performance-based design and testing methods for granular
TRANSCRIPT
Department of Civil, Construction and Environmental Engineering
Performance-based design and testing methods for granular road surface materials (IHRB Project TR-685)
2017 Mid-Continent Transportation Research SymposiumAugust 16, 2017
Cheng Li, Ph.D. Postdoctoral Research [email protected]
Project Principal Investigators:Drs. Jeramy Ashlock, Bora Cetin, and Charles Jahren
Department of Civil, Construction and Environmental Engineering
60% of IA road network is unpaved: 68,400 of 114,000 miles
2
3
5
6
$14,784 per mile for a 30ft wide road
Reference: Li, C., Ashlock, J. C., White, D. J., & Vennapusa, P. (2015). Low-Cost Rural Surface Alternatives: Demonstration Project., IHRBProject TR-664, Iowa Department of Transportation, Ames, IA, p. 242.
Department of Civil, Construction and Environmental Engineering
Performance and durability of granular road surface materials are a function of several influence factors
• Gradation (particle size distribution and top size)
• Plasticity (dust and stability)
• Aggregate Quality (degradation and abrasion)
• Aggregate morphology (shape and angularity)
7
Gradation
Department of Civil, Construction and Environmental Engineering
Particle packing of aggregate materials governs its mechanical performance, especially in wet conditions
9
Reference: Xiao, Y., E. Tutumluer, Y. Qian, and J. A. Siekmeier. Gradation Effects Influencing Mechanical Properties of Aggregate Base-Granular Subbase Materials in Minnesota. Transportation Research Record: Journal of the Transportation Research Board, 2267, 2012, pp. 14-26.
Coarse-graded Optimum Gradation Floating Aggregates
Department of Civil, Construction and Environmental Engineering
Typical state DOTs’ specifications for gradation and plasticity are not performance related
10
Sieve No. Iowa Class A or B
South Dakota
Illinois CA-6
Minnesota Class 1
Nebraska Rock
Missouri Grade B
1.5 100 1 100 100-90 100 100
3/4 in. 100-95 100 100 1/2 in. 90-70 90-60 3/8 in. 95-65 < 65 No. 4 55-30 78-50 56-30 85-40 60-20 No. 8 40-15 67-37
No. 10 70-25 30-0 25-5 No. 16 40-10 No. 40 35-13 45-10 No. 200 16-6 15-4 12-4 15-8 10-0 Plasticity
Index NA 12-4 9-2 NA NA NA
Department of Civil, Construction and Environmental Engineering
Hypothesis: the target is too wide and does not specify the most critical parameter: particle packing
11
#10
#40
#100
#200#41/2"
5/8"
SandGravel Silt
Grain Diameter (mm)
0.0010.010.1110100
Per
cent
Pas
sing
(%)
0102030405060708090
100Iowa DOT Class A & B
Existing Aggregate (varied gradations)
Existing AggregateVirgin AggregateVirgin with Existing
1"3/
4"
1/4"
#60
#20
3/8"
Clay
1.5"
Department of Civil, Construction and Environmental Engineering
Gravel Content (%)
10 20 30 40 50 60 70 80
Soak
ed C
BR (%
)
0
10
20
30
40
50
60
70
80
ExistingExisting+Virgin
100% Existing
100% Virgin
<#40, % vs Soaked CBR, %
Sand Content (%)
10 20 30 40 50 60
Soak
ed C
BR (%
)
0
10
20
30
40
50
60
70
80
ExistingExisting+Virgin
100% Virgin
100% Existing
Fines Content (%)
0 5 10 15 20 25 30
Soak
ed C
BR (%
)
0
10
20
30
40
50
60
70
80
ExistingExisting+Virgin
100% Existing
100% Virgin
An optimum gradation in terms of shear strength exists for a given well-graded granular material
12
Department of Civil, Construction and Environmental Engineering
Grain Diameter (mm)
0 5 10 15 20 25 30
Per
cent
Pas
sing
(%)
0102030405060708090
100
Gradations of well-graded granular materials can be described by two parameters using Fuller’s model
13
𝑝𝑝𝑖𝑖 = 𝐷𝐷𝑖𝑖𝐷𝐷𝑚𝑚𝑚𝑚𝑚𝑚
𝑛𝑛× 100
where subscript “𝑖𝑖” represents a particular sieve; 𝑝𝑝𝑖𝑖 = percentage passing the 𝑖𝑖 sieve;𝐷𝐷𝑖𝑖 = opening size of the 𝑖𝑖 sieve; 𝐷𝐷𝑚𝑚𝑚𝑚𝑚𝑚 = the maximum size of aggregate; and 𝑛𝑛 = shape factor of gradation curve.
Existing AggregateFuller's Model
pi = (Di / Dmax) n
R2 = 0.9803Dmax = 23.6 mmn = 0.2794G/S = 0.67
Department of Civil, Construction and Environmental Engineering
Fuller’s model with two parameters can be used to develop performance-based specifications
14
Department of Civil, Construction and Environmental Engineering
PSD curves with a wide range of shape factors can meet the current gradation specification band
15
#10
#40
#100
#200#41/2"
SandGravel Fines
Particle Size (mm)
0.010.1110100
Perc
ent
Pass
ing
(%)
0102030405060708090
100
Iowa DOT Spec. Band1"
3/4"
#60
#20
3/8"
n = 0.10
n = 0.15
n = 0.70
Department of Civil, Construction and Environmental Engineering
The current gradation specifications cannot ensure performance
16
Plasticity
Department of Civil, Construction and Environmental Engineering
Bentonite treatment significantly reduced dust and provided a much drier surface during thawing
#1A Dirty MacadamA-1-a (USCS: GP-GM)Percent of fines = 9.9%Percent of clay = 1.8%PI = NP
#1B Dirty Macadam+BentoniteA-2-6(0) (USCS: GC)Percent of fines = 14.3%Percent of clay = 5.0%LL=30, PL=12, and PI=18
#2 Dirty Macadam+Chloride
18
March 28, 2015
Reference: Li, C., P. K. R. Vennapusa, J. Ashlock, and D. J. White. (2017) Mechanistic-Based Comparisons for Freeze-Thaw Performance of Stabilized Unpaved Roads. Cold Regions Science and Technology, Vol. 141, 2017, pp. 97-108.
Department of Civil, Construction and Environmental Engineering
Bentonite was found to be effective to improve the stability and slaking of crushed limestone fines
19
Department of Civil, Construction and Environmental Engineering
Add 4% bentonite to the surface material passing #40 sieve or control the PI between 7 and 15
20
Moisture Content (%)
0 3 6 9 12 15 18
UC
S (M
Pa)
0.01
0.1
1
10
100
0% Bentonite2% Bentonite4% Bentonite6% Bentonite
Aver
age
Slak
ing
Tim
e (m
in.)
10
100
1000
OMC+2%After Drying
0% Bentonite
2% Bentonite
4% Bentonite
6% Bentonite
Department of Civil, Construction and Environmental Engineering
Different lab tests for measuring plasticity of soils were evaluated
21
Conventional Atterberg Limit tests
Fall Cone Liquid and Plastic Limit Test
Bar Linear Shrinkage(BLS) Test
Cas
agra
nde
Cup
AS
TM R
olle
r
Department of Civil, Construction and Environmental Engineering
South Africa DOT’s performance-related unpaved road surface material selection chart
22
(% passing1in . sieve % passing #10 sieve) % passing #4 sieveGrading Coefficient100
− ×=
Shrinkage Product Bar Linear Shrinkage % passing #40 sieve= ×
Reference: Paige-Green, P. The Influence of Geotechnical Properties on the Performance of Gravel Wearing Course Materials. Ph.D. Dissertation, University of Pretoria, South Africa, 1989.
Department of Civil, Construction and Environmental Engineering
The BLS can be correlated to the PI but is not a sensitive parameter indicating plasticity of soils
23
Bar Linear Shrinkage (%)
0 5 10 15 20 25
Pla
stic
Inde
x
0
5
10
15
20
25
30
35
40
Averagey = 4.84x-5.27; R2 = 0.62n = 36 (All data points)95% Prediction Interval3%
6%
9%
12% Bentonite
Note: Error bars show the maximum or minimum values.
Department of Civil, Construction and Environmental Engineering
The fall cone test is easier to learn and perform and has better repeatability than the conventional liquid limit test
Fall Cone Liquid Limit (%)
10 15 20 25 30 35 40 45 50
Cas
agra
nde
Cup
Liq
uid
Lim
it (%
)
10
15
20
25
30
35
40
45
50Averagey = 1.17x-5.82; R2 = 0.98n= 45 (All data points)95% Prediction Interval 1:1
Line
Note: Error bars show the maximum or minimum values.
0%
3%
6%
9%12%
Bentonite
24
Department of Civil, Construction and Environmental Engineering
The repeatability, reproducibility, and overall variability in the tests were statistically evaluated
Two-Way ANOVA-Based repeatability and reproducibility (R&R) Analysis𝑦𝑦𝑖𝑖𝑖𝑖𝑖𝑖 = 𝜇𝜇 + 𝛼𝛼𝑖𝑖 + 𝛽𝛽𝑖𝑖 + 𝛼𝛼𝛽𝛽𝑖𝑖𝑖𝑖 + 𝜀𝜀𝑖𝑖𝑖𝑖𝑖𝑖
where,𝜇𝜇 = an average measurement of all possible operators and all possible
parts,𝛼𝛼𝑖𝑖 = effects of different parts,𝛽𝛽𝑖𝑖 = effects of different operators,𝛼𝛼𝛽𝛽𝑖𝑖𝑖𝑖 = joint effects peculiar to particular part/operator combinations, and𝜀𝜀𝑖𝑖𝑖𝑖𝑖𝑖 = measurement error.
Reference: Vardeman, S. B., and Jobe, J. M., 1999, Statistical quality assurance methods for engineers, John Wiley, New York.
25
Department of Civil, Construction and Environmental Engineering
The overall errors caused by the test method itself and operators were statistically evaluated
Test Name No. of Samples
No. of Operators
No. of Repeated Tests
Casagrande Cup LL
5a 3b 3Fall Cone LLASTM Roller PLFall Cone PLBar Linear Shrinkage (BLS)
26
a An unpaved road surface material with 0%, 3%, 6%, 9%, and 12% bentoniteb The three operators were trained at the same time on all the five tests
The fall cone and ASTM roller devices yielded better repeatability and reproducibility than conventional tests
27
ParametersLiquid Limit Plastic Limit
CasagrandeCup
Fall Cone Test
ASTMRoller
Fall Cone Test
Repeatability 0.6% 0.5% 0.4%
NoCorrelation
Reproducibility 1.7% 0.5% 0.6%
Overall R&R 1.8% 0.7% 0.7%
% of overall R&Rdue to reproducibility 89% 50% 73%
Department of Civil, Construction and Environmental Engineering
Recommended lab tests for determining liquid and plastic limits of soils
Liquid limit Plastic Limit
28
Aggregate Quality and Morphology
Reference: Li, C., Ashlock, J., White, D., Jahren, C., and Cetin, B. (2017). "Gyratory Abrasion with 2D Image Analysis Test Method for Evaluation of Mechanical Degradation and Changes in Morphology and Shear Strength of Compacted Granular Materials." Construction and Building Materials, 152, 547—557, https://doi.org/10.1016/j.conbuildmat.2017.07.013
Department of Civil, Construction and Environmental Engineering
State DOTs and most researchers use LA abrasion tests to evaluate the quality of granular material
30
From https://www.youtube.com/watch?v=FUGUk-4wEyA
(From ASTM C131)
#12 sieve is used to separate the specimen
Department of Civil, Construction and Environmental Engineering
Five different materials types from three different sources yielded similar LA abrasion test results
31
Parameters Existing Surface Aggregate
Virgin Surface Aggregate
Road Rock
Class A Stone
Concrete Stone
Abbreviation ESA VSA RR CAS CSSource Granular road Quarry 1 Quarry 1 Quarry 2 Quarry 2
Gravel content (%) 24.0 68.7 65.2 42.9 96.3Sand content (%) 50.0 22.8 19.5 48.9 2.9Fines content (%) 26.0 8.5 15.3 8.2 0.8
Maximum aggregate size (mm) 25.4 38.1 38.1 25.4 25.4USCS symbol SM GP-GC GM SP-SM GP
LA abrasion loss 22.6 24.7 33.9 27.3 26.9
Sieve analysis and LA abrasion test results of the five granular materials
Department of Civil, Construction and Environmental Engineering
LA abrasion results could be governed by the testing mechanism instead of material properties
Particle Size (mm)
0.010.1110100
Per
cent
Pas
sing
(%)
0102030405060708090
100
ESA Specimen 1 ESA Specimen 2VSA Specimen 1VSA Specimen 2CS CAS RR (Grading A)
38.1
mm
25.4
mm
19.1
mm
12.7
mm
9.53
mm
0.07
5 m
m
SandGravel
0.15
mm
0.25
mm
0.43
mm
0.85
mm
4.75
mm
Fines
Grading B
1.7
mm
Grading A
32
#12
Department of Civil, Construction and Environmental Engineering
The gyratory compactor can better simulate actual field compaction and traffic loading conditions
33
Specimen
Rigid Mold
PDARam
1.25º
Bottom plate
Top plate
150 mm
170
mm
Resultant load
ey
ex
P3P1
P2
O
120°
x
yLoad cellGyratory Compactor
PDA
(a) (b)
(c)
Reference: Li, C., White, D., and Vennapusa, P. (2015). "Moisture-Density-Strength-Energy Relationships for Gyratory Compacted Geomaterials." Geotechnical Testing Journal, 38(4), 461—473, https://doi.org/10.1520/GTJ20140159
Department of Civil, Construction and Environmental Engineering
Using a single arbitrary sieve size to quantify the degradation of a material can be misleading
34
Particle Size (mm)
110100
Per
cent
Pas
sing
(%)
0102030405060708090
100Initial PSDFinal PSD
38.1
mm
25.4
mm
19.1
mm
12.7
mm
9.53
mm
4.75
mm
SandGravel
Road RockGravel = 65.2%Sand = 19.5%Fines = 15.3%
Difference in % passingat 4.75mm sieve = 1.8%
Hardin's Total BreakageBt=0.12
Department of Civil, Construction and Environmental Engineering
Changes in gradation and morphology of specimens can be determined using 2D Image analysis
35
(a) (b) (c)
Department of Civil, Construction and Environmental Engineering
Influences of aggregate morphology on performance of aggregate materials also need to be quantified
36
Reference: J. Zheng, R.D. Hryciw, (2015) Traditional soil particle sphericity roundness and surface roughness by computational geometry, Géotechnique 65 (6) 494–506.
Department of Civil, Construction and Environmental Engineering
The test results demonstrate the importance of testing full gradation for evaluating material’s degradation
37
38.1
mm
25.4
mm
19.1
mm
12.7
mm
9.53
mm
4.75
mm
SandGravel
n =
16n
= 11
n =
118
n =
78n
= 49
6n
= 36
2n
= 69
4n
= 51
0
n =
1342
n =
1847
(a)
(b)
Final n = 2808
Per
cent
Pas
sing
or R
etai
ned
0102030405060708090
100
Final % RetainedInitial % Retained
Initial PSDFinal PSD
Concrete StoneGravel = 96.3%Sand = 2.9%Fines = 0.8%
5th
25th
Median75th
95th
10th
90th
Initia No. of Aggregatesn = 2666
Bt=0.12
Particle Size (mm)
110100
Sph
eric
ity
0.10.20.30.40.50.60.70.80.9
Initial SphericityFinal Sphericity
Particle Size (mm)
110100
Sph
eric
ity
0.10.20.30.40.50.60.70.80.9
Initial SphericityFinal Sphericity
38.1
mm
25.4
mm
19.1
mm
12.7
mm
9.53
mm
4.75
mm
SandGravel
n =
1n
= 1
n =
8n
= 7
n =
77n
= 76
n =
139
n =
142
n =
821
n =
817
(a)
(b)
5th
25th
Median
75th
95th
10th
90th
Per
cent
Pas
sing
or R
etai
ned
0102030405060708090
100
Final % RetainedInitial % Retained
Initial PSDFinal PSD
Post-n = 1043
Existing Surface AggregateGravel = 24.0%
Sand = 50.0%Fines = 26.0%
Pre-number of particles n = 1046
Bt = 0.001
LA Abrasion: 26.9 LA Abrasion: 22.6
Department of Civil, Construction and Environmental Engineering
Using the image analysis can rapidly determine the particle size distribution of aggregate materials
38
Particle Size (mm)
110100
Per
cent
Pas
sing
0
10
20
30
40
50
60
70
80
90
100Sieve AnalysisBounding RectangleBest Fit EllipseFeret Diameter
38.1
mm
25.4
mm
19.1
mm
12.7
mm
9.53
mm
4.75
mm
SandGravel
Size (mm) CAS CS ESA VSA RR25.4 -0.2 -2.9 1.5 -6.4 5.019.1 -3.3 -4.8 4.7 2.1 -2.312.7 3.3 4.0 6.3 14.0 -11.59.53 4.7 4.0 9.4 9.1 -9.2
% passing from Feret Dia. - % passing from Sieve Analysis
Performances of Field Test Sections
A Microsoft Excel-based gradation optimization program was developed to recycle existing materials
40
District Project Date 5/19/2016County Note Designer Cheng Li
500 ft30 ft
Thickness of Existing Aggregate 2.30 in.Target Final Thickness 6.00 in.Compacted Dry Unit Weight 130 pcfMaximum Aggregate Size (Dmax) 1.00 in.PSD Shape Factor (n ) 0.40
Sieve No.
Sieve size (mm)
Target Gradation (%)
Existing Material
Gradation (%)
Target Virgin Material
Gradation (%)
Quarry Material A
Quarry Material B
Quarry Material C
Optimized Virgin Material Gradation (%)
Actual Final Gradation
(%)2 50.8 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
1.5 38.1 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.01 25.4 100.0 100.0 100.0 60.0 95.0 100.0 95.9 97.5
3/4 19.00 89.0 97.5 85.9 23.0 84.0 100.0 92.2 94.21/2 12.70 75.8 89.8 71.4 6.3 70.0 83.0 75.2 80.83/8 9.51 67.5 81.8 64.1 6.0 60.0 71.0 64.4 71.1#4 4.76 51.2 66.6 48.8 5.6 37.0 45.0 41.0 50.8#8 2.38 38.8 55.9 36.1 5.1 21.0 27.0 24.8 36.7
#16 1.19 29.4 48.4 25.5 4.6 14.0 20.0 18.4 29.9#30 0.595 22.3 41.9 17.5 4.1 11.0 16.0 14.8 25.2#50 0.297 16.9 35.7 12.0 3.4 8.4 14.0 12.9 21.7
#100 0.149 12.8 29.3 8.6 2.9 6.8 12.0 11.1 18.1#200 0.075 9.7 25.2 5.4 2.3 5.6 10.2 9.4 15.5
Proportion (%) 100 10 0 90 100.0000023Quantity (tons) 300.6 30.5 0.0 270.1
The optimal n value can be selected using the chart developed based on lab CBR test results in the IHRB Project TR-685.
Gradation Optimization for Unpaved Road Surface Materials
Pottawattamie
Average Road Width
TR-685 Test Section 2
Road LengthThis is a trial version of the program
0102030405060708090
100
0.01 0.1 1 10 100
Per
cent
Fin
er (%
)
Sieve Sizes (mm)
Existing GradationTarget Final GradationOptimized Virgin GradationActual Final Gradation
1.5 1
in3/41/23/8#4#8
#16
#30
#50
#100
#200
2 in
RUNINSTALL SOLVER
Department of Civil, Construction and Environmental Engineering
Field test sections constructed on County Road L66 in Pottawattamie County
41
3 in. Existing Aggregate4 in. Virgin Aggregate 1 +
Existing Aggregate
Subgrade (Average DCP-CBR = 14%)USCS: CL and AASHTO: A-7-6(26)
2 in. 3% Bentonite Treated Mixture
500 ftBentonite Treated
4 in. Virgin Aggregate 1
6 in. Existing Aggregate7 in. Virgin Aggregate 2 +
Existing Aggregate
500 ftOptimal Gradation
1000 ftOff-Spec Gradation
500 ft Existing Gradation
Department of Civil, Construction and Environmental Engineering
The as-built gradation of Section 2 is very close to the target optimal gradation (Top size=1in., n=0.4)
42
#10
#40
#100
#200#41/2"
5/8"
SandGravel Silt
Particle Size (mm)
0.0010.010.1110100
Per
cent
Pas
sing
(%)
0102030405060708090
100Iowa DOT Specification
Section 1 (n=0.36, Dmax=25.6mm)Section 2 (n=0.39, Dmax=24.2mm)Section 3 (n=0.30, Dmax=24.2mm)Section 4 (n=0.16, Dmax=22.8mm)
Target (n=0.40, Dmax=25.4mm)
1"3/
4"
1/4"
#60
#20
3/8"
Clay
1.5"
Department of Civil, Construction and Environmental Engineering
Field tests were conducted after construction and one seasonal freeze-thaw period on test sections
• DCP tests for measuring shear strength
• LWD tests for measuring composite stiffness
• Dustometer tests for measuring fugitive dust emissions
43
Department of Civil, Construction and Environmental Engineering
The trends of the in situ DCP test results generally agree with the predictions of the statistical model.
44
DC
P-C
BRA
GG (%
)
0
100
200
300
400Pr
edic
ted
Soak
ed C
BR (%
)
0
20
40
60
80
100
In Situ CBR (7/25/2016)Predicted Soaked CBR
In Situ CBR (5/5/2017)
S1 S2 S3 S4
BentoniteTreated
Optimal Gradation
Off-Spec Gradation
Existing Gradation
Department of Civil, Construction and Environmental Engineering
The stiffness of all sections decreased noticeably except for the section with optimal gradation
45
ELW
D (M
Pa)
0
20
40
60
80
100
120ELWD (7/25/2016)ELWD (5/5/2017)
BentoniteTreated
Optimal Gradation
Off-Spec Gradation
Existing Gradation
Department of Civil, Construction and Environmental Engineering
The bentonite and optimal gradation sections yielded much less dust than the control section
46
7/25/2016 5/5/2017Bentonite Treated 0.9 0.9Optimal Gradation 0.9 1.0Off-Spec Gradation 1.2 1.5Existing Gradation 3.2 4.2
Section NameFugitive Dust Emission (gram per 1000 ft)
Department of Civil, Construction and Environmental Engineering
Grading Coefficient
Shr
inka
ge P
rodu
ctBentonite (As-constructed)Roller Compacted (As-constructed)
Blended (As-constructed)Control
Good but dusty
Good
Slippery and dusty
Washboards and ravels
Erodible Ravels
15 350
0
100
365
250
Predicted performances of the test sections according to the South Africa performance chart
47
Increasing Coarseness/Gap
Incr
easi
ng P
last
icity
Bentonite-Treated SectionOptimal Gradation Section
Off-Spec Gradation SectionExisting Gradation Section
Department of Civil, Construction and Environmental Engineering
Our goal is to help to build granular roads with good performance, long service life, and low costs
• A performance-based design method was developed for granular road surface materials
• An MS Excel gradation optimization program was developed to recycle existing surface materials
• Lab test methods for determining plasticity of soils and evaluating quality of aggregate materials were evaluated
48
Department of Civil, Construction and Environmental Engineering
We welcome your
• Questions
• Suggestions
• Ideas
• Experience
49
Thank you!