Geosynthetics - 7 ICG - Delmas, Goure & Girard (eds) © 2002 Swets & Zeitinger, Lisse ISBN 80 5809 5231 Use of geosynthetics in vibro stone columns V. SIVAKUMAR & D. MCKELVEY, The Queen's University of Belfast, Beltast, Northern Ireland. ABSTRACT. Vibro stone columns ate a common method of ground sabilsation, They improve the bearing capacity and rede the setlement of foundations. The overall performance of the technique depends on several aspects such a the suength of the coum material, treagth of he surrounding clay, area eplacement ratio and cokimn length. A eronger isi materia will provide inreaged Iateral confinement othe stone column. This paper examines the effect of placing a geo-gié around the column in order to further chance the lateral confinement and therefore improve the overall performance of the wbro foundation. As par of the ongoing {escatch, tranal teste wete carried out on reinforced and unzeinforced sud columns which had been installed in clay samples (200 ‘aun in length and 100 mm in diameter), The composite sandiclay specimens were loaded under undrained conditions and were subjected to two different loading conditions uniform and foundation type (concentrated) loading. The test results indicate that the performance ofthe composite material ean be improved by renforeing the sand column with agee-grid 1 INTRODUCTION In ecen times increasing land prices, the limited availablity of Inthe las the decades, the vibro ste column tecique has sites andthe government initaive to redevelop inner etes has fered the building. industry to lock for cheaper land for constuction, Ata reall constuction is now being extied out on sites which, due to poor ground conditions, would not previously have boon considered economic to develop. These types of land include brownfield sites, filled ground, land containing peavlorganic deposits or land containing recently deposited alluvium, In recent years various ground. improvement teckniquet have been eniployed in order fo articialy prove ‘he gil properties in there sites ‘The ‘primary requirement of any ground improvement technique isto improve the compressibility characteristics ofthe isi soil and consequently reduce the expected foundation settlements, The most common methods avalable include vibro stone columns, piling, preloading and sand dain etc. This paper J concerned with viro stone. columns. Unlike cohesionless flavey soit ate lagely unaffected by induced vibrations and go thei properties ae net significantly improved by compaction alone. However when colimine of densely compacted course granular backfill are installed during the compaction prosees there is 4 substantial improvement in the ‘eength and setement characteris ofthe 40 elay. Unlike piles, which taneer the surface load toa stronger underlying Satan, bypassing the Weak material, stone columns ublise the Toad carying properties ofthe surrounding material (Fg 1). ‘When load is applied, a stone column develope end bearing pressures and side friction stresses similar to ale. However {he stone column also bulges, so trequres lateral suppert from the “surrounding soil (Fig. 1b) thus forming a composite soilsone column system. The increase i the lateral stresses in ‘he clay further leads to consolidation and subsequent relaxation of srcete, Tie process continues unt equlintim i reached. ‘The soil confines the sone column andthe passive resstance cases the granular eolimn to reinforce the ground and improve Dearing capacity and stiffness, become an environmentally acceptable and econemsically viable alterative to other conventional “methods of round Improvement. A number of processes have been developed over the years for installing the columns thete include the Wet and dry topfeed methods and the dry bottom-feed method (MeKelvey & Sivakumar (2000), Hu (1995), Greenwood & Kirsch (984), Harksdale & Bachue (1983), Hughes & Witere (574) | « Fare | 4) Pie) Vir tone ams The hottom-feed process (Fig. 2) isthe most commonly used method in the UK since i is applicable to wide range of site Conditions and seil strengths. It is also unaffected by the 1145,igure 2 Botiom fed mvthd (a) Poker pentaton and () sane lage presence of ground water. The modified bottom-feed vibrator Incorporates an addtional temic pipe so that the granular satenial may be supplied internally tothe nose ofthe poke. The sone i then discharged and compacted in its while the poker femaine in the hole. This ensures stability” throughout ‘onstruction of the columas ‘Vibro stone columns are suitable for supporting lightweight structures such as low-rise housing an industrial warehouses ‘Thoy have been installed successfully in soils with undeained shear strength (c) a low ae 7 kPa althonghthe technique isnot ‘usualy recommeded where cis les than 15 kPa because ofthe low radial suppor provided to the columns. This paper investigates the effect of placing a geosynthetie material atound the one column in exder to increae this lateral support and ence improve the load eanying capac. In an attempt to understand and predict the behaviour of ground seinforeed by stone columns, many studies based on Physical modeling, mathematial analysis and fll eae testing have been earied out during the last thrce decades. Previous Iaberatary based research, cared oat by Hughes & Wathers (1974), Charles & Watt (1983), Bachus& Barksdale (1984) and Hu (1995), has been invaluable fo establishing an understanding of the behaviour of stone columns. However, while many of ‘hese earty studies were based onexpesimentseamied out in one- dimensional loading chambers. this paper reports the performance of a stone column under taxi Toading 2 EXPERIMENTAL WORK “The experimental work involved a series of triaxial tests on samples of kaolin clay in which columns of sand were installed, ‘The kaolin war mixed at 1.3 timer the lgsid limit and consolidated under a vertical pessute of 200 KPa ina one~ ‘dimensional losding chamber, 100 mm in diameter and 500 ram deep (Fig. 3)- When consolidation had complete, the pressure the eylinder was reduced to zst0 under undrained conditions “The top cover was removed anda elieal auger wat used to hore cout a 32 mm diameter ole waile te sample remained in the loading chamber. frozen sand column was then inserted nfo the empty borehole. The methods adopted for preparing the sand columns ae outlined below ‘Unreinforced cohumns: The sand, prepared ata water content of 18%, was compacted in ayers into a tube (a thin plastic sheet folled around the helical auger to give a 32 mi inner diameter) ‘The filled tube was then ef in a Reeze for 24 hours, Whe i was taken out the tube was eut along ite length and the sand column was removed. A photograph of the fezsn column is —— == igure 5, Onemensions losing chamber shown in Figure da. The sand was dyed blue ung waterproof Ink fo at 10 differentiate the ‘column of sand front the surrounding clay. Reinforced columas: For the reinforced columns, a sleeve rade ftom georyntbtie sheet was placed inside the plastic tube, Both wete then rolled around the helical auger o give a 32 ‘um inner diameter. The sand again prepared at a water content ‘of 18%, was compacted in layers info the tube. Aer feezng the plate tube was cut along its length to remove the froze, geosynthetecovered sand colin. A. photograph of the Feinfored column is shown in Figure 4, © Figur 6. roan snd columns) Uncinrced and () Renee ‘The average bulk density ofthe sand prepared using this method war 19 Mpim'. Two column lengths were considered (a) 200 sam, repestoing a fully penetrating column and (b) 120 mm, representing floating’ or partially peneteating column, ‘When the column had been isalled the composite clayand sample was semoved ffom the conslidation chamber and 2 200, 1146rm high, 100 mm diameter specitnen was prepated fr testing in the tail ce Inthe tnaxil cell he specimen was consolidated under 100 ‘a of effective confining pressure n each et fll saturation of ‘he specimen was enauted by applying a back pore water pressure. Duet the differences between 1-Dimensiona loading fnd isotopic compression inthe triaxial cell, thie procedse produced a material with an undrained shear strength of| Spproximately 27 KPa and an OCR of 1.4 Following consolidation in the tvaxil cell the specimens were loaded under undrained conditions in two diferent ways: (@) the entre surf standard CU test and (8) the conte ofthe san representing 4 foundation type loading (Fg. 5). {he foundation was 40 mim, Teste wore cared out on specimens having either a reinforced or an unreinferced column of sand, 120 nm or 200 ram in length. In addition, a single test was conducted on a lay sample which dd not havea sand column, a axial eu of about 4%, Inthe ease ofthe 200 mm long geo rd reinforced column the devatorstesscoatinued to increase Sta slow rate; this increase became insignificant at an axial, strain of approximately 8%. A complex form of stress-strain behaviour tas observed forthe specimens having unreinforced columns. The results showed a reduction in atength followed by recovery. This may have been due tothe difference in the ngth characteristics ofthe sand andthe clay. Tigute 7 shows the undraned shear strength plotted against column length for bot eeinferced and unseinfored columns ‘The undrained shear strength of the clay without sand columns was approximately 27 APs, ‘The preseace of the 120 mm ‘reinforced columa inreated the strength slighlly. Any futher Increase inthe sregth aller reinforcing te coun with a geo guid wat marginal Inthe case of the 200 mm column the Increase in the shear stength is appreximately 40% for the tunreinforced column, sing to 125% when the column was teinforeed, The amount of inerease in undrained shea stength i significant between the reinforeed and unveinforeed cola this improvement is ene due tothe confinement provided by the geovgnds (Unter ening (@Foundon ding Pique $. Lowding pes inna et 3. RESULTS AND DISCUSSION Figure 6 shows the stess-train characteristics of the composite clayisand speeimens subjected te uniform loading. The thick lines represen the characteristics ofthe epecinene which had the geogtid reinforced columns (Fig. 4b) while the thin lines epresent the specimens which had the uneinforeed columne (ig 4a) Allsamples were sheared under undesined conditions at arate of 4% per day. In all cases, except in the specimens ed with 200 mm long columns, the pore water pressure nples reached entical state this occurred at Asal ines Pigue 7. Reaoship erweca undnined sbew stags and coh ne Figure 8 shows the load:isplacement characteristics forthe foundation type loading In this cas, load was applied to the cent aea ofthe sample aly. It was applied under undzained ‘conditions and he axial displacement rte wae set at 47s per ay. The load carrying capacity of the foundation increased by 60% in the case ofthe 120 mm coli and it further increased when the column wa extended Yo 200mm. The overall create in the load bearing capacity is significant for the fly penetating column reinforced wath a gee-grid. This increase was in the fonder of 175% However, at observed previously. for the specimen subjected to uniform loading, the invease in stength ‘was marginal when the 120mm long column was ceinforced, Tigize 9 shows the cross section ofthe clayisand specimen afer lesting under the foundation type loading. The cokimn Tength was initally 120 mm. The photograph cleanly shows the Tatra displacement o "bulging" of the sand column beneath the foundation. The deformed shape of the column is in close fsgreement with the findings from previous lborstoy studios (ia (1995) and Hughes & Withers (1978) 4 CONCLUSIONS The performance of soft clay installed with columns of sand reinforced with geo-gride was examined under two diferent loading condition uniform loading and foundation loading. The results indicate thatthe overall performance ofthe seil can be tnereased by approximately 17 1147reinforced with geo-grids. The practicality of installing reinforced columns need careful consideration 11000 Uareinfoneed Resnlorced 00 < Cana eigit2000 600 pe Lox 400 200 ° 5 10 1S Displacement mam Figure 8. Loné-iapacement elation for fomdion ype Tending Toanduion pe lode 5. ACKNOWLEDGEMENTS Funding for the work was provided by PJ. Carey (Contractors) Limited, UK and Keller Ground Enginering, UK. 6 REFERENCES Bache, RC. & Backdale, RD. 1984, Vere and neal eho of Tol se columns. nt Sal and Rock Reborcement Pree Tor Conf Pan 184 Bale, RD. Bathas RC. 1983, Design aed construction of stone oluneFin Report SCEGIT-AK104, Federal Highoay Chae, TA & Wats, KS 1988, Compeosby of sof clay renee ‘wh goth cohinine In So Mechaner and” Foundation Enginerng: Proc Buropean Con, Hela, 34-352, ‘rcenmond, DA. and Kish K. 1986, Speci pot eaneat by ‘abnor and dmamie aod In Advances in Ping and Grownd Treuiment for Foundations, Proc. Int Conf. In. Ov. Brg london 1-5, Ha W" 1995 Phos modeling of goup bebaiour of sane coke omits. Uni of Glasgow PRD Thee Juve 195. Hugh 3900. and Withers NU TEM. Reainng of 08 obese ‘oll with soot colinns. In. Grownd. Enginsring. Foundation Pubcon Lime 42-08 Mekelvey, D Swabia, V.2000, A review of te performace of bo" sooe calm Youndatons. Ta Ground" Dnprovement Techiguer Proc It Conf, Sogapore September 2000, 285 se 1148,