Effects of sample dimensions and shapes on measuring soil-water characteristic curves using pressure plate

Min Wang,Lingwei Kong,Meng Zang

aCivil and Computational Engineering Centre,College of Engineering,Swansea University,Singleton Park,Swansea SA2 8PP,UK

bState Key Laboratory of Geomechanics and Geotechnical Engineering,Institute of Rock and Soil Mechanics,Chinese Academy of Sciences,Wuhan 430071, China

Effects of sample dimensions and shapes on measuring soil-water characteristic curves using pressure plate

Min Wanga,*,Lingwei Kongb,Meng Zangb

aCivil and Computational Engineering Centre,College of Engineering,Swansea University,Singleton Park,Swansea SA2 8PP,UK

bState Key Laboratory of Geomechanics and Geotechnical Engineering,Institute of Rock and Soil Mechanics,Chinese Academy of Sciences,Wuhan 430071, China

A R T I C L E I N F O

Article history:

Received 8 December 2014

Received in revised form

8 January 2015

Accepted 13 January 2015

Available online 28 January 2015

Soil-water characteristic curve(SWCC)

Pressure plate

Mercury intrusion

Microstructure

Expansive soil

It is well known that soil-water characteristic curve(SWCC)plays an important role in unsaturated soil mechanics,but the measurement of SWCC is inconvenient.In laboratory it requires days of testing time. For fne-grained clays,it may last for a couple of months using pressure plate tests.In this study,the effects of sample dimensions and shapes on the balance time of measuring SWCCs using pressure plate tests and the shape of SWCCs are investigated.It can be found that the sample dimensions and shapes have apparent infuence on the balance time.The testing durations for circular samples with smaller diameters and annular samples with larger contact area are signifcantly shortened.However,there is little effect of sample dimensions and shapes on the shape of SWCCs.Its mechanism is explored and discussed in details through analysing the principle of pressure plate tests and microstructure of the sample.Based on the above fndings,it is found that the circular samples with smaller dimensions can accelerate the testing duration of SWCC using the pressure plate.

©2015 Institute of Rock and Soil Mechanics,Chinese Academy of Sciences.Production and hosting by Elsevier B.V.All rights reserved.

1.Introduction

In the past two decades,a considerable effort has been made for understanding the mechanics of unsaturated soils.A number of textbooks(Fredlund and Rahardjo,1993;Lu and Likos,2004)have been published whilst scores of national and international meetings have been held throughout the world where hundreds of papers(e.g.Alonso et al.,1990;Pietruszczak and Pande,1991;Delage et al.,1998)have been presented and published in prestigious internationaljournals.Generalconsensusamongstresearchers seems to centre around two new concepts which have been introduced to model the behaviour of unsaturated soils.These are (a)suction defned as the difference between air pressure(ua)and water pressure(uw)as an independent state variable,and(b)a relationship between water content or void ratio and suction, variously known as soil-water characteristic curve(SWCC)or soilwater retention curve(SWRC).Suction was introduced as an independent state parameter in what is popularly known as Barcelona basic model(BBM)(Alonso et al.,1990),whilst SWCC is used to supplement a constitutive model like BBM(Wheeler et al.,2003; Sheng et al.,2004)in which the hydro-mechanical coupling is incorporated.In addition,the unsaturated hydraulic conductivity function is often predicted from SWCC(Simms and Yanful,2002). As SWCC does play a vital role in unsaturated soil mechanics,its measurement is of great importance.

The pressure plate test(Richards,1941)is one of the testing methods in geotechnical engineering.It can be treated as an air intrusion test(where air is the non-wetting fuid)compared with mercury intrusion tests(where the non-wetting fuid,mercury, intrudes into the pores).This idea has also struck a couple of researchers since the 1980s(Prapaharan et al.,1985;Romero and Simms,2008),and they used the data of mercury intrusion porosimeter(MIP)to predict the SWCC of a soil.Like in mercury intrusion tests,the non-wetting fuid,air,cannot get access to pores freely and pressures need to be applied.Different pressures correspond to different pore sizes.Therefore,the outfow of water represents the volume of the pore of certain size under a specifed suction.It is known that the testing process is time-consuming especially for fne-grained clayey soils and this problem limits the utilisation of the pressure plate test.Thereby reducing the testing duration of the pressure plate test makes sense.

To date,the effect of sample height on the testing duration of pressure plate tests has been studied(e.g.Richards,1956;Dane and Hopmans,2002;Bittelli and Flury,2009).The sample dimensionsin the range of 5-8 cm in diameter and 1-6 cm in height were adopted.It is found that the time for reaching equilibrium increasesrapidly with the height of samples.To minimise the equilibration time,it is suggested that the height should therefore be kept relatively small(normally 1-2 cm).Although different diameters were used in their tests,the effect of sample diameters on pressure plate tests was not mentioned.

Table 1Basic physical properties of expansive soil.

The infuence of drying rate on the shrinkage characteristics of residual soils was explored by Krisdani et al.(2008).The cylindrical specimens were dried under an ambient condition,an accelerated evaporation condition using a fan,and an accelerated evaporation using a lamp.The rates of changes in void ratio and degree of saturation were highly affected by the evaporation rates.Similar results can be seen in expansive soils(Kong et al.,2009).In order to obtain different drying rates,the samples of an intact expansive soil were dried under the temperature of 25°C and relative humidity of 30%,60%,75%and90%,respectively,inaconstantclimatechamber.It is found that the lower drying rate leads to larger shrinkage deformation and less testing duration.Later,the effect of geometry factor including sample shapes and dimensions on the shrinkage of an expansive soil was investigated by Wang et al.(2012).The larger contact area between samples and gas can lead to a higher drying rate.In above literature,the drying process is triggered by water evaporation;while in the pressure plate test,it is promoted by the drainage of water due to gas pressure.Therefore,the effect of drying rate on the pressure plate extractor attracts the authors’attention.

It seems that sample shapes and dimensions will affect the testing duration of the pressure plate test,but to what extent?Will they infuence the shape of SWCCs?Based on these questions and the fnding thatthe high dryingrate can shortenthe testingduration of the soil shrinkage tests,the effects of sample dimensions and shapes on the pressure plate test are explored in an attempt to shorten the time of pressure plate tests made in this study.

2.Experimental method

2.1.Experimental soil

The soil tested in this study was sampled from 1.8-2.5 m below the surface of a slope at the Institute of Buffalo,Academy of Agriculture,in the suburb of Nanning,Guangxi Province,China.The basic physical properties and the swelling-shrinkage characteristics of this expansive soil are depicted in Tables 1 and 2,respectively. According to the criterion of swelling potential grade,this grey white expansive soil belongs to moderately strong expansive soil.

2.2.The pressure plate test

2.2.1.Sample preparation

For investigation of the infuence of sample dimensions on the balance time for measuring SWCCs,two kinds of undisturbed specimens are adopted,i.e.samples in different shapes(annular and circular shapes)and samples in different sizes.The intact samples are made through cutting the intact soil using selfdeveloped moulds(see Figs.1 and 2).The dimensions of small and large annular samples(20 mm in height)are 87.4 mm and 109.2 mm in outer diameter(OD),61.8 mm and 90 mm in inner diameter(ID),respectively.The diameters(d)of circular samples (20 mm in height)are 50.5 mm,61.8 mm,79.8 mm and 100 mm. For comparison,the volume and mass of both small and large annular samples are the same as the circular sample of 61.8 mm in diameter.Due to the difference of sample size and shape,the contact areas between samples and atmosphere are different. Therefore,the drying rates which are proportional to the ratio of contact area to volume are different during desiccation.Finally,the intact samples are saturated by the vacuum saturation method.

Table 2Swelling-shrinkage characteristics.

2.2.2.Test procedure

Fig.1.Self-developed moulds.(a)Moulds for annular samples;(b)Moulds for circular samples.

The pressure plate equipment used in this study is made by American Soil Moisture Equipment Corporation(see Fig.3).Itconsists of a pressure vessel which can be pressurised by nitrogen up to a pressure of 1500 kPa.A number of saturated soil samples in different shapes and sizes are placed on a ceramic disc which has a specifed air entry pressure value.The disc is connected to the atmosphereand water is allowed tofowout freely.Nitrogenpressure controlled by a nitrogen tank is applied in steps with pressure(P) being held at 10 kPa,30 kPa,50 kPa,100 kPa,300 kPa,500 kPa, 1200 kPa,respectively.During the whole test,the temperature in laboratory was controlled at around 20°C.In order to determine the balance time under each pressure,the samples are removed and weighted using electronic scales having accuracy of±0.01 g every 12 h for each pressure step.If the mass of the sample remains unchanged after 24 h,it is assumed to be in a state of equilibrium.

Fig.2.Samples in different shapes and sizes.(a)Annular samples;(b)Circular samples.

Fig.3.Pressure plate apparatus.

2.3.The mercury intrusion test

Saturated samples for the pressure plate test were chopped into small cubes with the side length of 1 cm using thin blade.Then, freeze the cubes rapidly in liquid nitrogen for 15 min and dry it through sublimation in vacuum for 24 h using refrigeration dryer under-50°C.This freeze-drying method can eliminate water from samples with minimal pore volume change(Delage et al., 1996).Finally,the mercury intrusion test was performed using the dried samples.The adopted MIP is American AutoPore IV9510 porosimeter with maximal aperture of 360μm and minimal aperture of 0.023μm,respectively.

3.Test results

Fig.4 shows the relationship between the water content and the elapsed time.The SWCCs(also called SWRC)of samples are depicted in Fig.5.From Fig.4,we can fnd that the drying process under each pressure can be divided into two periods,i.e.the rapid period and the slow one.With the growth of pressure,the time required for balance increases.It should be highlighted that the testing duration for this expansive soil lasts nearly 170 d,but there is no obvious difference among shapes of the SWCCs of all samples, which can be seen in Fig.5.Only the sample D(d=100 mm)has a larger variance due to the difference caused by saturation process.

Table 3 indicates the balance time of soil samples under each pressure(P).The balance time vs.pressure curve and the balance time-sample diameter curve are given in Figs.6 and 7,respectively. From Table 3 and Fig.6,it can be found that with the growth of gas pressure,the balance time increases continuously.Taking sample B for example,the balance time is only 5 d forP=10 kPa,when the pressure is increased to 300 kPa and 1200 kPa,the corresponding balance time increases to 12 d and 52 d,respectively.

Fig.4.Variation of water content vs.time.

The annulus 2,annulus 1 and sample B are designed to have the same volume and mass.Actually,their masses are 120.5 g,116.7 g and 121.17 g,respectively,as the error is introduced during sample preparation.Due to the difference of sample dimensions andshapes,the contact areas between samples and atmosphere are different.For annular samples,the bottom is in touch with ceramic plates and the contact areas between gas and samples are top surface,lateral and medial surfaces.In contrast,the contact areas of circular samples are only top and lateral surfaces.Therefore,the lengths of drainage paths are of certain distinction.Particularly, annulus 2 has the shortest drainage path and sample B has the longest one,and the relation of contact area per unit mass for annulus 2,annulus 1 and sample B is 2.2:1.8:1.The balance time for these samples reduces with the increase of contact area per unit mass and the decrease of the drainage path.The variance becomes more apparent when the pressure increases.

The relation of the contact area per unit mass for circular samples in different diameters(i.e.samples D,C,B and A)is 1.45:1.28:1.13:1.It can be also found that the balance time increases with the sample size under the same pressure(see Figs.6 and 7). This phenomenon is more obvious under higher pressures.Taking samples D and A for example,the balance time was reduced by 29% under gas pressure of 30 kPa from 8.5 d to 6 d.When the pressure was increased to 120 kPa,the balance time was reduced by 35% from 63 d to 41 d.In addition,when the pressure increases,the balance time increases rapidly.

Fig.5.Soil-water characteristic curves of(a)annular samples and(b)circular samples.

Fig.6.The balance time vs.pressure curves of various soil samples.

Fig.7.The balance time vs.sample diameter curves under different pressures.

4.Discussions

From the results above,we can fnd that annular samples and the circular sample in small diameter can shorten the testing duration of pressure plate tests,and the effect of the latter on testing duration is more apparent.However,little infuence of sample dimensions and shapes on the shape of SWCCs can be obtained.This fnding is very important for the SWCC test of fnegrained soils.The principle accounting for this phenomenon isthat pressure plate test is a gas intrusion test(where air is the nonwetting fuid)compared with mercury intrusion tests(where the non-wetting fuid,mercury,intrudes into the pores).During the whole desiccation,the gas will intrude into large pores and expel water in large pores frst,and with the increase of pressures,gas will enter small pores gradually.Because the mercury intrusion test is commonly used for determining the pore size distribution of porous media,SWCC is the refection of the pore size distribution of the sample.This can be proved through comparing mercury intrusion data with SWCC data.

Table 3Balance time under each pressure(unit:d).

The following equation used in mercury intrusion porosimetry can be adopted for calculating pore size in pressure plate test:

whereris the soil pore radius,Tsis the surface tension (Ts=0.480 N/m for mercury intrusion,and 0.07275 N/m for gas intrusion),andαis the contact angle between the soil particle and mercury(α=140°for mercury intrusion,and 180°for gas intrusion).

The cumulative outfow of water is the cumulative pore volume. Therefore,the pore size distribution curve can be obtained through SWCC test and this result is compared with mercury intrusion data in Fig.8.It can be found that both curves match well.The pore radius ranges from 102 nm to 143,293 nm.From the pore radius distribution curve(Fig.9),we can fnd that the pores in SWCC are only half of the mercury intrusion data,and the fractional pore volume decreases quickly with the pore size within the resolution of pressure plate tests.The pore distribution in these samples should be identical.This is the reason why the samples with different sizes and shapes could be used for the pressure plate test.

The factors determining the desiccation period comprise the drainage length and pore aperture.Due to the difference of sample shapes,the contact areas between samples and atmosphere are different.It results in annulus 2 with the largest contact area having the shortest drainage path,while sample B with the smallest contact area has the largest drainage length.Similarly,for circular samples,the one with smaller diameter has larger contact area per unit mass and shorter drainage path.In addition,it can be found that higher gas pressure corresponds to smaller pore size and more pore volume within the range of 10-1200 kPa from Fig.9.This is the main reason why it requires more desiccation time under higher gas pressure.As all samples will be subjected to the same gas pressures of 10 kPa,30 kPa,50 kPa,100 kPa,300 kPa,500 kPa, and 1200 kPa,only the drainage length needs to be considered. Based on the experimental results,it can be seen that the circular sample with small diameter and annular samples are promising in shortening the testing duration of the pressure plate tests.And the circular sample with small dimension has more effect than annular samples.

Fig.8.The cumulative pore volume-radius curve.

Fig.9.The pore diameter histogram from mercury intrusion porosimeter.

5.Conclusions

In this study,the effects of sample shapes and dimensions on SWCC tests have been investigated and the principles of the pressure plate test are discussed.Some conclusions can be drawn as follows:

(1)The annular samples with large contact area and circular samples with small dimensions require short testing duration in pressure plate tests,and the effect of the latter is more apparent.

(2)Through comparing the mechanisms and results of pressure plate and mercury intrusion tests,the pressure plate test can be treated as the gas intrusion test which can be used for determining pore size distribution of soils like the mercury intrusion test.

(3)The infuence of sample dimensions and shapes on the shape of SWCCs is negligible.The mechanism is that the pressure plate test is a pore size distribution test.

(4)Based on the above conclusions,the circular sample with small dimensions is recommended for accelerating the pressure plate test.

Confict of interest

The authors wish to confrm that there are no known conficts of interest associated with this publication and there has been no signifcant fnancial support for this work that could have infuenced its outcome.

Acknowledgements

The research work was supported by the National Natural Science Foundation of China(Grant No.10872210)and the State Key Laboratory of Geomechanics and Geotechnical Engineering(Grant No.Y11002).The authors are verygratefultotheir fnancial support. Thanks are extended to Professor Emeritus G.N.Pande in Swansea University,UK and reviewers of this journal for their valuable suggestions on forming this paper.

Alonso EE,Gens A,Josa A.A constitutive model for partially saturated soils.Geotechnique 1990;40(3):405-30.

Bittelli M,Flury M.Errors in water retention curves determined with pressure plates.Soil Science Society of America Journal 2009;73(5):1453-60.

Dane JH,Hopmans JW.Pressure plate extractor.In:Dane JH,Topp GC,editors. Methods of soil analysis.Madison,WI,USA:Soil Science Society of America; 2002.

Delage P,Audiguier M,Cui YJ,Howat DM.Microstructure of a compacted silt.Canadian Geotechnical Journal 1996;33(1):150-8.

Delage P,Howat MD,Cui YJ.The relationship between suction and swelling propertiesinaheavilycompactedunsaturatedclay.EngineeringGeology 1998;50(1-2):31-48.

Fredlund DG,Rahardjo H.Soil mechanics for unsaturated soils.New York,USA:John Wiley and Sons;1993.

Kong LW,Li XW,Guo AG,Zhang Y.Preliminary study on engineering behaviours and water retention characteristics of expansive soil under infuence of drying rate.Chinese Journal of Geotechnical Engineering 2009;31(2):335-40(in Chinese).

Krisdani H,Rahardjo H,Leong EC.Effects of different drying rates on shrinkage characteristics of a residual soil and soil mixtures.Engineering Geology 2008;102(1-2):31-7.

Lu N,Likos WJ.Unsaturated soil mechanics.New York,USA:John Wiley and Sons; 2004.

Pietruszczak S,Pande GN.On the mechanics of partially saturated soils.Computers and Geotechnics 1991;12(1):55-71.

Prapaharan S,AltschaeffA,Dempsey B.Moisture curve of compacted clay:mercury intrusion method.Journal of Geotechnical Engineering 1985;111(9):1139-43.

Richards LA.A pressure-membrane extraction apparatus for soil solution.Soil Science 1941;51(5):377-86.

Richards LA.Sample retainers for measuring water retention by soil.Soil Science Society of America Journal 1956;20(2):301-3.

Romero E,Simms PH.Microstructure investigation in unsaturated soils:a review with special attention to contribution of mercury intrusion porosimetry and environmental scanning electron microscopy.Geotechnical and Geological Engineering 2008;26(6):705-27.

Sheng D,Sloan SW,Gens A.A constitutive model for unsaturated soils:thermomechanical and computational aspects.Computational Mechanics 2004;33(6): 453-65.

Simms P,Yanful E.Predicting soil-water characteristic curves of compacted plastic soils from measured pore-size distributions.Geotechnique 2002;52(4):269-78. Wang M,Kong LW,Zang M.Effects of drying environment and sample type on the shrinkage properties of expansive soil.In:Kong LW,editor.Proceedings of the 4th International Conference on Problematic Soils,Wuhan,China;2012.

Wheeler SJ,Sharma RS,Buisson MSR.Coupling of hydraulic hysteresis and stressstrain behaviour in unsaturated soils.Geotechnique 2003;53(1):41-54.

Min Wang,Ph.D.candidate in Swansea University,UK,obtained his B.S.degree in Civil Engineering from China University of Mining and Technology in 2009 and his M.S.in Geotechnical Engineering(disaster prevention and mitigation engineering)from Institute of Rock and Soil Mechanics, Chinese Academy of Sciences in 2012.He spent the frst year (2012.09-2013.09)of his Ph.D.study in the University of Birmingham,UK,and the last two years in Swansea University,UK.His research interests include(a)problematic soil mechanics,(b)fuid-solid coupling,(c)particle erosion and hydraulic fracture,(d)numerical methods including discrete element method,lattice Boltzmann method, bonded particle method and their coupling.His current Ph.D.project is:Coupled bonded particle and lattice Boltzmann method and its application to geomechanics.

*Corresponding author.Tel.:+44 07710606945.

E-mail address:sacewangmin@gmail.com(M.Wang).

Peer review under responsibility of Institute of Rock and Soil Mechanics,Chinese Academy of Sciences.

1674-7755©2015 Institute of Rock and Soil Mechanics,Chinese Academy of Sciences.Production and hosting by Elsevier B.V.All rights reserved.

http://dx.doi.org/10.1016/j.jrmge.2015.01.002