzella [13] studied the influence of flowslide volumeinitial height slope of slip bed and the slope break onmovement characteristics but the relative significance ofvarious factors governing the movement of flowslide has notbeen revealed although the factors considered are com-prehensive An orthogonal experiment was used to study themotion characteristics (velocity acceleration) of flowslideunder the influence of various factors and found that slopeangle was the largest influencing factor [33] which isconsistent with the findings of this article though the re-search content was different

e morphology of the double upheaval of flowslide wasalso observed in an experiment similar to the current study[1] which examined wave propagation in the macroscopicview at experiment was also carried out with bulk ma-terials A wave would be formed when the experimentalmaterials slid or fell into the accumulation area the initialwave would accelerate suddenly as it was pushed by asubsequent wave propagating forward e wave crestformed when motion stopped due to resistance may be oneof the reasons for the formation of upheaval and the wavetrough can be regarded as wash [1] in this paper Accordingto the motion rule of the object the flowslide is partiallyreflected at the slope break [22] and then propagates in theform of waves which were observed somewhat during theexperiment of larger slope angles Macroscopic wavepropagation may be the cause of the formation of upheavaland wash on the surface of flowslide but if the phenomenonis deeply understood from the interior of sliding body theinternal shear behavior in the movement of flowslide may bereasonable In addition the morphology of actual flowslides

is also affected by properties of sliding materials [14 19 20]With an increase of the mass of fine particles the length andwidth of the flowslide deposit increase first and then de-crease [14] erefore the difference of properties ofgranular materials is contributing to the formation of thedouble upheaval

e apparent friction coefficient is often involved in thestudy of flowslide but the focus here is on its relationshipwith volume [5 7 9 21 38] ere are few studies that useslope angle as a variable including Lied and Bakkehoi [47]and Okura [3] both of which proposed that the apparentfriction coefficient of flowslide was directly proportional tothe slope angle but the results were based on the analysis ofonly about 20 flowslides so the conclusion was limited eanalysis in this paper involved 99 landslides in three areasand the conclusion is consistent with the viewpoint of Crosta[22] in finding that the apparent friction coefficient increasesnonlinearly as the slope angle increases In the comparisonmade in this study spatial and lithological differences in theapparent friction coefficient are also found

In this paper an experimental idea of constant height ofthe center of gravity is provided but when the angle changesthe sliding distance of the flowslide on the inclined plate(Figure 4(b)) will inevitably increase and the loss of kineticenergy due to friction will increase accordingly ereforefurther improvement of the experimental device is needed toreduce the kinetic energy loss due to the increase in thesliding distance on the inclined plate Modeling should alsoinclude more influencing factors

5 Conclusions

(1) e combined influence factor obtained by theTaguchi method demonstrates that when flowslidevolume is 54times10minus 3m3 the slope angle is 70deg theslope height is 090m and the changes of depositmorphology of the flowslide is the largest e ex-treme differences of maximum length (Ld) maxi-mum thickness (Dd) area (A) and the length-widthratio (LdWd) at three different levels under variousfactors are 277 551 454 and 17 respectively ecalculation results indicate that the factor that has thegreatest impact on the deposit morphology offlowslide is slope angle followed by volume andslope height

(2) With an increase in slope angle the width lengtharea and length-width ratio of flowslide deposits firstincrease and then decrease and the maximum ac-cumulation thickness first decreases and then in-creases And there should be a critical angle in thechange laws of the flowslide deposit morphologythat is between 60deg and 70deg under the experimentalconditions When the volume is 36times10minus 3m3 andthe slope height is 090m the slope angles at 40deg and50deg both result in a single upheaval while slopeangles of 60deg and 70deg both result in a double upheavalwhich has been explained by shear behavior in thesliding body based on the Mohr-Coulomb criterion

12 Advances in Civil Engineering

(3) e apparent friction coefficient of flowslide in-creases nonlinearly as the slope angle increases andshows spatial and lithological difference e greaterthe friction coefficient of the sliding material thehigher the initial benchmark of the curve the largerthe coefficient of velocity restitution the larger theabscissa corresponding to the position of inflectionpoint when the curve changes from flat to rapid

Data Availability

e data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

e authors declare that they have no conflicts of interest

Acknowledgments

is study would not have been possible without financialsupport from the Special Fund for the National NaturalScience Foundation of China under Grant nos 4170229841790442 and 41602359 as well as the Project Supported byNatural Science Basic Research Plan in Shaanxi Province ofChina under Grant no 2017JQ4020

Supplementary Materials

In this section the data on the slope angles and the apparentfriction coefficients are provided as a formation of tableese data are related to 44 rock flowslides in theWenchuanSichuan province China 28 soil flowslides in in the JingyangSouth platform Shaanxi province China and 27 chalkflowslides along the coast of northern Europe (Supple-mentary Materials)

References

[1] O Roche M Attali A Mangeney and A Lucas ldquoOn the run-out distance of geophysical gravitational flows insight fromfluidized granular collapse experimentsrdquo Earth PlanetaryScience Letters vol 311 no 3-4 pp 375ndash385 2011

[2] D Guo M Hamada C He Y Wang and Y Zou ldquoAnempirical model for landslide travel distance prediction inWenchuan earthquake areardquo Landslides vol 11 no 2pp 281ndash291 2014

[3] Y Okura H Kitahara A Kawanami and U KurokawaldquoTopography and volume effects on travel distance of surfacefailurerdquo Engineering Geology vol 67 no 3-4 pp 243ndash2542003

[4] Z Han Y Li J Huang et al ldquoNumerical simulation for run-out extent of debris flows using an improved cellular au-tomaton modelrdquo Bulletin of Engineering Geology and theEnvironment vol 76 no 3 pp 961ndash974 2017

[5] F Guzzeti F Ardizzone M Cardinali M Rossi andD Valigi ldquoLandslide volumes and landslide mobilizationrates in Umbria Central Italyrdquo Earth Planetary Science Let-ters vol 279 no 3-4 pp 222ndash229 2009

[6] J D Chaytor S Uri A R Solow and B Andrews ldquoSizedistribution of submarine landslides along the US Atlanticmarginrdquo Marine Geology vol 264 no 1-2 pp 16ndash27 2009

[7] H Qiu P Cui A D Regmi et al ldquoInfluence of topographyand volume on mobility of loess slides within different slipsurfacesrdquo Catena vol 157 pp 180ndash188 2017

[8] J Hu J Peng G Wang I Javed Y Wang and W LildquoDistribution and characteristics of landslide in LoessPlateau a case study in Shaanxi provincerdquo EngineeringGeology vol 236 pp 89ndash96 2018

[9] L Xu F C Dai Q M Gong L G am and H MinldquoIrrigation-induced loess flow failure in Heifangtai Platformnorth-west Chinardquo Environmental Earth Sciences vol 66no 6 pp 1707ndash1713 2012

[10] R Q Huang W H Liu J P Zhou and X J Pei ldquoRolling testson movement characteristics of rock blocksrdquo Journal ofGeotechnical Engineering vol 29 no 9 pp 1296ndash1302 2007in Chinese

[11] I Manzella and V Labiouse ldquoPhysical modelling to betterunderstand rock avalanchesrdquo in Physical Modelling in Geo-technics Two Volume Set Proceedings of the 7th InternationalConference on Physical Modelling in Geotechnics (ICPMG2010) vol 2 pp 1259ndash1265 Taylor amp Francis Group LondonUK 2010

[12] G B Crosta F V De Blasio M De Caro G VolpiS Imposimato and D Roddeman ldquoModes of propagationand deposition of granular flows onto an erodible substrateexperimental analytical and numerical studyrdquo Landslidesvol 14 no 1 pp 47ndash68 2017

[13] I Manzella and V Labiouse ldquoEmpirical and analytical ana-lyses of laboratory granular flows to investigate rock ava-lanche propagationrdquo Landslides vol 10 no 1 pp 23ndash362013

[14] Y-X Hu H-B Li S-C Qi G Fan and J-W ZhouldquoGranular effects on depositional processes of debris ava-lanchesrdquo KSCE Journal of Civil Engineering vol 24 no 4pp 1116ndash1127 2020

[15] I Goktepeli U Atmaca and A Cakan ldquoInvestigation of heattransfer augmentation between the ribbed plates via Taguchiapproach and computational fluid dynamicsrdquo Journal ofEermal Science vol 29 no 3 pp 647ndash666 2020

[16] M Farin A Mangeney and O Roche ldquoFundamental changesof granular flow dynamics deposition and erosion processesat high slope angles insights from laboratory experimentsrdquoJournal of Geophysical Research Earth Surface vol 119 no 3pp 504ndash532 2014

[17] M Kesseler V Heller and B Turnbull ldquoA laboratory-nu-merical approach for modelling scale effects in dry granularslidesrdquo Landslides vol 15 no 11 pp 2145ndash2159 2018

[18] A Dufresne and T R Davies ldquoLongitudinal ridges in massmovement depositsrdquo Geomorphology vol 105 no 3-4pp 171ndash181 2009

[19] T De Haas L Braat J R F W Leuven I R Lokhorst andM G Kleinhans ldquoEffects of debris flow composition onrunout depositional mechanisms and deposit morphology inlaboratory experimentsrdquo Journal of Geophysical ResearchEarth Surface vol 120 no 9 pp 1949ndash1972 2015

[20] C Goujon N omas and B Dalloz-Dubrujeaud ldquoMono-disperse dry granular flows on inclined planes role ofroughnessrdquo Ee European Physical Journal E vol 11 no 2pp 147ndash157 2003

[21] X-Y Fan S-J Tian and Y-Y Zhang ldquoMass-front velocity ofdry granular flows influenced by the angle of the slope to therunout plane and particle size gradationrdquo Journal of MountainScience vol 13 no 2 pp 234ndash245 2016

[22] G B Crosta F V D Blasio M Locatelli S Imposimato andD Roddeman ldquoLandslides falling onto a shallow erodible

Advances in Civil Engineering 13

sliding body In this paper the Mohr-Coulomb criterion isused in the analysis of the formation of single and doubleupheaval as follows

τ c + σ tanφ (2)Taking the numbers① and② as two microelements in

the sliding body under a stress condition whereby themaximum principal stress σ1 of the microelement at theslope break is parallel to the direction of the inclined plateand the confining pressure σ2 σ3 is perpendicular to thedirection of the inclined plate (the microelements are smallenough so it is considered that σ2 σ3) shear behavioroccurs [22 30 31]

e angle between the shear plane and the maximumprincipal stress is defined as 45degminus (φ2) and the friction angleφ in the sand material is 36deg For a microelement ① atsmaller slope angles of 40deg and 50deg the counterclockwiseangles between the shear action direction and the overallmovement direction of the flowslide are 167deg and 157degrespectively tending to horizontal which is approximatelythe same as the direction of the maximum principal stress inthe state of the microelement ② which will be a source ofpower for the morphology of single upheaval Afterwardsthe microelements located on the horizontal plate movealong the shear plane at an angle of 27deg counterclockwisewith the movement direction of the sliding body on thehorizontal plate under the obstruction of the other particlesin the front and the push of rear particles similar to a thrustnappe in macroscopic performance resulting in depositmorphology as a single upheaval

e formation of double upheaval is more complicatedHere it will be described according to the order of the frontupheaval the wash and the back upheaval First based on thesame calculation method as for the smaller angles thecounterclockwise angles between the shear-action direction ofthe microelement ① and the movement direction of thesliding body on the horizontal plate are 93deg and 83deg at thelarger slope angles of 60deg and 70deg respectively At this pointthe sliding body deposits part of its load under the actions ofshear force and thrust force to form a rudiment of the frontupheaval while the part of the sliding body on the horizontalplate keeps moving under the inertial force and the thrust

force of the back part sliding body (the force state is similar tothe microelement②) and continues development of the frontupheaval gradually Next themicroelements at the back of thefront upheaval stop their shearing motion when the equi-librium condition of equation (3) is satisfied climbing isstopped and the wash is formed However mechanicalconditions still allow the microelements to move horizontallyWhen the equilibrium condition of equation (4) is satisfiedthe horizontal movement stops Assuming that the micro-elements are small enough the equilibrium condition can belisted as follows

m middot g τ middot sin 45deg minusφ2

1113874 1113875 (3)

m middot a τ middot cos 45deg minusφ2

1113874 1113875 (4)

where m is the mass of the microelement g is the accel-eration of gravity τ is shear stress a is the horizontal ac-celeration of the microelement and φ is the friction angle ofsand material

Eventually under the balance of equations (3) and (4)the sliding body that is still moving at the trailing edge is verysmall as the speed so the shear force provided can be re-duced accordingly At this time if the internal shearing forceof the sliding body at the trailing edge is greater than theshearing strength of the sliding body the subsequent slidingbody is deposited at the foot of the slope to form the backupheaval and conducts a localized shear motion based on theshear force similar to the microelement ① however if theshear force is less than the shear strength of the sliding bodythen the sliding body deposits at the slope break to form theback upheaval In general the order of the formation processof the double upheaval is front upheaval wash and backupheaval

e formation of double upheaval morphology is alsoowing to properties of the granular material In this studythere is a double upheaval morphology nevertheless this isnot in the study by Hu [14] e range of the granularparticle from 5times10minus 6m to 4times10minus 4m in this article is ob-viously less than the range of the granular particle from10minus 3m to 16times10minus 2m in the research by Hu [14] ereforethe internal friction angle and cohesion of the experimentalmaterial in this study are larger than those in the research byHu [14]emore the internal friction angle and cohesion ofgranular materials are the more the difficult to disturb thestate of the deposit morphology [19] is When the frontupheaval is formed the granular material with a particle sizeof 5times10minus 6m to 4times10minus 4m is more likely to keep its mor-phology of the front upheaval in the disturbance of the backsliding body compared with a particle size of 10minus 3m to16times10minus 2m If the backsliding body can not disturb theformed front upheaval it will accumulate in the slope breakforming the back upheaval meanwhile whereupon thedouble upheaval is formed

For these change laws as the slope angle increases themaximum width maximum length and area of the depositfirst increase and then decrease similar to the results by Hu[14] at the condition of the mass of fine particles changes

Phase 1Phase 2Phase 3

m1 m2

σ1

σ1

σ2

σ2

α

ατ1 τ1

12

Figure 13 Schematic illustration of shear force of flowslide

10 Advances in Civil Engineering

Based on those results a conclusion was proposed that there isan optimum proportion of fine and coarse particles in thegranular materials because the shear strength and frictioncoefficient are increasing with the mass of fine particles in-creases [41 42] In the condition that the internal frictionangle and cohesion are the smallest the maximum widthmaximum length and area are maximum Hence the de-duction mentioned before is reasonable that there should be acritical slope angle in the changes of the deposit morphology

36 Apparent Friction Coefficient To further understand therules that govern changes in the apparent friction coefficientwith slope angle and the distribution characteristics of theapparent friction coefficient in different lithologies and dif-ferent regions the apparent friction coefficients and corre-sponding slope angle data were collected for research fromthree places ere are 28 soil flowslides in the Jingyang Southplatform (34deg29prime32PrimeN 108deg50prime13Prime E) Shaanxi province China[30] consisting of loesse accumulation area is the even riverterrace comprised of the loess deposit underlain by the siltyclay deposit with interbedded sand layers occasionally Alongthe coast of northern Europe(51deg05prime03Prime N 1deg20prime36Prime E) thereare 27 chalk flows [22] with a porosity ranging from 9 to 52and dry density ranging from 2413 to 1265 kNm3 InWenchuan (31deg00prime00Prime N 103deg40prime00Prime E) Sichuan provinceChina [2] there are 44 rock flowslides that occurred in themountainous area and consisted of the sand-mudstonemainlye accumulation area is relatively flat though it is hard tocompare with flowslides that occurred in Jingyang and coast ofnorthern Europe An important reason why the above threeflowslide groups are taken as research objects is that they havesimilar flow characteristic and long runout distance In ad-dition the topographical condition of the accumulation area issimilar to the model test conditions It can be seen fromFigure 14 that the overall relationship between the apparentfriction coefficient of flowslide and the slope angles shows anonlinear growth relationship is conclusion is consistentwith Crosta [22] e growth relationship is as follows

H

L

tan θ1 + (cos θ + ε sin θ)2 middot ((tan θμ) minus 1)1113960 1113961

(5)

where θ represents the slope angle H represents the slopeheight L represents the moving distance of the flowslide ε isthe coefficient of velocity restitution when the sliding bodyreaches the slope break to change its moving direction tohorizontal μ is the friction coefficient

e difference in regional lithology determines thedifference between the friction coefficient μ and the co-efficient of velocity restitution ε so two different μ valuesand three different ε values are set in this paper When μ is02 and ε is 0 02 or 04 the curve better reflects theoverall trend of the three natural flowslide groups Whenthe friction coefficient μ between the inclined plate andthe experimental material is 055 and ε is 04 the curveshows good correlation with the experimental points Itcan be said that the friction coefficient determines theheight of the initial benchmark of the curve e larger thevalue of the friction coefficient the higher the initial

benchmark of the curve e coefficient of velocity res-titution determines the change of the inflection pointwhen the curve changes from flat to rapid e larger thevalue of the coefficient of velocity restitution the largerthe horizontal coordinates of the inflection point

From the point of energy conversion energy loss at theslope break is closely related to terrain and lithology ekinetic energy loss is small when the junction of slope and theaccumulation area is a curved slope break when the junctionhas sharp edges the kinetic energy loss is larger [13 22]Under the experimental conditions in which slope height Hand volume are constant as the slope angle increases energyloss also increases due to the collision of the flowslide and thehorizontal plate at the slope break [43 44] which results inshorter distance of movement of the sliding body e slopeangle in the chalk flow area is too large and the energy loss isexcessive though the sliding path is lubricated by water be-neath the accumulation area 1-2m e flowslides in Wen-chuan affected by lubrication due to water are few Besidesthere are many obstacles in the movement path because theseflowslides are located in a mountainous area though theaccumulation area is relatively flat so energy loss is largeHowever for the flowslides in Jingyang the path is flat andstraight as well as lubricated by the liquefied effect [31] andthe energy loss is small e flowslides in the model exper-iment fall between the three natural flowslide groups to theleft of chalk flows above loess flowslides and intersectingwith rock flowslides e apparent friction coefficient in thecondition of this model experiment is not the smallest al-though the condition is ideal flat for the reason that there areno lubrications erefore the lubrication effect plays a keyrole in affecting the apparent friction coefficient

4 Discussion

In this paper the Taguchi method was employed inno-vatively to study quantitatively the order of significance

0 20 40 60 80 10000

05

10

15

20

Wenchuan rock flowslidesJingyang loess flowslides

Chalk flowsExperimental flowslides

HL

Slope angle (deg)

ε = 0

ε = 02

ε = 0

2

ε = 04

ε = 0

4

ε = 0μ = 055

μ = 02

tan θ1 + [(cos θ + εsin θ)2 middot (tan θμ ndash1)]

ndashHL =

Figure 14 Comprehensive relationship between slope angle andapparent friction coefficient Formula comes from Crosta [22]

Advances in Civil Engineering 11

and the maximal responses of combinations of factorsaffecting flowslide deposit morphology e calculationresults in Figure 5 confirm that slope angle is the mostinfluential factor e maximum influencing combinationis a volume of 54 times10minus 3 m3 at a slope angle of 70deg and aslope height of 090m which will be useful to evaluate thehazards of potential flowslides Based on the acceptedknowledge that slope angle is the largest influencingfactor further experimental research was performed econditions of small slope angles and large slope anglesshow interesting phenomena of single upheaval anddouble upheaval respectively the Mohr-Coulomb crite-rion is used to explain the action of the shear motionduring the process e comprehensive study of apparentfriction coefficient shows a nonlinear increase with slopeangle as well as spatial and lithological differences indistribution which are affected by the friction coefficientand the coefficient of velocity restitution

Similar model experiments have been carried out toanalyze the movement rule of flowslide [4 12 21] anddeposit morphology [3 16 18 22 25] Previous studiesfocused on the influence of a single factor and the researchcontent is more in-depth [8 11 45 46] but the researchprocess is not systematic enough as flowslide movement anddeposit morphology are determined by multiple factorsincluding volume slope angle slope height sliding bodymaterial and slip bed material Of course some scholarshave conducted multifactor experiments for exampleManzella [13] studied the influence of flowslide volumeinitial height slope of slip bed and the slope break onmovement characteristics but the relative significance ofvarious factors governing the movement of flowslide has notbeen revealed although the factors considered are com-prehensive An orthogonal experiment was used to study themotion characteristics (velocity acceleration) of flowslideunder the influence of various factors and found that slopeangle was the largest influencing factor [33] which isconsistent with the findings of this article though the re-search content was different

e morphology of the double upheaval of flowslide wasalso observed in an experiment similar to the current study[1] which examined wave propagation in the macroscopicview at experiment was also carried out with bulk ma-terials A wave would be formed when the experimentalmaterials slid or fell into the accumulation area the initialwave would accelerate suddenly as it was pushed by asubsequent wave propagating forward e wave crestformed when motion stopped due to resistance may be oneof the reasons for the formation of upheaval and the wavetrough can be regarded as wash [1] in this paper Accordingto the motion rule of the object the flowslide is partiallyreflected at the slope break [22] and then propagates in theform of waves which were observed somewhat during theexperiment of larger slope angles Macroscopic wavepropagation may be the cause of the formation of upheavaland wash on the surface of flowslide but if the phenomenonis deeply understood from the interior of sliding body theinternal shear behavior in the movement of flowslide may bereasonable In addition the morphology of actual flowslides

is also affected by properties of sliding materials [14 19 20]With an increase of the mass of fine particles the length andwidth of the flowslide deposit increase first and then de-crease [14] erefore the difference of properties ofgranular materials is contributing to the formation of thedouble upheaval

e apparent friction coefficient is often involved in thestudy of flowslide but the focus here is on its relationshipwith volume [5 7 9 21 38] ere are few studies that useslope angle as a variable including Lied and Bakkehoi [47]and Okura [3] both of which proposed that the apparentfriction coefficient of flowslide was directly proportional tothe slope angle but the results were based on the analysis ofonly about 20 flowslides so the conclusion was limited eanalysis in this paper involved 99 landslides in three areasand the conclusion is consistent with the viewpoint of Crosta[22] in finding that the apparent friction coefficient increasesnonlinearly as the slope angle increases In the comparisonmade in this study spatial and lithological differences in theapparent friction coefficient are also found

In this paper an experimental idea of constant height ofthe center of gravity is provided but when the angle changesthe sliding distance of the flowslide on the inclined plate(Figure 4(b)) will inevitably increase and the loss of kineticenergy due to friction will increase accordingly ereforefurther improvement of the experimental device is needed toreduce the kinetic energy loss due to the increase in thesliding distance on the inclined plate Modeling should alsoinclude more influencing factors

5 Conclusions

(1) e combined influence factor obtained by theTaguchi method demonstrates that when flowslidevolume is 54times10minus 3m3 the slope angle is 70deg theslope height is 090m and the changes of depositmorphology of the flowslide is the largest e ex-treme differences of maximum length (Ld) maxi-mum thickness (Dd) area (A) and the length-widthratio (LdWd) at three different levels under variousfactors are 277 551 454 and 17 respectively ecalculation results indicate that the factor that has thegreatest impact on the deposit morphology offlowslide is slope angle followed by volume andslope height

(2) With an increase in slope angle the width lengtharea and length-width ratio of flowslide deposits firstincrease and then decrease and the maximum ac-cumulation thickness first decreases and then in-creases And there should be a critical angle in thechange laws of the flowslide deposit morphologythat is between 60deg and 70deg under the experimentalconditions When the volume is 36times10minus 3m3 andthe slope height is 090m the slope angles at 40deg and50deg both result in a single upheaval while slopeangles of 60deg and 70deg both result in a double upheavalwhich has been explained by shear behavior in thesliding body based on the Mohr-Coulomb criterion

12 Advances in Civil Engineering

(3) e apparent friction coefficient of flowslide in-creases nonlinearly as the slope angle increases andshows spatial and lithological difference e greaterthe friction coefficient of the sliding material thehigher the initial benchmark of the curve the largerthe coefficient of velocity restitution the larger theabscissa corresponding to the position of inflectionpoint when the curve changes from flat to rapid

Data Availability

e data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

e authors declare that they have no conflicts of interest

Acknowledgments

is study would not have been possible without financialsupport from the Special Fund for the National NaturalScience Foundation of China under Grant nos 4170229841790442 and 41602359 as well as the Project Supported byNatural Science Basic Research Plan in Shaanxi Province ofChina under Grant no 2017JQ4020

Supplementary Materials

In this section the data on the slope angles and the apparentfriction coefficients are provided as a formation of tableese data are related to 44 rock flowslides in theWenchuanSichuan province China 28 soil flowslides in in the JingyangSouth platform Shaanxi province China and 27 chalkflowslides along the coast of northern Europe (Supple-mentary Materials)

References

[1] O Roche M Attali A Mangeney and A Lucas ldquoOn the run-out distance of geophysical gravitational flows insight fromfluidized granular collapse experimentsrdquo Earth PlanetaryScience Letters vol 311 no 3-4 pp 375ndash385 2011

[2] D Guo M Hamada C He Y Wang and Y Zou ldquoAnempirical model for landslide travel distance prediction inWenchuan earthquake areardquo Landslides vol 11 no 2pp 281ndash291 2014

[3] Y Okura H Kitahara A Kawanami and U KurokawaldquoTopography and volume effects on travel distance of surfacefailurerdquo Engineering Geology vol 67 no 3-4 pp 243ndash2542003

[4] Z Han Y Li J Huang et al ldquoNumerical simulation for run-out extent of debris flows using an improved cellular au-tomaton modelrdquo Bulletin of Engineering Geology and theEnvironment vol 76 no 3 pp 961ndash974 2017

[5] F Guzzeti F Ardizzone M Cardinali M Rossi andD Valigi ldquoLandslide volumes and landslide mobilizationrates in Umbria Central Italyrdquo Earth Planetary Science Let-ters vol 279 no 3-4 pp 222ndash229 2009

[6] J D Chaytor S Uri A R Solow and B Andrews ldquoSizedistribution of submarine landslides along the US Atlanticmarginrdquo Marine Geology vol 264 no 1-2 pp 16ndash27 2009

[7] H Qiu P Cui A D Regmi et al ldquoInfluence of topographyand volume on mobility of loess slides within different slipsurfacesrdquo Catena vol 157 pp 180ndash188 2017

[8] J Hu J Peng G Wang I Javed Y Wang and W LildquoDistribution and characteristics of landslide in LoessPlateau a case study in Shaanxi provincerdquo EngineeringGeology vol 236 pp 89ndash96 2018

[9] L Xu F C Dai Q M Gong L G am and H MinldquoIrrigation-induced loess flow failure in Heifangtai Platformnorth-west Chinardquo Environmental Earth Sciences vol 66no 6 pp 1707ndash1713 2012

[10] R Q Huang W H Liu J P Zhou and X J Pei ldquoRolling testson movement characteristics of rock blocksrdquo Journal ofGeotechnical Engineering vol 29 no 9 pp 1296ndash1302 2007in Chinese

[11] I Manzella and V Labiouse ldquoPhysical modelling to betterunderstand rock avalanchesrdquo in Physical Modelling in Geo-technics Two Volume Set Proceedings of the 7th InternationalConference on Physical Modelling in Geotechnics (ICPMG2010) vol 2 pp 1259ndash1265 Taylor amp Francis Group LondonUK 2010

[12] G B Crosta F V De Blasio M De Caro G VolpiS Imposimato and D Roddeman ldquoModes of propagationand deposition of granular flows onto an erodible substrateexperimental analytical and numerical studyrdquo Landslidesvol 14 no 1 pp 47ndash68 2017

[13] I Manzella and V Labiouse ldquoEmpirical and analytical ana-lyses of laboratory granular flows to investigate rock ava-lanche propagationrdquo Landslides vol 10 no 1 pp 23ndash362013

[14] Y-X Hu H-B Li S-C Qi G Fan and J-W ZhouldquoGranular effects on depositional processes of debris ava-lanchesrdquo KSCE Journal of Civil Engineering vol 24 no 4pp 1116ndash1127 2020

[15] I Goktepeli U Atmaca and A Cakan ldquoInvestigation of heattransfer augmentation between the ribbed plates via Taguchiapproach and computational fluid dynamicsrdquo Journal ofEermal Science vol 29 no 3 pp 647ndash666 2020

[16] M Farin A Mangeney and O Roche ldquoFundamental changesof granular flow dynamics deposition and erosion processesat high slope angles insights from laboratory experimentsrdquoJournal of Geophysical Research Earth Surface vol 119 no 3pp 504ndash532 2014

[17] M Kesseler V Heller and B Turnbull ldquoA laboratory-nu-merical approach for modelling scale effects in dry granularslidesrdquo Landslides vol 15 no 11 pp 2145ndash2159 2018

[18] A Dufresne and T R Davies ldquoLongitudinal ridges in massmovement depositsrdquo Geomorphology vol 105 no 3-4pp 171ndash181 2009

[19] T De Haas L Braat J R F W Leuven I R Lokhorst andM G Kleinhans ldquoEffects of debris flow composition onrunout depositional mechanisms and deposit morphology inlaboratory experimentsrdquo Journal of Geophysical ResearchEarth Surface vol 120 no 9 pp 1949ndash1972 2015

[20] C Goujon N omas and B Dalloz-Dubrujeaud ldquoMono-disperse dry granular flows on inclined planes role ofroughnessrdquo Ee European Physical Journal E vol 11 no 2pp 147ndash157 2003

[21] X-Y Fan S-J Tian and Y-Y Zhang ldquoMass-front velocity ofdry granular flows influenced by the angle of the slope to therunout plane and particle size gradationrdquo Journal of MountainScience vol 13 no 2 pp 234ndash245 2016

[22] G B Crosta F V D Blasio M Locatelli S Imposimato andD Roddeman ldquoLandslides falling onto a shallow erodible

Advances in Civil Engineering 13

Based on those results a conclusion was proposed that there isan optimum proportion of fine and coarse particles in thegranular materials because the shear strength and frictioncoefficient are increasing with the mass of fine particles in-creases [41 42] In the condition that the internal frictionangle and cohesion are the smallest the maximum widthmaximum length and area are maximum Hence the de-duction mentioned before is reasonable that there should be acritical slope angle in the changes of the deposit morphology

36 Apparent Friction Coefficient To further understand therules that govern changes in the apparent friction coefficientwith slope angle and the distribution characteristics of theapparent friction coefficient in different lithologies and dif-ferent regions the apparent friction coefficients and corre-sponding slope angle data were collected for research fromthree places ere are 28 soil flowslides in the Jingyang Southplatform (34deg29prime32PrimeN 108deg50prime13Prime E) Shaanxi province China[30] consisting of loesse accumulation area is the even riverterrace comprised of the loess deposit underlain by the siltyclay deposit with interbedded sand layers occasionally Alongthe coast of northern Europe(51deg05prime03Prime N 1deg20prime36Prime E) thereare 27 chalk flows [22] with a porosity ranging from 9 to 52and dry density ranging from 2413 to 1265 kNm3 InWenchuan (31deg00prime00Prime N 103deg40prime00Prime E) Sichuan provinceChina [2] there are 44 rock flowslides that occurred in themountainous area and consisted of the sand-mudstonemainlye accumulation area is relatively flat though it is hard tocompare with flowslides that occurred in Jingyang and coast ofnorthern Europe An important reason why the above threeflowslide groups are taken as research objects is that they havesimilar flow characteristic and long runout distance In ad-dition the topographical condition of the accumulation area issimilar to the model test conditions It can be seen fromFigure 14 that the overall relationship between the apparentfriction coefficient of flowslide and the slope angles shows anonlinear growth relationship is conclusion is consistentwith Crosta [22] e growth relationship is as follows

H

L

tan θ1 + (cos θ + ε sin θ)2 middot ((tan θμ) minus 1)1113960 1113961

(5)

where θ represents the slope angle H represents the slopeheight L represents the moving distance of the flowslide ε isthe coefficient of velocity restitution when the sliding bodyreaches the slope break to change its moving direction tohorizontal μ is the friction coefficient

e difference in regional lithology determines thedifference between the friction coefficient μ and the co-efficient of velocity restitution ε so two different μ valuesand three different ε values are set in this paper When μ is02 and ε is 0 02 or 04 the curve better reflects theoverall trend of the three natural flowslide groups Whenthe friction coefficient μ between the inclined plate andthe experimental material is 055 and ε is 04 the curveshows good correlation with the experimental points Itcan be said that the friction coefficient determines theheight of the initial benchmark of the curve e larger thevalue of the friction coefficient the higher the initial

benchmark of the curve e coefficient of velocity res-titution determines the change of the inflection pointwhen the curve changes from flat to rapid e larger thevalue of the coefficient of velocity restitution the largerthe horizontal coordinates of the inflection point

From the point of energy conversion energy loss at theslope break is closely related to terrain and lithology ekinetic energy loss is small when the junction of slope and theaccumulation area is a curved slope break when the junctionhas sharp edges the kinetic energy loss is larger [13 22]Under the experimental conditions in which slope height Hand volume are constant as the slope angle increases energyloss also increases due to the collision of the flowslide and thehorizontal plate at the slope break [43 44] which results inshorter distance of movement of the sliding body e slopeangle in the chalk flow area is too large and the energy loss isexcessive though the sliding path is lubricated by water be-neath the accumulation area 1-2m e flowslides in Wen-chuan affected by lubrication due to water are few Besidesthere are many obstacles in the movement path because theseflowslides are located in a mountainous area though theaccumulation area is relatively flat so energy loss is largeHowever for the flowslides in Jingyang the path is flat andstraight as well as lubricated by the liquefied effect [31] andthe energy loss is small e flowslides in the model exper-iment fall between the three natural flowslide groups to theleft of chalk flows above loess flowslides and intersectingwith rock flowslides e apparent friction coefficient in thecondition of this model experiment is not the smallest al-though the condition is ideal flat for the reason that there areno lubrications erefore the lubrication effect plays a keyrole in affecting the apparent friction coefficient

4 Discussion

In this paper the Taguchi method was employed inno-vatively to study quantitatively the order of significance

0 20 40 60 80 10000

05

10

15

20

Wenchuan rock flowslidesJingyang loess flowslides

Chalk flowsExperimental flowslides

HL

Slope angle (deg)

ε = 0

ε = 02

ε = 0

2

ε = 04

ε = 0

4

ε = 0μ = 055

μ = 02

tan θ1 + [(cos θ + εsin θ)2 middot (tan θμ ndash1)]

ndashHL =

Figure 14 Comprehensive relationship between slope angle andapparent friction coefficient Formula comes from Crosta [22]

Advances in Civil Engineering 11

and the maximal responses of combinations of factorsaffecting flowslide deposit morphology e calculationresults in Figure 5 confirm that slope angle is the mostinfluential factor e maximum influencing combinationis a volume of 54 times10minus 3 m3 at a slope angle of 70deg and aslope height of 090m which will be useful to evaluate thehazards of potential flowslides Based on the acceptedknowledge that slope angle is the largest influencingfactor further experimental research was performed econditions of small slope angles and large slope anglesshow interesting phenomena of single upheaval anddouble upheaval respectively the Mohr-Coulomb crite-rion is used to explain the action of the shear motionduring the process e comprehensive study of apparentfriction coefficient shows a nonlinear increase with slopeangle as well as spatial and lithological differences indistribution which are affected by the friction coefficientand the coefficient of velocity restitution

Similar model experiments have been carried out toanalyze the movement rule of flowslide [4 12 21] anddeposit morphology [3 16 18 22 25] Previous studiesfocused on the influence of a single factor and the researchcontent is more in-depth [8 11 45 46] but the researchprocess is not systematic enough as flowslide movement anddeposit morphology are determined by multiple factorsincluding volume slope angle slope height sliding bodymaterial and slip bed material Of course some scholarshave conducted multifactor experiments for exampleManzella [13] studied the influence of flowslide volumeinitial height slope of slip bed and the slope break onmovement characteristics but the relative significance ofvarious factors governing the movement of flowslide has notbeen revealed although the factors considered are com-prehensive An orthogonal experiment was used to study themotion characteristics (velocity acceleration) of flowslideunder the influence of various factors and found that slopeangle was the largest influencing factor [33] which isconsistent with the findings of this article though the re-search content was different

e morphology of the double upheaval of flowslide wasalso observed in an experiment similar to the current study[1] which examined wave propagation in the macroscopicview at experiment was also carried out with bulk ma-terials A wave would be formed when the experimentalmaterials slid or fell into the accumulation area the initialwave would accelerate suddenly as it was pushed by asubsequent wave propagating forward e wave crestformed when motion stopped due to resistance may be oneof the reasons for the formation of upheaval and the wavetrough can be regarded as wash [1] in this paper Accordingto the motion rule of the object the flowslide is partiallyreflected at the slope break [22] and then propagates in theform of waves which were observed somewhat during theexperiment of larger slope angles Macroscopic wavepropagation may be the cause of the formation of upheavaland wash on the surface of flowslide but if the phenomenonis deeply understood from the interior of sliding body theinternal shear behavior in the movement of flowslide may bereasonable In addition the morphology of actual flowslides

is also affected by properties of sliding materials [14 19 20]With an increase of the mass of fine particles the length andwidth of the flowslide deposit increase first and then de-crease [14] erefore the difference of properties ofgranular materials is contributing to the formation of thedouble upheaval

e apparent friction coefficient is often involved in thestudy of flowslide but the focus here is on its relationshipwith volume [5 7 9 21 38] ere are few studies that useslope angle as a variable including Lied and Bakkehoi [47]and Okura [3] both of which proposed that the apparentfriction coefficient of flowslide was directly proportional tothe slope angle but the results were based on the analysis ofonly about 20 flowslides so the conclusion was limited eanalysis in this paper involved 99 landslides in three areasand the conclusion is consistent with the viewpoint of Crosta[22] in finding that the apparent friction coefficient increasesnonlinearly as the slope angle increases In the comparisonmade in this study spatial and lithological differences in theapparent friction coefficient are also found

In this paper an experimental idea of constant height ofthe center of gravity is provided but when the angle changesthe sliding distance of the flowslide on the inclined plate(Figure 4(b)) will inevitably increase and the loss of kineticenergy due to friction will increase accordingly ereforefurther improvement of the experimental device is needed toreduce the kinetic energy loss due to the increase in thesliding distance on the inclined plate Modeling should alsoinclude more influencing factors

5 Conclusions

(1) e combined influence factor obtained by theTaguchi method demonstrates that when flowslidevolume is 54times10minus 3m3 the slope angle is 70deg theslope height is 090m and the changes of depositmorphology of the flowslide is the largest e ex-treme differences of maximum length (Ld) maxi-mum thickness (Dd) area (A) and the length-widthratio (LdWd) at three different levels under variousfactors are 277 551 454 and 17 respectively ecalculation results indicate that the factor that has thegreatest impact on the deposit morphology offlowslide is slope angle followed by volume andslope height

(2) With an increase in slope angle the width lengtharea and length-width ratio of flowslide deposits firstincrease and then decrease and the maximum ac-cumulation thickness first decreases and then in-creases And there should be a critical angle in thechange laws of the flowslide deposit morphologythat is between 60deg and 70deg under the experimentalconditions When the volume is 36times10minus 3m3 andthe slope height is 090m the slope angles at 40deg and50deg both result in a single upheaval while slopeangles of 60deg and 70deg both result in a double upheavalwhich has been explained by shear behavior in thesliding body based on the Mohr-Coulomb criterion

12 Advances in Civil Engineering

(3) e apparent friction coefficient of flowslide in-creases nonlinearly as the slope angle increases andshows spatial and lithological difference e greaterthe friction coefficient of the sliding material thehigher the initial benchmark of the curve the largerthe coefficient of velocity restitution the larger theabscissa corresponding to the position of inflectionpoint when the curve changes from flat to rapid

Data Availability

e data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

e authors declare that they have no conflicts of interest

Acknowledgments

is study would not have been possible without financialsupport from the Special Fund for the National NaturalScience Foundation of China under Grant nos 4170229841790442 and 41602359 as well as the Project Supported byNatural Science Basic Research Plan in Shaanxi Province ofChina under Grant no 2017JQ4020

Supplementary Materials

In this section the data on the slope angles and the apparentfriction coefficients are provided as a formation of tableese data are related to 44 rock flowslides in theWenchuanSichuan province China 28 soil flowslides in in the JingyangSouth platform Shaanxi province China and 27 chalkflowslides along the coast of northern Europe (Supple-mentary Materials)

References

[1] O Roche M Attali A Mangeney and A Lucas ldquoOn the run-out distance of geophysical gravitational flows insight fromfluidized granular collapse experimentsrdquo Earth PlanetaryScience Letters vol 311 no 3-4 pp 375ndash385 2011

[2] D Guo M Hamada C He Y Wang and Y Zou ldquoAnempirical model for landslide travel distance prediction inWenchuan earthquake areardquo Landslides vol 11 no 2pp 281ndash291 2014

[3] Y Okura H Kitahara A Kawanami and U KurokawaldquoTopography and volume effects on travel distance of surfacefailurerdquo Engineering Geology vol 67 no 3-4 pp 243ndash2542003

[4] Z Han Y Li J Huang et al ldquoNumerical simulation for run-out extent of debris flows using an improved cellular au-tomaton modelrdquo Bulletin of Engineering Geology and theEnvironment vol 76 no 3 pp 961ndash974 2017

[5] F Guzzeti F Ardizzone M Cardinali M Rossi andD Valigi ldquoLandslide volumes and landslide mobilizationrates in Umbria Central Italyrdquo Earth Planetary Science Let-ters vol 279 no 3-4 pp 222ndash229 2009

[6] J D Chaytor S Uri A R Solow and B Andrews ldquoSizedistribution of submarine landslides along the US Atlanticmarginrdquo Marine Geology vol 264 no 1-2 pp 16ndash27 2009

[7] H Qiu P Cui A D Regmi et al ldquoInfluence of topographyand volume on mobility of loess slides within different slipsurfacesrdquo Catena vol 157 pp 180ndash188 2017

[8] J Hu J Peng G Wang I Javed Y Wang and W LildquoDistribution and characteristics of landslide in LoessPlateau a case study in Shaanxi provincerdquo EngineeringGeology vol 236 pp 89ndash96 2018

[9] L Xu F C Dai Q M Gong L G am and H MinldquoIrrigation-induced loess flow failure in Heifangtai Platformnorth-west Chinardquo Environmental Earth Sciences vol 66no 6 pp 1707ndash1713 2012

[10] R Q Huang W H Liu J P Zhou and X J Pei ldquoRolling testson movement characteristics of rock blocksrdquo Journal ofGeotechnical Engineering vol 29 no 9 pp 1296ndash1302 2007in Chinese

[11] I Manzella and V Labiouse ldquoPhysical modelling to betterunderstand rock avalanchesrdquo in Physical Modelling in Geo-technics Two Volume Set Proceedings of the 7th InternationalConference on Physical Modelling in Geotechnics (ICPMG2010) vol 2 pp 1259ndash1265 Taylor amp Francis Group LondonUK 2010

[12] G B Crosta F V De Blasio M De Caro G VolpiS Imposimato and D Roddeman ldquoModes of propagationand deposition of granular flows onto an erodible substrateexperimental analytical and numerical studyrdquo Landslidesvol 14 no 1 pp 47ndash68 2017

[13] I Manzella and V Labiouse ldquoEmpirical and analytical ana-lyses of laboratory granular flows to investigate rock ava-lanche propagationrdquo Landslides vol 10 no 1 pp 23ndash362013

[14] Y-X Hu H-B Li S-C Qi G Fan and J-W ZhouldquoGranular effects on depositional processes of debris ava-lanchesrdquo KSCE Journal of Civil Engineering vol 24 no 4pp 1116ndash1127 2020

[15] I Goktepeli U Atmaca and A Cakan ldquoInvestigation of heattransfer augmentation between the ribbed plates via Taguchiapproach and computational fluid dynamicsrdquo Journal ofEermal Science vol 29 no 3 pp 647ndash666 2020

[16] M Farin A Mangeney and O Roche ldquoFundamental changesof granular flow dynamics deposition and erosion processesat high slope angles insights from laboratory experimentsrdquoJournal of Geophysical Research Earth Surface vol 119 no 3pp 504ndash532 2014

[17] M Kesseler V Heller and B Turnbull ldquoA laboratory-nu-merical approach for modelling scale effects in dry granularslidesrdquo Landslides vol 15 no 11 pp 2145ndash2159 2018

[18] A Dufresne and T R Davies ldquoLongitudinal ridges in massmovement depositsrdquo Geomorphology vol 105 no 3-4pp 171ndash181 2009

[19] T De Haas L Braat J R F W Leuven I R Lokhorst andM G Kleinhans ldquoEffects of debris flow composition onrunout depositional mechanisms and deposit morphology inlaboratory experimentsrdquo Journal of Geophysical ResearchEarth Surface vol 120 no 9 pp 1949ndash1972 2015

[20] C Goujon N omas and B Dalloz-Dubrujeaud ldquoMono-disperse dry granular flows on inclined planes role ofroughnessrdquo Ee European Physical Journal E vol 11 no 2pp 147ndash157 2003

[21] X-Y Fan S-J Tian and Y-Y Zhang ldquoMass-front velocity ofdry granular flows influenced by the angle of the slope to therunout plane and particle size gradationrdquo Journal of MountainScience vol 13 no 2 pp 234ndash245 2016

[22] G B Crosta F V D Blasio M Locatelli S Imposimato andD Roddeman ldquoLandslides falling onto a shallow erodible

Advances in Civil Engineering 13

and the maximal responses of combinations of factorsaffecting flowslide deposit morphology e calculationresults in Figure 5 confirm that slope angle is the mostinfluential factor e maximum influencing combinationis a volume of 54 times10minus 3 m3 at a slope angle of 70deg and aslope height of 090m which will be useful to evaluate thehazards of potential flowslides Based on the acceptedknowledge that slope angle is the largest influencingfactor further experimental research was performed econditions of small slope angles and large slope anglesshow interesting phenomena of single upheaval anddouble upheaval respectively the Mohr-Coulomb crite-rion is used to explain the action of the shear motionduring the process e comprehensive study of apparentfriction coefficient shows a nonlinear increase with slopeangle as well as spatial and lithological differences indistribution which are affected by the friction coefficientand the coefficient of velocity restitution

Similar model experiments have been carried out toanalyze the movement rule of flowslide [4 12 21] anddeposit morphology [3 16 18 22 25] Previous studiesfocused on the influence of a single factor and the researchcontent is more in-depth [8 11 45 46] but the researchprocess is not systematic enough as flowslide movement anddeposit morphology are determined by multiple factorsincluding volume slope angle slope height sliding bodymaterial and slip bed material Of course some scholarshave conducted multifactor experiments for exampleManzella [13] studied the influence of flowslide volumeinitial height slope of slip bed and the slope break onmovement characteristics but the relative significance ofvarious factors governing the movement of flowslide has notbeen revealed although the factors considered are com-prehensive An orthogonal experiment was used to study themotion characteristics (velocity acceleration) of flowslideunder the influence of various factors and found that slopeangle was the largest influencing factor [33] which isconsistent with the findings of this article though the re-search content was different

e morphology of the double upheaval of flowslide wasalso observed in an experiment similar to the current study[1] which examined wave propagation in the macroscopicview at experiment was also carried out with bulk ma-terials A wave would be formed when the experimentalmaterials slid or fell into the accumulation area the initialwave would accelerate suddenly as it was pushed by asubsequent wave propagating forward e wave crestformed when motion stopped due to resistance may be oneof the reasons for the formation of upheaval and the wavetrough can be regarded as wash [1] in this paper Accordingto the motion rule of the object the flowslide is partiallyreflected at the slope break [22] and then propagates in theform of waves which were observed somewhat during theexperiment of larger slope angles Macroscopic wavepropagation may be the cause of the formation of upheavaland wash on the surface of flowslide but if the phenomenonis deeply understood from the interior of sliding body theinternal shear behavior in the movement of flowslide may bereasonable In addition the morphology of actual flowslides

is also affected by properties of sliding materials [14 19 20]With an increase of the mass of fine particles the length andwidth of the flowslide deposit increase first and then de-crease [14] erefore the difference of properties ofgranular materials is contributing to the formation of thedouble upheaval

e apparent friction coefficient is often involved in thestudy of flowslide but the focus here is on its relationshipwith volume [5 7 9 21 38] ere are few studies that useslope angle as a variable including Lied and Bakkehoi [47]and Okura [3] both of which proposed that the apparentfriction coefficient of flowslide was directly proportional tothe slope angle but the results were based on the analysis ofonly about 20 flowslides so the conclusion was limited eanalysis in this paper involved 99 landslides in three areasand the conclusion is consistent with the viewpoint of Crosta[22] in finding that the apparent friction coefficient increasesnonlinearly as the slope angle increases In the comparisonmade in this study spatial and lithological differences in theapparent friction coefficient are also found

In this paper an experimental idea of constant height ofthe center of gravity is provided but when the angle changesthe sliding distance of the flowslide on the inclined plate(Figure 4(b)) will inevitably increase and the loss of kineticenergy due to friction will increase accordingly ereforefurther improvement of the experimental device is needed toreduce the kinetic energy loss due to the increase in thesliding distance on the inclined plate Modeling should alsoinclude more influencing factors

5 Conclusions

(1) e combined influence factor obtained by theTaguchi method demonstrates that when flowslidevolume is 54times10minus 3m3 the slope angle is 70deg theslope height is 090m and the changes of depositmorphology of the flowslide is the largest e ex-treme differences of maximum length (Ld) maxi-mum thickness (Dd) area (A) and the length-widthratio (LdWd) at three different levels under variousfactors are 277 551 454 and 17 respectively ecalculation results indicate that the factor that has thegreatest impact on the deposit morphology offlowslide is slope angle followed by volume andslope height

(2) With an increase in slope angle the width lengtharea and length-width ratio of flowslide deposits firstincrease and then decrease and the maximum ac-cumulation thickness first decreases and then in-creases And there should be a critical angle in thechange laws of the flowslide deposit morphologythat is between 60deg and 70deg under the experimentalconditions When the volume is 36times10minus 3m3 andthe slope height is 090m the slope angles at 40deg and50deg both result in a single upheaval while slopeangles of 60deg and 70deg both result in a double upheavalwhich has been explained by shear behavior in thesliding body based on the Mohr-Coulomb criterion

12 Advances in Civil Engineering

(3) e apparent friction coefficient of flowslide in-creases nonlinearly as the slope angle increases andshows spatial and lithological difference e greaterthe friction coefficient of the sliding material thehigher the initial benchmark of the curve the largerthe coefficient of velocity restitution the larger theabscissa corresponding to the position of inflectionpoint when the curve changes from flat to rapid

Data Availability

e data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

e authors declare that they have no conflicts of interest

Acknowledgments

is study would not have been possible without financialsupport from the Special Fund for the National NaturalScience Foundation of China under Grant nos 4170229841790442 and 41602359 as well as the Project Supported byNatural Science Basic Research Plan in Shaanxi Province ofChina under Grant no 2017JQ4020

Supplementary Materials

In this section the data on the slope angles and the apparentfriction coefficients are provided as a formation of tableese data are related to 44 rock flowslides in theWenchuanSichuan province China 28 soil flowslides in in the JingyangSouth platform Shaanxi province China and 27 chalkflowslides along the coast of northern Europe (Supple-mentary Materials)

References

[1] O Roche M Attali A Mangeney and A Lucas ldquoOn the run-out distance of geophysical gravitational flows insight fromfluidized granular collapse experimentsrdquo Earth PlanetaryScience Letters vol 311 no 3-4 pp 375ndash385 2011

[2] D Guo M Hamada C He Y Wang and Y Zou ldquoAnempirical model for landslide travel distance prediction inWenchuan earthquake areardquo Landslides vol 11 no 2pp 281ndash291 2014

[3] Y Okura H Kitahara A Kawanami and U KurokawaldquoTopography and volume effects on travel distance of surfacefailurerdquo Engineering Geology vol 67 no 3-4 pp 243ndash2542003

[4] Z Han Y Li J Huang et al ldquoNumerical simulation for run-out extent of debris flows using an improved cellular au-tomaton modelrdquo Bulletin of Engineering Geology and theEnvironment vol 76 no 3 pp 961ndash974 2017

[5] F Guzzeti F Ardizzone M Cardinali M Rossi andD Valigi ldquoLandslide volumes and landslide mobilizationrates in Umbria Central Italyrdquo Earth Planetary Science Let-ters vol 279 no 3-4 pp 222ndash229 2009

[6] J D Chaytor S Uri A R Solow and B Andrews ldquoSizedistribution of submarine landslides along the US Atlanticmarginrdquo Marine Geology vol 264 no 1-2 pp 16ndash27 2009

[7] H Qiu P Cui A D Regmi et al ldquoInfluence of topographyand volume on mobility of loess slides within different slipsurfacesrdquo Catena vol 157 pp 180ndash188 2017

[8] J Hu J Peng G Wang I Javed Y Wang and W LildquoDistribution and characteristics of landslide in LoessPlateau a case study in Shaanxi provincerdquo EngineeringGeology vol 236 pp 89ndash96 2018

[9] L Xu F C Dai Q M Gong L G am and H MinldquoIrrigation-induced loess flow failure in Heifangtai Platformnorth-west Chinardquo Environmental Earth Sciences vol 66no 6 pp 1707ndash1713 2012

[10] R Q Huang W H Liu J P Zhou and X J Pei ldquoRolling testson movement characteristics of rock blocksrdquo Journal ofGeotechnical Engineering vol 29 no 9 pp 1296ndash1302 2007in Chinese

[11] I Manzella and V Labiouse ldquoPhysical modelling to betterunderstand rock avalanchesrdquo in Physical Modelling in Geo-technics Two Volume Set Proceedings of the 7th InternationalConference on Physical Modelling in Geotechnics (ICPMG2010) vol 2 pp 1259ndash1265 Taylor amp Francis Group LondonUK 2010

[12] G B Crosta F V De Blasio M De Caro G VolpiS Imposimato and D Roddeman ldquoModes of propagationand deposition of granular flows onto an erodible substrateexperimental analytical and numerical studyrdquo Landslidesvol 14 no 1 pp 47ndash68 2017

[13] I Manzella and V Labiouse ldquoEmpirical and analytical ana-lyses of laboratory granular flows to investigate rock ava-lanche propagationrdquo Landslides vol 10 no 1 pp 23ndash362013

[14] Y-X Hu H-B Li S-C Qi G Fan and J-W ZhouldquoGranular effects on depositional processes of debris ava-lanchesrdquo KSCE Journal of Civil Engineering vol 24 no 4pp 1116ndash1127 2020

[15] I Goktepeli U Atmaca and A Cakan ldquoInvestigation of heattransfer augmentation between the ribbed plates via Taguchiapproach and computational fluid dynamicsrdquo Journal ofEermal Science vol 29 no 3 pp 647ndash666 2020

[16] M Farin A Mangeney and O Roche ldquoFundamental changesof granular flow dynamics deposition and erosion processesat high slope angles insights from laboratory experimentsrdquoJournal of Geophysical Research Earth Surface vol 119 no 3pp 504ndash532 2014

[17] M Kesseler V Heller and B Turnbull ldquoA laboratory-nu-merical approach for modelling scale effects in dry granularslidesrdquo Landslides vol 15 no 11 pp 2145ndash2159 2018

[18] A Dufresne and T R Davies ldquoLongitudinal ridges in massmovement depositsrdquo Geomorphology vol 105 no 3-4pp 171ndash181 2009

[19] T De Haas L Braat J R F W Leuven I R Lokhorst andM G Kleinhans ldquoEffects of debris flow composition onrunout depositional mechanisms and deposit morphology inlaboratory experimentsrdquo Journal of Geophysical ResearchEarth Surface vol 120 no 9 pp 1949ndash1972 2015

[20] C Goujon N omas and B Dalloz-Dubrujeaud ldquoMono-disperse dry granular flows on inclined planes role ofroughnessrdquo Ee European Physical Journal E vol 11 no 2pp 147ndash157 2003

[21] X-Y Fan S-J Tian and Y-Y Zhang ldquoMass-front velocity ofdry granular flows influenced by the angle of the slope to therunout plane and particle size gradationrdquo Journal of MountainScience vol 13 no 2 pp 234ndash245 2016

[22] G B Crosta F V D Blasio M Locatelli S Imposimato andD Roddeman ldquoLandslides falling onto a shallow erodible

Advances in Civil Engineering 13

(3) e apparent friction coefficient of flowslide in-creases nonlinearly as the slope angle increases andshows spatial and lithological difference e greaterthe friction coefficient of the sliding material thehigher the initial benchmark of the curve the largerthe coefficient of velocity restitution the larger theabscissa corresponding to the position of inflectionpoint when the curve changes from flat to rapid

Data Availability

e data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

e authors declare that they have no conflicts of interest

Acknowledgments

is study would not have been possible without financialsupport from the Special Fund for the National NaturalScience Foundation of China under Grant nos 4170229841790442 and 41602359 as well as the Project Supported byNatural Science Basic Research Plan in Shaanxi Province ofChina under Grant no 2017JQ4020

Supplementary Materials

In this section the data on the slope angles and the apparentfriction coefficients are provided as a formation of tableese data are related to 44 rock flowslides in theWenchuanSichuan province China 28 soil flowslides in in the JingyangSouth platform Shaanxi province China and 27 chalkflowslides along the coast of northern Europe (Supple-mentary Materials)

References

[1] O Roche M Attali A Mangeney and A Lucas ldquoOn the run-out distance of geophysical gravitational flows insight fromfluidized granular collapse experimentsrdquo Earth PlanetaryScience Letters vol 311 no 3-4 pp 375ndash385 2011

[2] D Guo M Hamada C He Y Wang and Y Zou ldquoAnempirical model for landslide travel distance prediction inWenchuan earthquake areardquo Landslides vol 11 no 2pp 281ndash291 2014

[3] Y Okura H Kitahara A Kawanami and U KurokawaldquoTopography and volume effects on travel distance of surfacefailurerdquo Engineering Geology vol 67 no 3-4 pp 243ndash2542003

[4] Z Han Y Li J Huang et al ldquoNumerical simulation for run-out extent of debris flows using an improved cellular au-tomaton modelrdquo Bulletin of Engineering Geology and theEnvironment vol 76 no 3 pp 961ndash974 2017

[5] F Guzzeti F Ardizzone M Cardinali M Rossi andD Valigi ldquoLandslide volumes and landslide mobilizationrates in Umbria Central Italyrdquo Earth Planetary Science Let-ters vol 279 no 3-4 pp 222ndash229 2009

[6] J D Chaytor S Uri A R Solow and B Andrews ldquoSizedistribution of submarine landslides along the US Atlanticmarginrdquo Marine Geology vol 264 no 1-2 pp 16ndash27 2009

[7] H Qiu P Cui A D Regmi et al ldquoInfluence of topographyand volume on mobility of loess slides within different slipsurfacesrdquo Catena vol 157 pp 180ndash188 2017

[8] J Hu J Peng G Wang I Javed Y Wang and W LildquoDistribution and characteristics of landslide in LoessPlateau a case study in Shaanxi provincerdquo EngineeringGeology vol 236 pp 89ndash96 2018

[9] L Xu F C Dai Q M Gong L G am and H MinldquoIrrigation-induced loess flow failure in Heifangtai Platformnorth-west Chinardquo Environmental Earth Sciences vol 66no 6 pp 1707ndash1713 2012

[10] R Q Huang W H Liu J P Zhou and X J Pei ldquoRolling testson movement characteristics of rock blocksrdquo Journal ofGeotechnical Engineering vol 29 no 9 pp 1296ndash1302 2007in Chinese

[11] I Manzella and V Labiouse ldquoPhysical modelling to betterunderstand rock avalanchesrdquo in Physical Modelling in Geo-technics Two Volume Set Proceedings of the 7th InternationalConference on Physical Modelling in Geotechnics (ICPMG2010) vol 2 pp 1259ndash1265 Taylor amp Francis Group LondonUK 2010

[12] G B Crosta F V De Blasio M De Caro G VolpiS Imposimato and D Roddeman ldquoModes of propagationand deposition of granular flows onto an erodible substrateexperimental analytical and numerical studyrdquo Landslidesvol 14 no 1 pp 47ndash68 2017

[13] I Manzella and V Labiouse ldquoEmpirical and analytical ana-lyses of laboratory granular flows to investigate rock ava-lanche propagationrdquo Landslides vol 10 no 1 pp 23ndash362013

[14] Y-X Hu H-B Li S-C Qi G Fan and J-W ZhouldquoGranular effects on depositional processes of debris ava-lanchesrdquo KSCE Journal of Civil Engineering vol 24 no 4pp 1116ndash1127 2020

[15] I Goktepeli U Atmaca and A Cakan ldquoInvestigation of heattransfer augmentation between the ribbed plates via Taguchiapproach and computational fluid dynamicsrdquo Journal ofEermal Science vol 29 no 3 pp 647ndash666 2020

[16] M Farin A Mangeney and O Roche ldquoFundamental changesof granular flow dynamics deposition and erosion processesat high slope angles insights from laboratory experimentsrdquoJournal of Geophysical Research Earth Surface vol 119 no 3pp 504ndash532 2014

[17] M Kesseler V Heller and B Turnbull ldquoA laboratory-nu-merical approach for modelling scale effects in dry granularslidesrdquo Landslides vol 15 no 11 pp 2145ndash2159 2018

[18] A Dufresne and T R Davies ldquoLongitudinal ridges in massmovement depositsrdquo Geomorphology vol 105 no 3-4pp 171ndash181 2009

[19] T De Haas L Braat J R F W Leuven I R Lokhorst andM G Kleinhans ldquoEffects of debris flow composition onrunout depositional mechanisms and deposit morphology inlaboratory experimentsrdquo Journal of Geophysical ResearchEarth Surface vol 120 no 9 pp 1949ndash1972 2015

[20] C Goujon N omas and B Dalloz-Dubrujeaud ldquoMono-disperse dry granular flows on inclined planes role ofroughnessrdquo Ee European Physical Journal E vol 11 no 2pp 147ndash157 2003

[21] X-Y Fan S-J Tian and Y-Y Zhang ldquoMass-front velocity ofdry granular flows influenced by the angle of the slope to therunout plane and particle size gradationrdquo Journal of MountainScience vol 13 no 2 pp 234ndash245 2016

[22] G B Crosta F V D Blasio M Locatelli S Imposimato andD Roddeman ldquoLandslides falling onto a shallow erodible

Advances in Civil Engineering 13

substrate or water layer an experimental and numericalapproachrdquo IOP Conference Series Earth and EnvironmentalScience vol 26 no 1 Article ID 012004 2015

[23] C Quantin P Allemand and C Delacourt ldquoMorphology andgeometry of valles marineris landslidesrdquo Planetary and SpaceScience vol 52 no 11 pp 1011ndash1022 2004

[24] B K Lucchitta ldquoLandslides in valles marineris marsrdquo Journalof Geophysical Research vol 84 no B14 pp 8097ndash8113 1979

[25] B K Lucchitta ldquoMorphology of chasma walls Marsrdquo Journalof Research of the US Geological Survey vol 6 no 5pp 651ndash662 1978

[26] J J Wu Z W Xie Y L Huang and X M Wu ldquoSensitivityanalysis on stress of flexible hinges to design parameters basedon Taguchi methodrdquo Journal of Mechanical Strength vol 38no 2 pp 252ndash258 2016

[27] S D Kumar P R Vundavilli A Mandal S Mantry andM Chakraborty ldquoErosion response of thixoformed A356-5TiB2 in situ composite using Taguchirsquos experimental designrdquoTribology Transactions vol 60 no 1 pp 39ndash46 2017

[28] M Bagci and H Imrek ldquoApplication of Taguchi method onoptimization of testing parameters for erosion of glass fiberreinforced epoxy composite materialsrdquo Materials amp Designvol 46 pp 706ndash712 2013

[29] M Bagci and H Imrek ldquoErosion wear performance of boraxfilled novel hybrid composites by using the Taguchi experi-mental designrdquo Industrial Lubrication and Tribology vol 68no 1 pp 134ndash140 2016

[30] Z Duan W-C Cheng J-B Peng Q-Y Wang andW ChenldquoInvestigation into the triggering mechanism of loess land-slides in the south Jingyang platform Shaanxi provincerdquoBulletin of Engineering Geology and the Environment vol 78no 7 pp 4919ndash4930 2019

[31] P Ma J Peng Q Wang Z Duan Z Meng and Z JianqildquoLoess landslides on the South Jingyang platform in Shaanxiprovince Chinardquo Quarterly Journal of Engineering Geologyand Hydrogeology vol 52 no 4 pp 547ndash556 2019

[32] P Asteriou H Saroglou and G Tsiambaos ldquoGeotechnicaland kinematic parameters affecting the coefficients of resti-tution for rock fall analysisrdquo International Journal of RockMechanics and Mining Sciences vol 54 pp 103ndash113 2012

[33] R Q Huang and W H Liu ldquoIn-situ test study of charac-teristics of rolling rock blocks based on orthogonal designrdquoChinese Journal of Rock Mechanics Engineering Geologyvol 28 no 5 pp 882ndash891 2009 in Chinese

[34] L-T Zhan Z Zhang Y-M Chen et al ldquoe 2015 Shenzhencatastrophic landslide in a construction waste dump re-constitution of dump structure and failure mechanisms viageotechnical investigationsrdquo Engineering Geology vol 238pp 15ndash26 2018

[35] C S Chen Q Q Li J S Zhu et al ldquoFormation of the 2015Shenzhen landslide as observed by SAR shape-from-shadingrdquoScientific Reports vol 7 no 1 Article ID 43351 2017

[36] J N Hutchinson S G Evans and J V Degraff ldquoChalk flowsfrom the coastal cliffs of Northwest Europerdquo in CatastrophicLandslides vol 15 pp 257ndash302 Geological Society ofAmerica Boulder Colo USA 2002

[37] A Henaff Y Lageat S Costa and E Plessis ldquoLe recul desfalaises crayeuses du Pays de Cauxdetermination des proc-essus drsquoerosion et quantification des rythmes drsquoevolutionRetreat of chalk cliffs in the Pays de Caux processes andratesrdquo Geomorphologie Relief Processus Environnementvol 8 no 2 pp 107ndash118 2002

[38] A Lucas A Mangeney and J P Ampuero ldquoFrictional ve-locity-weakening in landslides on earth and on other

planetary bodiesrdquo Nature Communications vol 5 no 1pp 1ndash9 2014

[39] A N Edwards and J M N T Gray ldquoErosion-depositionwaves in shallow granular free-surface flowsrdquo Journal of FluidMechanics vol 762 pp 35ndash67 2015

[40] E D Fernacuteandez-Nieto J Garres-Dacuteıaz A Mangeney andG Narbona-Reina ldquoA multilayer shallow model for drygranular flows with the micro (I) rheology application to granularcollapse on erodible bedsrdquo Journal of Fluid Mechanicsvol 798 pp 643ndash681 2015

[41] J-J Wang H-P Zhang S-C Tang and Y Liang ldquoEffects ofparticle size distribution on shear strength of accumulationsoilrdquo Journal of Geotechnical and Geoenvironmental Engi-neering vol 139 no 11 pp 1994ndash1997 2013

[42] L J SuW H ZhouW B Chen and X Jie ldquoEffects of relativeroughness and mean particle size on the shear strength ofsand-steel interfacerdquo Measurement vol 122 pp 330ndash3462018

[43] L-P Li S-Q Sun S-C Li Q-Q Zhang C Hu and S-S ShildquoCoefficient of restitution and kinetic energy loss of rockfallimpactsrdquo KSCE Journal of Civil Engineering vol 20 no 6pp 2297ndash2307 2015

[44] K Sharma S S Mallick and A Mittal ldquoA study of energy lossdue to particle to particle and wall collisions during fluidizeddense-phase pneumatic transportrdquo Powder Technologyvol 362 pp 707ndash716 2020

[45] H Qiu A D Regmi P Cui S Hu YWang and Y He ldquoSlopeaspect effects of loess slides and its spatial differentiation indifferent geomorphologic typesrdquo Arabian Journal of Geo-sciences vol 10 no 15 p 344 2017

[46] C M Lo M L Lin C L Tang and J C Hu ldquoA kinematicmodel of the Hsiaolin landslide calibrated to the morphologyof the landslide depositrdquo Engineering Geology vol 123 no 1-2 pp 22ndash39 2011

[47] K Lied and K Bakkehoslashi ldquoEmpirical calculations of snow-avalanche run-out distance based on topographic parame-tersrdquo Journal of Glaciology vol 26 no 94 pp 165ndash177 1980

14 Advances in Civil Engineering