advances in mechanical engineering 2016, vol. 8(9) 1–12 an … · 2016-11-08 · research article...

Research Article

Advances in Mechanical Engineering2016, Vol. 8(9) 1–12� The Author(s) 2016DOI: 10.1177/1687814016669635aime.sagepub.com

An empirical methodology forprediction of shape and flow rate of afree-falling non-submerged liquid andcasting iron stream

PJ Gamez-Montero1, R Castilla1, J Freire2, M Khamashta2 and E Codina1

AbstractThe work presented in this article demonstrates the use of an empirical and simplified approach based on an opticaltechnique and a home-made ad hoc code that give knowledge of the shape and falling velocity of a free-falling non-sub-merged liquid stream to predict its typology and flow rate. The visualization photographic technique is a non-intrusiverobust technique which can be applied to high-temperature liquids in harsh environments, such as an iron stream in afoundry. This technique allows predicting the liquid stream boundaries and contours without any type of treatment onthe fluid. As a result of employing this empirical methodology, three flow typologies for a water stream are proposedand demonstrated experimentally. Comparisons with experimental data reveal satisfactory estimations of mean flowquantities. Finally, the approach used based on experimental visualization is carried out in an iron stream of a foundry,not being disruptive to in-situ foundry operations and showing its potential to improve performance of the cast parts’properties during the casting phase and revealing to be a useful tool for process optimization.

KeywordsCast iron, liquid stream, stream shape, air entrainment, flow rate

Date received: 12 June 2016; accepted: 23 August 2016

Academic Editor: Shan-Tung Tu

Introduction

The cast iron foundry industry produces cast parts forautomotive, agriculture, transportation, energy, aero-space, and manufacturing industry. All foundry pro-cesses generate a certain level of rejection, lostirrecoverably, that is closely related to the type of cast-ing, the processes used, and the equipment available.As the quality demands from end users of castingsincrease, it is essential that cast iron technology movesforward together with green manufacturing as a firststep toward sustainability.

Extensively, work has been carried out by foundrycommunity to minimize the casting defects, such as por-osity, slag, and clogging, but only few literatures areconcerned with the casting of the metal itself which isrelated to a more versatile melt control technology.

The flow control system in a tundish has a signifi-cant impact on the quality and level of rejection on thepouring of the castings.1 If the metal does not flow in aconsistent stream, it could result in casting defects fromoxidation, air entrapment, and erosion of the pouringmold, among others.

1LABSON, Department of Fluid Mechanics, Universitat Politecnica de

Catalunya, Terrassa, Spain2LABSON, Department of Mechanical Engineering, Universitat

Politecnica de Catalunya, Terrassa, Spain

Corresponding author:

PJ Gamez-Montero, LABSON, Department of Fluid Mechanics,

Universitat Politecnica de Catalunya, Campus Terrassa, Colom 11, 08222

Terrassa, Spain.

Email: [email protected]

Creative Commons CC-BY: This article is distributed under the terms of the Creative Commons Attribution 3.0 License

(http://www.creativecommons.org/licenses/by/3.0/) which permits any use, reproduction and distribution of the work without

further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/

open-access-at-sage).

by guest on October 8, 2016ade.sagepub.comDownloaded from

http://ade.sagepub.com/

Much of the understanding of the relationshipamong stream shape, flow conditions, and casting prop-erties relates to human visual control in the pouring ofthe casting. A main goal of this work is to achieve a bet-ter knowledge of the stream shape and flow conditionsthat will improve the integrity and consistency of themetal flow into the mold.

The physics of liquid jets in applications or academicsituations such as the jet breakup, critical limit, jet dis-persion, fragment sizes, and drop shapes is of greatinterest from fundamental to technological aspects.2,3

A quite extensive literature to review regarding jets’characteristics to processes such as atomized sprays, jetcutting, jet cleaning, and fuel injectors, among others,can be found. Concentrating exclusively on verticalliquid jets, these studies mainly deal with the plungingjet flow patterns, entrained air bubble sizes, and theinfluence of the jet velocity and variations of jet fallinglengths on the jet penetration depth. Contrariwise, acast iron stream differs significantly from those studies.

Jeffrey et al.4 presented a method to predict the velo-city decay of a free-falling jet into a plunge pool. Amodel of a plunging jet was constructed to simulate afree-falling fully air entrained developed jet and a non-aerated undeveloped jet. The velocity decay data werecollected by means of a pitot tube and the air entrainedby means of an air concentration probe. Despite beinga free-falling jet, it was rectangular in shape and intru-sive techniques were used. Therefore, it must be dis-carded in a cast iron stream.

Non-invasive visualization techniques are widelyused for the examination of the flow fields, such as par-ticle image velocimetry (PIV) and time-resolved particleimage velocimetry (TR-PIV), which introduce tracers,and they are chosen in such a way that their movementrepresents the movement of the fluid.5,6 Melling7

demonstrated that proper flow seeding is particularlycritical with PIV. Experimental investigations were per-formed to investigate the flow-field structure on a sub-merged water jet impinging normally on a smoothquiescent and flat surface and the geometry in a sub-merged entry nozzle in a water model.8–10 Jiang et al.11

presented the advanced applications of PIV techniqueto measure the flow field of different axial planes, axialtime-average velocity, and the length of the initial sec-tions of horizontal low-pressure jet flows emitted bythe sprinkler with different nozzle geometric para-meters and working pressures. A sophisticated multi-grid cross-correlation digital particle image velocimetry(MCCD-PIV) is used to investigate the stability andstructure of low Reynolds number axisymmetric jets toa qualitative examination of the development of thejet.12 Gong et al.13 conducted a visualization method tocapture the micro-surface wave in a high-speed verticaljet characterizing diverse wave patterns. However, thesestudies are not concerned with non-submerged driven

by gravity jet in the atmosphere and even less withregard to high-temperature cast iron stream in real andpractical working conditions in a commercial foundry.

Moreover, the air entrainment within the liquidstream limits the optical methods such as PIV. One nat-ural limitation is that the presence of phase boundariesin multiphase flow blocks the optical path to the mea-surement location.14 Subsequently, to overcome thisphenomenon, visual techniques can be used. A com-bined experimental and numerical investigation is car-ried out by Qu et al.15 Bubble sizes were estimatedfrom single images by means of image processingthrough subsequent background subtraction, cell seg-mentation, bubble detection, and bubble size calcula-tion. A similar technique is carried out to visualize therandom motion of a single bubble and to obtainempirical relationship between penetration depth andair entrainment rate.16,17

Conversely, a cast iron stream differs significantlyfrom the studies presented above. As far as authors’knowledge, there is still a lack of studies concerning theshape prediction and flow rate estimation of a free-falling non-submerged vertical casting iron stream dri-ven by gravity in the atmosphere from the tundishnozzle exit to the mold. Summing up, the tracer parti-cles in the PIV technique cannot be used in a cast ironsteam at 1450�C and also applies to intrusive tech-niques. As a matter of fact, to overcome the maindrawbacks of cast iron, it is not necessary to obtain avery precise quantitative description of the liquid velo-city vectors in an iron stream.18 Furthermore, thedimensions of the nozzle exit and the mean diameter ofthe liquid stream differ considerably from the liquid jetapplications reported, and the influence of the airentrainment into the stream cannot be neglected.

Due to that, the keystone of this work is that anempirical and simplified approach based on an opticaltechnique gives knowledge of the shape and fallingvelocity of a vertical liquid stream that will be very use-ful to predict reasonably its flow condition and flowrate. As a bottom line, this methodology has a greatfield of application and can be implemented in a realcast iron stream on a current foundry. The approachused based on experimental visualization and home-made ad hoc code could lead to an improving perfor-mance of the cast parts’ properties during the castingphase and can be considered as a useful tool for processoptimization.

Materials and methods

There are significant differences between water and castiron; as fluid properties, the most notable differencesare density and surface tension, and as operating condi-tions, fluid temperature. However, water and cast iron

2 Advances in Mechanical Engineering



have similar kinematic viscosity that together with thedifficulty of real industrial condition experiments haveled dynamic water modeling in partial-to-full scalemodels a universally accepted method for observingand evaluating flow conditions of molten metals achiev-ing notable results and quality improvements.

Using a water model simulates experimental investi-gations on a cast iron stream driven by gravity in theatmosphere from the nozzle exit of the tundish. Theexperimental apparatus consists of two independentunits, the fluid cycle and the visualization system.

Experimental setup

The fluid cycle consists in a full-scale foundry pour box(1, see Figure 1) constructed using poly(methyl metha-crylate) (PMMA). A centrifugal pump (2) feeds thepour box through a filter (3) and flow rates are adjusta-ble by a valve (4) mounted on the pipe (5) between thepump and the pour box. It can also maintain a con-stant water height while casting by means of an over-flow pipe (6). The pour box is fitted with a replicastopper rod (7) and slender nozzle (8), also fabricatedof PMMA. The stopper rod can be closed on the topwith a rubber plug (9) avoiding the air to enter throughits interior, which allows simulating non-ideal castingconditions afflicted with typical slag build-up in thestopper rod plugging it. The nozzle exit of radiusR0=13mm (at z=0, see Figure 3) leads the liquidstream (10) into the static air with a maximum fallingdistance of 750mm, where gravitational forces drivethe stream to a reservoir (11) with negligible influenceof air friction and air movement.

The main dimensions of the model are as follows:the tank of 500 3 500 3 400mm3 and the pour boxwith a prismatic shape (1) of 300 3 250 3 250mm3

and [250 3 250mm2 being the exit part (12) where thenozzle is located. Between them a wire screen flow tran-quilizer (13) is placed to breakup and straight the flow.The maximum capacity is 150L of water, for which thetemperature is measured. A fully automated-operatedlift-actuation mechanism (14) is used to seat and unseatthe stopper rod in the nozzle to simulate openings andshutoffs and also establish the constant stopper rod ele-vation h during the casting. The experimental flow rateis calculated by means of filling a volume on a precisionscale and continuously recording the fluid weight exit-ing from the nozzle.

The stream shape formation was observed experi-mentally by the visualization system consisting on ahigh-speed video camera Photron Ultima APX-RS (15)with 17.5mmpixels in a 1024 3 1024 pixels sensor ableto run 3000 fps at full resolution and 250,000 fps at lim-ited resolution. The images were obtained with maxi-mum resolution, that is, 1024 3 1024 pixels at 500 and1000Hz with a 105mm objective. The buffer memoryof the camera allows recording up to 2048 images perexperiment, with duration of slightly more than 4 and2 s, respectively.

With regard to illumination, a modification of thebacklighting is applied: the sidelit. The background is ablack opaque sheet (16) and a directional light from theside through a white styrene sheeting sheet (17) createsa diffuse and clean light on the stream. The sidelitarranges the light in such a way that it leaves the dark-ness behind the stream so that very well illuminatededges and area are achieved. The collected digitalimages are processed and analyzed on a computer (18).

Test fluid

The test fluid is ordinary tap water at temperature 20�Cwith density r = 998kgm23, viscosity m = 1mPa s,and surface tension s = 73mNm21; no other sub-stance is added to the water.

Experimental methodology

The height from liquid free surface to nozzle exitremains constant during all the experiments at thevalue of H0=230mm (see Figure 1). Then, maximumtheoretical velocity at the nozzle exit is U0 = 2.1m s21

(see Figure 3), achieving a recording of slightly morethan 4 s of the vertical liquid stream at Nfps=500 fpswith maximum resolution.

A millimeter ruler was placed in the plane of thestream for calibrating the distance in the images obtain-ing a relation pixel to millimeter of Npxmm=14px/mm.The estimated maximum vertical displacement of thestream is zmax=73mm (see Figure 1) with an esti-mated average time to go out of images of 0.0365 s

Figure 1. Sketch of the experimental setup.

Gamez-Montero et al. 3



= 18 frames, considering it an acceptable precisionbased on the maximum theoretical velocity and testsperformed.

Visualization photographic technique

From the grey-scale digital images in TIFF format (0:blackO 255: white), the stream contour and edgesinformation can be extracted by the use of a steepestimage-processing grey-scale gradient methodology thatauthors have named visualization photographic tech-nique (VPT). The stream edges can be determinedwithin an uncertainty of about 1 pixel at a total contourof 700pixels.

The VPT methodology is described as follows (seeFigure 2):

1. The sidelit provides a diffuse and clean light onthe stream which is digital recorded in a frameimage.

2. The TIFF digital image is read, cropped out,and masked the areas out of interest.

3. A histogram of intensity values of the imageshows the intensity levels that allow setting upthe threshold for the segmentation.

4. The resulting image is dilated and eroded with adisk-shaped filter in order to define edgesboundaries of the stream.

5. The detected boundaries must be selected anddivide in two parts, left and right hand sidecontours.

6. Pixel coordinates are computed to calculate thestream width in all its falling height and coordi-nates are transformed from image space (pixelvalues) to space coordinates.

7. Statistical evaluation, plots, and videos are pro-cessed for all the images of the correspondingexperiment.

Jet dynamics

When a liquid pours vertically from an outlet into sta-tionary air, a free-falling non-submerged vertical liquidjet is formed, initially of constant radius that acceler-ates, stretches, and narrows under the influence ofgravity. The stability of the jet is mainly subjected toeffects of gravity, surface tension, viscosity, and inertialeffects.

Plateau19 first characterized this stability in 1873through experimental observation, and he noted theinstability arose when the liquid column lengthexceeded the column diameter by a factor of about3.13, which was later corroborated by Lord Rayleigh.20

Jet shape function and stability

The theoretical characterization of a jet shape has beenof significant interest of study by many researchersowing to its application on engineering practice. In thisarticle, based on Massalha and Digilov’s21 work, thetheoretical problem of a free-falling non-submergedvertical liquid stream as isothermal, laminar and incom-pressible Newtonian fluid of density r, viscosity m, andsurface tension s injected vertically downward (follow-ing the direction of gravity, g) into stationary air withmean velocity U0 from a circular nozzle orifice of radiusR0 is addressed (see Figure 3).

Here, the effects of the gravity, the surface tension,and the viscosity in comparison with inertia are repre-sented by the Froude number (Fr), the Weber number(We), and the Reynolds number (Re) as

Fr=U2

0

2R0g, We=

2R0rU20

s, Re=

2R0rU0

mð1Þ

The Bond number (Bo), defined asBo=We=Fr= 4R2

0rg=s, can also be used to charac-terize the relative effects of gravity with respect to

Figure 2. VPT procedure on a free-falling water stream.




surface tension. Then, the jet shape function as a rela-tionship between the jet radius r and the axial distancez from the exit of nozzle orifice is obtained as

~z=aFr1

~r4� 1

� �+

12

Bo

1

~r� 1

� �ð2Þ

where ~z= z=R0 and ~r = r=R0 are the reduced jet lengthand jet radius, respectively, and a is a correction para-meter of the velocity profile to be determined fromexperiment (1�a� 2).

At large Bo, when the surface tension effect becomesnegligible compared to the gravity, the gravity and theviscous resistance govern the jet. Moreover, the viscos-ity relative to gravity becomes negligible toRe=Fr� 32d, where d is a correction factor of the vis-cous losses to be determined from experiment (d\1),and then the jet shape function reduces to theWeisbach equation

~z=aFr1

~r4� 1

� �ð3Þ

Together with the jet shape, the stability of theliquid jets plays an important role. Viscous jets, existingslowly from a circular nozzle, are dominated by viscous

forces upstream and by inertial forces downstream.Furthermore, the source of instability can also beattributed to surface tension, flow rate, and nozzle dia-meter and configuration. With regard to the transitionfrom jetting to dripping, Le Dizes22 showed that if thebasic falling capillary jet in the limit has approximatelyan axisymmetric plug profile and Re� 1 and Fr� 1,it becomes locally, absolutely unstable at the orifice fora critical value of We’ 6:25. Clanet and Lasheras23

demonstrated that the maximum nozzle radius is lim-ited to 4mm for water, because above this limit airenters the tube and changes the problem.

Jet shape description

In the liquid stream studied in this work, nozzle dia-meter, stream radius, height, and velocity are quite farfrom the jets encountered in the previous works. Thedata of our configuration are used to calculate the criti-cal parameters with experiments with slow velocitiesunder 1m s21.24 Moreover, the stopper can be pluggedor unplugged, allowing entering air into the stream.

The shape and stability of liquid jets of great butfinite radius and length drawn down by gravity remainsunmentioned in the literature on jet dynamics. Here,the dimensionless group numbers computed from equa-tion (1) are Fr=17.7, We=1604, Re=55121,Bo=90.7, and Re=Fr= 3114� 32. Subsequently,inertial forces downstream dominates the stream and,in this case, equation (3) is reduced to an approxima-tion of the stream shape function. Likewise, the streamheight recorded in each image does not exceed the col-umn diameter by factor of 3.

Convincingly, this work has converged to experi-mental work, which provides the stream diameter as afunction of length calculated using the digital imageswith the home-made ad hoc MATLAB� code pro-gramed by the present authors. Additionally, the char-acteristics of the nozzle orifice radius, stream radius,and length are indirectly included in the stream shapeand then, an empirical methodology to estimate theflow rate using the falling velocity is derived whichagrees moderately well with the experimental data.

Results and discussions

This section explores the significance of the results ofthis work.

Water stream shape and stability: empiricalmethodology

The empirical methodology to the water stream shapewas applied to Nim=900 images, being approximately

Figure 3. Sketch of the fluid jet.




50 cycles of the maximum vertical displacement of thestream, which was proved to have sufficient accuracy.It is not the aim of this work to perform neither a com-prehensive experimental study on the stationary streamshape nor an experimental stability model.

Three different constant positions of the stopper rodelevation have been studied: h=1mm, h=3mm, andh=6mm. They have been chosen as the most repre-sentative shape and stability modes and typologies ofthe water stream as a function of stopper rod elevation.

Figure 4. Evolution of the width contour (stream diameter) of the water stream when falling from the nozzle to the maximumvertical displacement for all images for the three stopper rod elevations, with and without plug: (a) h = 1 mm without plug, (b)h = 1 mm with plug, (c) h = 3 mm without plug, (d) h = 3 mm with plug, (e) h = 6 mm without plug, and (f) h = 6 mm with plug.




Besides, for each elevation, two cases are performed:with rubber plug (no air through stopper rod interior)and without rubber plug (allowing air through stopperrod interior) to experimentally simulate the clogging ofthe stopper rod during casting iron operation.

Figure 4 superimposes the evolution of the widthcontour (stream diameter) of the water stream whenfalling from the nozzle to the maximum vertical displa-cement for all images. The stream height has been nor-malized with the maximum vertical displacement andthe stream width with the nozzle diameter. Figure 4(a),(c), and (d) shows the great influence of the stopper rodelevation on the performance of each water stream. Asthe elevation increases, the stream diameter becomeswider and unstable with the falling distance.

Figure 4 also shows a blue dashed and dotted line,which is the jet shape function of equation (2). Withinall range of elevations, a narrowing occurs just at theexit of the nozzle. However, an important expansionoccurs at 1mm elevation, and at 3 and 6mm elevations

the stream approaches the jet shape function. This ten-dency is corroborated in Figure 5 at three differentstream vertical positions (initial z=0, half z= zmax/2and final z= zmax) and its mean value.

The presence of this expansion is owing to the flowconditions generated between nozzle and stopper rodsection that causes the entrainment of air to the interiorof the stream in its free-falling trajectory. At the eleva-tion of 3mm, the good agreement occurs for the no-plug configuration with the theoretical jet shape,despite the width contour scatter as it depicts inFigure 6. An example of the importance of the influ-ence on the typology of the stream is shown in Figure 7by means of histograms.

Water stream flow rate estimation: experimentalvalidation

An experimental validation of the VPT is to estimatethe flow rate by means of the presented empirical

Figure 5. Evolution of the width contour (stream diameter) of the water stream at initial, half, and final vertical displacement forthe three stopper rod elevations with plug: (a) h = 1 mm with plug, (b) h = 3 mm with plug, and (c) h = 6 mm with plug.




methodology based on a simplified approach describedas follows:

1. Find the maximum width contour of the streamall over its height (z=0O zmax) for each frameimage.

2. Calculate the number of vertical pixels betweentwo consecutive maximum width contours. Thisnumber of vertical pixels will be used to evaluatethe time of falling between two consecutive max-imums of the width contour of the stream.

3. Calculate the mean value of the number of verti-cal pixels Npx per frame image.

4. Estimate the free-fall velocity as Npx=(NpxmmNfps).5. Calculate the mean of the width contour of the

stream all over its height (z=0O zmax) for eachframe image, being an estimation of the dia-meter of the stream.

6. Estimate the cross-sectional area of the streamas a circular section.

7. Estimate the flow rate of the stream as the mul-tiplication of free-fall velocity and the area ofthe stream.

Figure 8 depicts the flow rate estimation for the sixexperiments performed, summarized in Figure 4 andnormalized by the theoretical flow rate.

This experiment demonstrates the capabilities of thepresented empirical methodology to predict flow rateand shows the differences in performance of the waterstream as a function of the stopper rod elevation, mainlyregarding to air entrainment. The elevation of 1mmwithout plug presents the maximum error on flow rateestimation, around 72% as it is depicted in Figure 9. Aspreviously demonstrated, this configuration entrains airto the center of the stream turning into an annular cross

Figure 6. Scattering of the stream width at half verticaldisplacement (z = zmax/2) with plug: (a) h = 3 mm with plug and(b) h = 6 mm with plug.

Figure 7. Histogram of the stream width at half verticaldisplacement (z = zmax/2) without plug: (a) h = 1 mm withoutplug: annular cross section typology and (b) h = 6 mm withoutplug: circular cross section typology.




section. Hence, the width contour as stream diameter toestimate the flow quantities drives to an important error.However, this error is not useless, on the contrary, it hasa critical significance: it must be taken as an alarm in thein-situ casting process because the iron mass flow fromthe tundish would not fulfill the corresponding moldand, consequently, the casting parts will be rejected.

However, this error becomes minor as the stopperrod height increases as shown in Figure 9. The elevationof 6mm, with and without plug, is the best configura-tion once the influence of the stopper rod is negligibleand the nozzle becomes the slot of a gravity-drivenstream issuing the pouring box.

Casting iron stream characterization: practicedevelopment

The main difference in iron stream properties, besides thetemperature, is its surface tension. Data of the surface ten-sion of cast iron have been proven experimentally, andvalues between 990 and 1465mNm21 are obtained as aresult of its relationship with the graphite shape.25

The empirical methodology to the iron stream shapewas applied on real images from the optical sensing sys-tem on a foundry. This system is used to late-steam ironinoculation and also to visual inspection of the continu-ous casting flow rate by manual control of the stopper

Figure 8. Flow rate quantities estimation.

Figure 9. Relative error on flow rate estimation versus stopper rod elevation, with and without plug.




rod elevation. The VPT and the empirical methodologyare applied to the real images to identify the iron streamboundaries and contour, as it is shown in Figure 10.

Currently, the optical sensing system allows a maxi-mum of 19 images per mold, which is proved not tohave enough accuracy, as is it is depicted in Figure 11.The aim of this work is to perform neither a compre-hensive experimental study on the stationary streamshape nor an experimental stability model of an ironstream, but to show the potential of the proposedapproach to setup on casting iron stream quality.

Conclusion

An empirical methodology experimentally validated tocharacterize the shape and estimate the flow rate of afree-falling non-submerged liquid stream has been pre-sented. The VPT is well-suited as a non-intrusive tech-nique that allows predicting the liquid streamboundaries and contours without any type of treatmenton the fluid and independent of its temperature. As afruit of these measurements, the results lead to threetypologies of the water stream:

� Annular cross section typology. Maximum influ-ence of the low stopper rod elevation. The nar-rowed section between nozzle and stopper rod,and the air entrainment into the interior of thestream drive the stream in its free-falling. The

Figure 10. VPT procedure on a casting iron stream.

Figure 11. Evolution of the width contour of the iron streamat initial, half, and final vertical displacement.




influence of the stopper rod with and withoutplug is also critical.

� Contour shape instability typology. Combinedinfluence of the stopper rod and the nozzle atmedium stopper rod elevation. A quasi-stationary periodic instability on the boundariesof the stream in its free-falling is presented. Theinfluence of the stopper rod with or without plugloses significance.

� Circular cross section typology. Maximum influ-ence of the nozzle at high stopper rod elevation.The nozzle drives the free-falling characteristicsof the stream and no air is encountered. Theinfluence of the stopper rod, with or withoutplug, is negligible.

With regard to the plug cases, the configuration withplug presents less flow rate estimation error for lowstopper rod elevation (50%) compared to the config-uration without plug (72%). However, the low stopperrod elevation is a transitory flow configuration on cast-ing iron in a foundry, which means that this error onthe flow rate estimation to this elevation turns out to beof less importance. As the stopper elevation increases,this error becomes much smaller without plug (2%)than with plug (13%). Keeping in mind that the config-uration with plug simulates the real clogging stopperrod, it means that special care must be taken in thefoundry with the stopper rod containing slag.

Whether this empirical methodology holds for cast-ing iron stream or not is demonstrated with the pre-sented empirical methodology. Nevertheless, moreexperiments have to be performed with an improvedoptical sensing system on the foundry to obtain moreaccurate results.

An important conclusion is that the presentedempirical methodology can be installed and performedin a foundry in a real-time cast iron industrialprocedure, despite the hard operation conditions,improving the quality and reducing rejection of castiron products.

Acknowledgements

The authors gratefully acknowledge the lab technicians MrJaume Bonastre and Mr Justo Zoyo for providing laboratoryresources and support in the experimental work. The authorswould also like to acknowledge the support of theFundiciones de Roda (Spain).

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest withrespect to the research, authorship, and/or publication of thisarticle.

Funding

The author(s) disclosed receipt of the following financial sup-port for the research, authorship, and/or publication of thisarticle: This research program has received funding from theEuropean Union Seventh Framework Programme (FP7/2007-2013) under grant agreement no. 314540.

References

1. Lavers JD and Kadar L. Application of electromagneticforces to reduce tundish nozzle clogging. Appl Math

Model 2004; 28: 29–45.2. Rajaratnam N. Turbulent jets. New York: Elsevier, 1976.

3. Eggers J and Villermaux E. Physics of liquid jets. RepProg Phys 2008; 71: 1–79.

4. Jeffrey GB, Steven RA and Rodney JW. Predictingplunge pool velocity decay of free falling, rectangular jet.J Hydraul Eng 1998; 124: 1043–1048.

5. Raffel M, Willert CE, Wereley S, et al. Particle image

velocimetry: a practical guide. New York: Springer, 2007.6. Adrian RJ and Westerweel J. Particle image velocimetry.

New York: Cambridge University Press, 2011.7. Melling A. Tracer particles and seeding for particle

image velocimetry. Meas Sci Technol 1997; 8: 1406–1416.8. Hammad KJ and Milanovic I. Effect of Reynolds num-

ber on the turbulent flow structure in the near-wallregion of an impinging round jet. In: Proceedings of

AJK2011-FED, Hamamatsu, Japan, 24–29 July 2011,pp.2875–2882. New York: ASME.

9. Qu X, Goharzadeh A, Khezzar L, et al. Experimentalcharacterization of air-entrainment in a plunging jet. ExpTherm Fluid Sci 2013; 44: 51–61.

10. Chaudhary R, Lee Go-Gi, Thomas BG, et al. Transientmold fluid flow with well- and mountain-bottom nozzlesin continuous casting of steel. Metall Mater Trans B

2008; 39: 870–884.11. Jiang Y, Li H, Xiang Q, et al. Investigation on spatial

breakup characteristics of low-pressure jets with particleimage velocimetry experiment and volume of fluid-levelset simulation. Adv Mech Eng 2016; 8: 1–10.

12. O’Neill P, Soria J and Honnery D. The stability of lowReynolds number round jets. Exp Fluids 2004; 36:473–483.

13. Gong C, Yang M, Wang Y, et al. Wavelength measure-ment of a liquid jet based on spectral analysis of imageintensity. Adv Mech Eng 2015; 7: 1–10.

14. Brucker Ch. PIV in two-phase flows (Lecture Series 2000–01). Sint-Genesius-Rode: Von Karman Institute for FluidDynamics, 2000.

15. Qu XL, Khezzar L, Danciu D, et al. Characterization ofplunging liquid jets: a combined experimental and numer-ical investigation. Int J Multiphas Flow 2011; 37: 722–731.

16. Roy AK, Maitib B and Dasb PK. Visualisation of airentrainment by a plunging jet. Proc Eng 2013; 56:468–473.

17. Harby K, Chiva S and Munoz-Cobo JL. An experimen-tal study on bubble entrainment and flow characteristicsof vertical plunging water jets. Exp Therm Fluid Sci 2014;57: 207–220.




18. Foseco. Iron turbulence analysis via foundry water modelII: advanced testing. Foundry Pract 2014; 261: 8–15.

19. Plateau J. Experimental and theoretical statics of liquids

subject to molecular forces only. Paris: Gauthier-Villars,1873.

20. Rayleigh L. On the instability of jets. Proc London Math

Soc 1878; s1-10: 4–13.21. Massalha T and Digilov RM. The shape function of a

free-falling laminar jet: making use of Bernoulli’s equa-tion. Am J Phys 2013; 81: 733–737.

22. Le Dizes S. Global modes in falling capillary jets. Eur JMech B: Fluid 1997; 16: 761–778.

23. Clanet C and Lasheras JC. Transition from dripping tojetting. J Fluid Mech 1999; 383: 307–326.

24. Sauter US and Buggisch HW. Stability of initially slowviscous jets driven by gravity. J Fluid Mech 2005; 533:237–257.

25. Shi D, Li D, Gao G, et al. Relation between surface ten-sion and graphite shape in cast iron. Mater Trans 2008;49: 2163–2165.




advances in mechanical engineering 2016, vol. 8(9) 1–12 an … · 2016-11-08 · research article...

Documents