Mechanisms of longitudinal cracks along pavement shoulder in ...

Unsaturated Soils: Theory and Practice 2011Jotisankasa, Sawangsuriya, Soralump and Mairaing (Editors)Kasetsart University, Thailand, ISBN 978-616-7522-77-7Mechanisms of longitudinal cracks along pavement shoulder in centralThailand Mechanisms of longitudinal cracks along pavement shoulder in centralThailandApiniti Jotisankasa, Barames Vadhanabhuti & Krit LousuphapGeotechnical Engineering Research and Development Center, Department of Civil Engineering, KasetsartUniversity, Bangkok, Thailand, fengatj@ku.ac.thA. Jotisankasa, B. Vadhanabhuti & K. LousuphapAuckpath Geotechnical SawangsuriyaEngineering Research and Development Center, Department of Civil Engineering, KasetsartBureau University, of Bangkok, Road Thailand, Research fengatj@ku.ac.thand Development, Department of Highways, Bangkok, Thailand,sawangsuriya@gmail.comA. SawangsuriyaBureau of Road Research and Development, Department of Highways, Bangkok, Thailand,sawangsuriya@gmail.comABSTRACT: Cracks along asphaltic-concrete shoulder of highways are one of the most common infrastructuredamages found in the central part of Thailand where sedimentary soils of high plasticity were widely depositedand sometimes used as subgrade materials. This paper reports on a case study of highway route no.357 in Supanburi province with emphasis on the mechanism of the longitudinal cracks initiated by desiccationcracking of the clay subgrade, as well as infiltration-induced slope instability. The studies included field investigation,monitoring of pore water pressure and suction, soil-water characteristic curve determination, aswell as the suction monitored-direct shear tests. Seepage and deformation analyses were used to demonstratethe initiation of crack induced by evaporation from the embankment. Limit Equilibrium method was thenused to analyse the influence of water-filled crack on the stability of the embankment.KEYWORDS: longitudinal shoulder cracks; shrink-swell clay; unsaturated soils; strength; embankment stability;suction1 INTRODUCTIONThe longitudinal shoulder crack along highways isquite a common problem in central Thailand wheresedimentary soils of high plasticity were widely depositedand sometimes used as subgrade materials(Figure 1). A number of failure mechanisms havebeen postulated and studied (e.g. Taesiri et al., 2004,DOH, 2007), such as, instability of embankment onsoft clays, under-compaction of subgrade materialalong shoulder, and heavy traffic load in excess ofthe design value.This study will deal with a case of longitudinalshoulder cracks along highway which was foundedon stiff clay foundation and where medium-to-highplasticity clay was used as subgrade. Interestingly,the cracks started to appear in some locations evenbefore this highway was opened to public use (Figure1). In addition, severe cracks along highway reportedlyappeared to exacerbate after a period offlooding and subsequent drying. Climatic influenceon volumetric shrinkage and shallow slope stabilityof the subgrade material was thus believed to playan important role. Moreover, due to lack of suitableconstruction materials in the area, high plasticitymaterials might be increasingly chosen as subgradealong highway shoulder. It is thus of great importanceto understand the mechanism of such shouldercracks so that appropriate remedial and preventivemeasures can be arrived at.Several researchers have studied the climatic influenceon longitudinal cracks of highway shoulderand slope deformation. Luo & Prozzi (2008) foundthat the non-uniform matric suction change in thesubgrade could cause shrinkage stresses in the subgradesoil and pavement structure. This stress concentrationcould be large enough to drive the crackto propagate further to the shoulder surface. Take &Bolton (2004) observed that seasonal pore pressurecycles in slope could also drive the mechanism ofslope creep leading eventually to a progressive failurethrough the development of cracks towards thecrest of a slope and a shear rupture at a toe.Figure 1. Cracks along shoulders of Highway route 357(km32+909), Suparnburi, Central Thailand (the right figureshowing a test pit)699

In this paper, unsaturated soil mechanics will beused to explain such mechanism of cracks alonghighway shoulder, involving shrink-swell clay subgrade,in central Thailand. Influence of drying oninitiation of cracks will be explored based on seepageand deformation analyses. The fully saturatedand unsaturated shear strength parameters are usedto determine the stability of slope which exacerbatedby water infiltration, and strength softening.Table __________________________________________________1. Summary of subgrade properties at the test site T1Specific LL PL PI Wn γ d SPTGravity, % % % % t/m 3 Blow/ftGs __________________________________________________2.71 43-53 17-21 26-34 20-27 1.49- 101.69LL = Liquid limit, PL = Plastic Limit, PI = Plasticity Index,Wn = Natural water content, sampled on 13-16 August 2009(during rainy season), γ d = Dry density, SPT (uncorrected)2 TEST SITE2.1 Highway route 357Highway route no. 357 (Suparnburi by-pass) locatedin Suparnburi province, central region of Thailandwas selected as the test site where severe longitudinalcracking along asphalt pavement shoulder wasencountered prior to trafficking. The test sites consistedof two adjacent sections: one section havingsevere cracks (sta. 32+600 to sta. 33+100) designatedas T-1 and another having minor cracks (sta.31+600 to sta.32+100) designated as T-2. This paperhowever focuses on the section T-1 only.2.2 Field investigation and basic propertiesAn extensive field investigation programme (includingborehole/test pits sampling, field vane shear,SPT, Kunzelstab penetration tests) revealed that thefoundation soil consists of stiff clay and very stiffclay to the depth of 5.5 metres below which existsthe medium-dense to dense sand. The subgrade materialis classified as a medium-to-high plasticityclay (CL-CH) with thickness of about 2m. Its basicproperties are summarized in Table 1. The soil mapfrom Department of Land Development (used foragricultural purpose) also indicated that Vertisol existedin shallow depth around the vicinity of thehighway. The depth of cracks at Site T1 (Figure 1)appeared to be at least 48 cm and the longitudinallength varied from 2.3 to 90 metres.The maximum dry density, γ dmax , from modifiedProctor tests was 1.74 t/m 3 and the optimum moisturecontent, OMC, was 13.3%. Therefore, the percentcompaction is 86-97% and the soil, especiallyclose to surface, must have wetted up from its ascompactedstate and perhaps have expanded. Detailedsounding tests using Kunzelstab light weightpenetrometer along the slope of embankment indicatesthat the soft zone ran sub-parallel to the slopeand down to depth of 0.5 to 1 metre, as shown inFigure 3. The cracks thus seemed to be associatedwith shallow movement.0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.50-0.5-1-1.5Distance, m-2Depth, m-2.5-3-3.5-4-4.5-5Figure 2. A typical cross-section of the test section.This highway consists of two carriage lanes ineach direction. The carriage lanes were constructedusing the concrete pavement, while the shoulderswere constructed using the asphalt pavement. Thewidths of inner and outer shoulders are 1.5 and 2.5m, respectively. The side slope is about 2 horizontalto 1 vertical (2:1) and covered with grass on bothsides of the slopes. Figure 2 illustrates a typicalcross-section of the test section.-5.50 5 10 15 20 25 30 35 40 45 50Figure 3. Blow counts (per 20cm) of Kunzelstab tests alongembankment slope (a light weight dynamic penetrometer, withweight of 10kg, and a falling height of 0.5metre)Pore-water pressure distribution in the slope wasmonitored at five locations along the slope cross sectionusing Kasetsart University (KU)-made tensiometersas installed by Jotisankasa et al. (2010)approach. Figure 4 shows the pore water pressureand total hydraulic head in October 2009, close tothe end of rainy season in that year. It can be seen700

that the effect of rain infiltration was to diminish allthe suction (resulting in positive pore water pressure)and the major direction of seepage was subverticaldownwards. The value of pore water pressurein August 2009 which was the middle of rainyseason (not shown here due to limited paper length)varied between -70 to 9 kPa. This trend clearlyshows the seasonal variation of pore water pressurein the pavement shoulder.almost reached zero. The vertical displacement duringshearing (not shown here) also indicated that thesamples were compressed by nearly the sameamount for all three vertical stresses.Shear Stress (kPa)403530252015Normal Stress = 62 kPaNormal Stress = 31 kPaNormal Stress = 16 kPa10500 2 4 6 8 10 12 14 16Cumulative Displacment (mm.)a)a)Shear Strength (kPa)706050403020Peak Shear StrengthPeak Shear Strength(Site T-1)Residual Shear Strength(Site T-1)Residual Shear Strengthc′ (kPa) φ′ (deg) c′ (kPa) φ′ (deg)3 26 2 1710b)Figure 4. Field observation results: a) Contour of pore waterpressure in Site T1 (kPa), and b) Total hydraulic head, duringOctober 2009, the arrow indicates the likely seepage direction3 MATERIAL PROPERTIES3.1 Fully saturated shear strengthIt is well known that for slope failure in high plasticitysoils, the average mobilized strength will be lyingbetween the peak and residual values due to theprogressive failure and shear softening phenomena(e.g. Skempton, 1985, Mesri & Shanien, 2003, Starket al., 2005). In this study, the consolidated draineddirect shear tests (with a shear rate of 0.005mm/min) were used to investigate both peak & “residual”shear strength envelopes of the fully saturatedsamples as shown in Figure 5. The undisturbedsamples were collected from the test pit at depth of1.5 m close to the pavement sholder, using a lineddriven tube with an inner diameter of 63mm. For residualstrength determination, the cut-plane reversalmethod was used in this study as a preliminary approach.It can be seen in Figure 5 that upon reaching theresidual state after the shear plane was formed, theshear strength dropped dramatically and its cohesionNormal Stress (kPa)b)Figure 5 Results from fully saturated direct shear tests (zero suction),a) Stress-displacement, b) Failure envelopeSr, %e10090807060504030201000.70.60.50.40.30.20.100 10 20 30 40 50 60 70wettingdrying0.1 1 10 100 1000 10000 100000 1E+06wettingdryingsuction, kPa00.1 1 10 100 1000 10000 100000 1E+06suction, kPaFigure 6 Soil-water characteristic and e-log suction curvea)b)701

laboratory tests as shown in Figure 6. The dryingpath was chosen since it was the evaporation processthat was to be simulated. The saturated permeability,k s , was determined for undisturbed sample (38mmdiameter) using the constant head method in a triaxialapparatus. As for the hydraulic properties ofpavement surface materials (i.e. base, subbase andsand cushion), the parameters were assumed basedon the particle size distribution of crushed rock subbaseas well as the void ratio observed in the fieldusing Arya & Paris (1982) approach. The permeabilityvariations with suction for all materials were estimatedbased on Jackson (1972) approach as shownin Figure 9b.from the drying test which corresponds to the inverseof slope in Figure 6a. It should also be notedthat for the suction range of zero to 2000 kPa, thesoil is still nearly fully saturated (Sr >90%) (seeFigure 6). Thus, the soil remained nearly saturatedfor most of the drying path obtained from the seepageanalysis. In such case, the principle of effectivestress and the expression H=E/(1-2ν) would stillhold true. The value of Young’s modulus, E is thencalculated based on this expression. Since no totalvertical stress change was imposed in this analysis,the conventional characterization of the Young’smodulus using triaxial tests was not relevant and notused here.Volumetric water contentq = 0q = 5E-5 m/daya)k (m/sec)0.40.35a)0.30.25Clay-Subgrade0.2Pavement - Crushed rock0.150.10.0500.01 0.1 1 10 100 1000 10000 100000Suction, kPa0.01 0.1 1 10 100 1000 10000 1000001.E-021.E-04Clay Subgrade1.E-061.E-08Pavement-Crushed rock1.E-101.E-12b)1.E-141.E-161.E-18Section-1b)Suction,kPaFigure 9. SWCC and Permeability functions used in the analysisA transient analysis was performed with upwardflux rate of 5E-5 m/day imposed on the slope sidebut not on the road top surface where no flow conditionwas prescribed. The initial pore water pressurecondition is hydrostatic but assuming a maximumsuction of 10kPa within the subgrade zone aboveground water. The finite element mesh and the resultantpore water pressure contour after 150 days ofevaporation are shown in Figure 10. The suction inclay subgrade increased to about 150 kPa while inthe pavement material the suction appeared to increaseto about 1,500 kPa due to the high contrast inthe SWCC of the two materials.Fredlund & Rahardjo (1993) proposed a simplifiedelastic equation for incremental stress-strain relationship,accounting for change in suction asshown in Equation 3 (Geo-slope international,2008). The modulus with respect to suction, H, usedhere was obtained from drying test results as shownin Figure 6a using the Equation 4. Figure 11 showsthe variation of secant H with suction as obtainedElevation, mDay zeroDay 150-2000 -1500 -1000 -500 0Pore water pressure, kPa13.512.511.510.510Figure 10. a) Finite Element Mesh; b) Pore-water pressure contour(unit in kPa) after 150 day of drying; c) Pore water pressuredistribution at the shoulder edge (Section-1) from numericalseepage simulation at day zero and day 150c)14131211703

6 CONCLUSIONThe use of high plasticity clay as subgrade materialbeneath pavement shoulder is one of the majorcauses of longitudinal crack problem commonlyfound in central Thailand. A comprehensive programmeof field investigation, laboratory tests andnumerical analyses was conducted, the results ofwhich have revealed the mechanism of such longitudinalcracks in a case of highway route no. 357, Supanburi.Field observation and seepage analysis suggestedthat the active zone of pore-water pressure changestook place along the shallow depth of about 0.5-1metre. The deformation analysis with pore waterpressure change from seepage analysis indicate theslope shrinkage and the zone of negative horizontalstress or crack zone being sub-parallel to the slopesurface. This crack zone was then considered to befilled with ground water in the Limit Equilibriumstability analysis. The resultant Factor of Safety wasfound to be about 1 in the case of mobilized strengthreaching the residual state.Based on the findings of this study, it will be ableto suggest the ways to preventing the shoulder crackproblem in the future such as a more precautious criterionof material selection for shoulder subgrade.This testing and analysis methodology can also beused in designing other ways to prevent extrememoisture fluctuation underneath the pavement in thefuture.Modeling with SIGMA/W 2007; Stability Modeling withSLOPE/W 2007. Fourth edition.Jackson, R.D. (1972). On the calculation of hydraulic conductivity.Soil Sci. Soc. Am. Proc. 36, 350-383.Jotisankasa, A., Porlila, W., Soralump, S., and Mairiang W.2007. Development of a low cost miniature tensiometer andits applications. Proc. 3rd Asian Conference on UnsaturatedSoils (Unsat-Asia 2007), Nanjing, China, 21-23 April1007: 475-480.Jotisankasa, A. and Mairaing, W. 2010. Suction-monitored directshear testing of residual soils from landslide-prone areas,Journal of Geotechnical and Geoenvironmental Engineering,ASCE, Vol. 136, No. 3:533-537.Mesri, G., and Shanien, M. 2003. “Residual Shear StrengthMobilized in First-Time Slope Failure,” Journal of Geotechnicaland Geoenvironmental Engineering, Vol.129,No.1, ASCE, 12-31 pp.Skempton, A.W. 1985. Residual strength in landslides, foldedstrata and the laboratory. Geotechnique 35, 1-18.Stark, T.D., Choi, H., and McCone, S. 2005.” Drained ShearStrength Parameter for Analysis of Landslides,” Journal ofGeotechnical and Geoenvironmental Engineering, Vol.131, No.5, ASCE, 575-588 pp.Taesiri, Y..Jittrikul, P, Boonkam, J, 2004, Remedial measuresfor highway shoulder damage. Internal seminar report. Bureauof Materials, Analysis and Inspection, Department ofHighways, Thailand. (in Thai)Take, W.A. & Bolton, M.D. 2004. Identification of seasonalslope behaviour mechanisms from centrifuge case studies.Advances in geotechnical engineering. The Skempton ConferenceVolume 2.Thomas TelfordACKNOWLEDGEMENTSThe authors acknowledge the financial support providedby the Bureau of Road Research and Development,Department of Highways, Thailand and theKasetsart University Research and Development Institute(KURDI). Geotechnical Engineering Researchand Development Center is thankfully acknowledgedfor its permission to use softwareSEEP/W, SIGMA/W and SLOPE/W 2007.REFERENCESArya, L. M., and J. F. Paris. 1981. A physicoempirical modelto predict the soil moisture characteristic from particle-sizedistribution and bulk density data. Soil Sci. Soc. Am. J.45:1023-1030.DOH (Bureau of Materials, Analysis and Inspection, Departmentof Highways, Thailand), 2007. Manual for inspectionand evaluation of road surface damages (in Thai)Luo, R. and Prozzi, J.A. 2008. Development of LongitudinalCracks on Pavement over Shrinking Subgrade. TransportationResearch Board Annual Meeting 2008 Paper #08-1034.Fredlund, D. G. & Rahardjo, H. 1993. Soil mechanics for unsaturatedsoils. New York: Wiley.Geo-slope international (2010). An Engineering methodology;Seepage Modeling with SEEP/W 2007; Stress-Deformation705