Design Procedures for Slide Suppressor Walls

Transcription

1 TRANSPORTATION RESEARCH RECORD Design Procedres for Slide Sppressor Walls W M ISENHOWER, s G WRIGHT, AND M K KAYYAL A slide sppre sor wau is a retaining wall that is embedded in a slope lhal bas failed Slide sppre sor wall are sed to repair hallow slope fa ilres in areas where right-of-way is restricted and tlle slope cannot be flatlened The design proc dre for slide sppressor wajjs as mcs that earth pressre acting on the wall is eqal lo a hydrostatic pres re of a fl id with a density eqal lo the total nit weight of soil The performance of' the spporting drilled shas and load-carryig capacity of the wall panels were evalated for a range of wall geometrics Design charts for wall spported by 18-in and 24-in shas are presented A cost stdy fond that Ude sppressor walls cost abot $1 to $18/ 1 and are more economical than conventional earth-retaining strctres This paper presents a procedre developed by the Texas Siate Department of Highway and Pblic Tran portation (SDHPT) for the de ign of " lide sppressor ' waus Thi design procedre is based on a simplified method for estimating the magnitde of earth pre Sres acting on a wall The method ed to estimate the earth pres res acting on a slide ppressor wall is reviewed This is followed by a di c sio n of (a) the analysis of the wall slab and spporting drilled has, (b) the election of the size of wa ll slab, and (c) spacing between the drilled sha s Finally, a comparison is made between the costs of slide sppressor walls and conventional retaining walls BACKGROUND The stability of embankments and ct slopes in highly plastic clay has b een a contining mainte nance problem in Texas Many reason for the problem exi t; however, only a few can be effectively addres ed by state officials In areas where slope maintenance costs are high, two case of the problem 'tand ot One is the poor qality of soil encontered Texas has la rge expanses of highly plastic, expansive clays that have low drained friction angle (commonly between 12 degrees and 2 degrees) and negligible effective cohesion A second case is the inflence of the weather o n the poor-qality soils across the state A soil deposit might case significant problems in areas where the winter season brings high rainfall and many freeze-thaw cycles, and case fewer problems in areas with a more moderate climate The areas with the greatest problems have the following featres in common: W M Isenhower, Highway Design Division D-8PD, Texa State D p a rtm e nt of Highway and Pblic Transporioo, 11th and Brzos Astin, Tex 7 71 S G Wright and M K K11yyal, T he Umversity of Texas al Astin Astin, Tex Expansive clays, Wet winters with many freeze-thaw cycles, Dry smmers with little rain, and Moderate (3 horizontal to 1 vertical) to steep slopes The variation in weather over a year cases the clays to expand when wet and to crack wh n dry With later cycle of wetting and drying, the zone of cracki ng and weathering extend deeper into the fill or ct Abot 1 to 3 year ae r constrction the clay ha " decompacted," becoming loo e and having a low shear strength Later, sally dring a period of continal rainfall in early winter, the clay becomes satrated and a shallow md-flow, face-failre occrs on the slope While many slides have occrred in slopes steeper than 3 horizontal to 1 vertical, many slides have also occrred in slopes that are close lo this teepne s This is srprising ince a l pe stability analy i can show th at a slope in " n rmal oil," with the expected range in shearing properties, hold have an adeqate factor of safety against failre Stdi s of slope fail res have fond that th hear trength vale backcalclated from lope failres are I we r than tho e measred on laboratory-prepared samples of the same soils (1,2) In addition, many failres do not extend beyond the toe and crest of the slope This type of failre geometry is characteristic of soils with low cohesion vales The occrrence of the failres is direct proof that the shearing properties of the soils have altered, with effective cohesion approaching zero and effective friction angles increasing somewhat, and that weathering has an effect that is dependent on the prevailing climatic conditions If an adeqate amont of right-of-way is available, the most economical way to stabilize slope failres is to fl atten the lope In area whe re restricted right-of-way prevent flattening of the slope, Texas SDHPT ha d lide ppres or walls to remedy slope fa ilres A lid ppre sor wall i a re taining wall bried in the lope Typically the lid ppre or wall is located at one-third of the slope height as illstrated in Figre 1 Becase it i embedded, the wall can mobilize the sliding re istance of the downslope soils and need onl y to add enogh additional resistance against sliding to provide stability The embedment of the wall a llow rhe wall to be lighter in section than a conventional retaining wall and thereby less expensive to constrct Slide sppressor walls have several advantages over conventional retaining strctres The most important advantage is cost Slide sppressor walls can be bilt for abot 5 to 6 percent of the co t of a conventional retaining strctre A second advantage is that it is possible to con trct the walls qickly by sing prefabricated wall sections A third advan-

2 16 TRANSPORTATION RESEARCH RECORD 1242 Drilled shas \ / Precast concrete /,_- panels Drilled sha _ FIGURE 1 Typical slide sppressor wall sed for slope repair: top, plan view; bottom, profile (cross section) tage i that aer wall constrction, the slope geometry closely re embles that of the original slope and access is nre tricted for mowing and other maintenance The design procedre presented in this paper is intended for se by personn I who are not g technical specialists These personnel will not have condcted a geotechnical investigation beyond determining the depth of slide and determining the Atterberg limits of the soils in the slope As a reslt, several conservative assmptions related to the earth pressre calclations and oil propertie have been made The prpose for the di cssion of the method sed to develop the design procedre is to gide Lhe reader in de ig11ing lide sppressor walls whenever more detailed information is available CALCULATION OF EARTH PRESSURES The development of the method sed to calclate the earth pre res acting on a slide sppre or wa ll is di c sed by the athors in a companion paper in thi Rec rd The eanh pressres acting on walls installed in slopes that have failed can be calclated, for all practical prposes, assming that the soil i cohesionles For a jst-stable (i e, factor of afety = 1), co he ion less oil, the angle of internal frictioll is eqal to the I pe angle The companion paper fond that the earth pressre coefficient will be within 2 percent of nity, and will be within 1 percent of nity for slope angles of less than 2 degrees These vales sggest that the earth pressres acting on a slide sppressor wall are nearly hydrostatic For prposes of wall design in marginally stable slopes, the earth pressre coefficient of nity was assmed Ths, the earth pressre per nit width (P) acting on a slide sppressor wall bilt at one-third of the slope height is compted from 1 p = --yh2 2 where -y is the total nit weight of the soil and h is the depth of embedment The depth of embedment is selected on consideration of the depth of the slide observed in the field Typically, a wall hpight i st q1-!a!! the depth cf lide y!t:3 :lbgt 1 If the depth of slide is deep and not a shallow failre, a general lope fail re is likely, and a slide sppressor wall may be inadeqate for repai1 WALL PANELS The wall panels of the slide sppressor walls are constrcted sing prestressed concrete slabs The details for these slabs (1)

3 Isenhower et al were selected from the solid flat slabs presented in the PC! Design Handbook (3) Solid flat slabs with thicknesses of 4 in, 6 in, and 8 in were considered In determining the loadcarry capacity of a slab, the safe sperimposed service loading on a slab was increased over the vale shown in the handbook by the amont of the dead load for each slab The reason for this increase is that the wall slabs will be bried and will not have to spport the dead loads of 5 psf for a 4-in slab, 75 psf for a 6-in slab, and 1 psf for an 8-in slab The allowable bending moments in the slabs were calclated from the safe sperimposed service load pls dead load vales for each combination of slab length and slab height The maximm bending moment was calclated sing where the maximm bending moment will exceed the capacity of the strongest 8-in slab The information in Tables 1 and 2 was combined to determine the slab thickness and reinforcement strand pattern designation for several combinations of panel length and height Tables 3 throgh 5 show the reinforcement strand pattern designations for the varios combinations of panel length and panel height for each wall thickness The reinforcement strand pattern designation is a description of the amont and size of reinforcement in the slab The tens digit designates the nmber of strands per foot width of slab The ones digit designates the diameter of the strand in 16ths of an inch The sffix S designates that the strands are 17 wl2 8 (2) TABLE 3 SOLID FLAT SLAB STRAND PATTERN DESIGNATION FOR 4-IN WALL PANELS (TYPE FS4) where w is the average load per nit area and l is the length between spporting drilled shas A smmary of the allowable bending moments calclated for 4-, 6-, and 8-in-thick slabs is shown in Table 1 The maximm bending moments in the wall slabs are dependent on the magnitde of the average earth pressre acting on the wall and the span between the spporting drilled shas The maximm bending moments in the wall were calclated sing the assmption that the soil acts as a flid with a density eqal to the total nit weight of 125 pcf The maximm bending moments for different size wall panels are shown in Table 2 No moments are shown in Table 2 for the cases TABLE 1 ALLOWABLE BENDING MOMENTS IN WALL PANELS Strand Pattern Desionation Allowable Bending Moment" -lb Wall Thickness in Inches S S S S S a from safe sperimposed service load pls dead load, f' c = 5 psi, low-relaxation strand TABLE 2 MAXIMUM BENDING MOMENTS IN WALL PANELS PER WALL HEIGHT Lenqth Maximm Bending Moments -lb Length Strand Pattern Designation S 66-S 58-S 68-S 1 66-S 58-s 68-S S 58-s S S TABLE 4 SOLID FLAT SLAB STRAND PATTERN DESIGNATION FOR 6-IN WALL PANELS (TYPE FS6) Length Strand Pattern Designation S 76-S 76-S 1 66-S 76-S S 76-S S 76-S S S s TABLE 5 SOLID FLAT SLAB STRAND PATTERN DESIGNATION FOR 8-IN WALL PANELS (TYPE FS8) Length Strand Pattern Designation S 1 66-S 66-s 66-S S 66-S 66-S 76-S S 66-S 76-S 68-S S 66-S 76-S 78-S S 66-S 76-S 78-S S 66-S 78-S S 66-S 58-S S 76-S S 68-S S 2 76-S 1

4 18 straight in the slab The strands have a 15-in of cover on the tension side of the slab and are made of low-relaxation steel The compressive strength of the concrete is 5, psi ANALYSIS OF DRILLED SHAFT BEHAVIOR The objective of the analysis of drilled sha behavior was to establish a relationship between the depth of sha reqired to spport the wall adeqately given the height of wall and pressre acting on the wall This relationship is sed in the cost analysis to select the depths of shas reqired to spport wall panels of varios lengths The analysis of the spporting drilled shas took two steps The first step was to evalate the nonlinear bending stiffness and maximm allowable bending moment for the drilled shas being considered The second step was to analyze the performance of drilled shas sbjected to a distribted lateral load that modeled the loading on a slide sppressor wall Calclation of the nonlinear bending stiffnesses of the varios size of shas sed in the stdy were made sing STIFFl ( 4) Analyses of the laterally loaded drilled shas were made sing LPILEl (5) Bending Stiffness of Shas The analysis of laterally loaded drilled shas reqires vales for moment of inertia and modls of elasticity to calclate the bending stiffness of the shas For steel piles, the modls of elasticity is well known and the moments of inertia for standard shapes can be fond in steel design handbooks In contrast, no comparable reference is available for concrete shas becase of the wide variation in material properties and strctral details The bending stiffness of a drilled sha depends on the compressive strength of the concrete, yield strength of the reinforcing steel, the arrangement of reinforcement, the combination of axial load and bending moment, and whether the section is cracked in the zone of tension STIFFl calclates the bending stiffness taking the above featres into accont The nonlinear bending stiffnesses for 18-in and 24-in drilled shas are shown as a fnction of maximm concrete strain in Figres 2 and 3 These shas have a compressive strength of a_ 'f t A 11 ow ob l '1 Moment for LJgs 1 gn!! '"" 5 D 2 EI for Design o 6 12 o o 3 Mox I mm Stro l n - J nchgs/ Inch FIGURE 2 Eighteen-in drilled shas with 1 percent 6-ksi reinforced steel E l o " :,_ t g' " cil TRANSPORTATION RESEARCH RECORD Moment 2 48 a_ ' N " g 15 Al lowoblc;;i Moment f"or Qgs19n 36 t 1 24 " cil SOD --'"-'- -- a o o 3 Maximm Strain - lnc:tigs/lnch FIGURE 3 Twenty-for-in drilled shas with 1 percent 6-ksi reinforced steel El g a_ ' t concrete eqal to 3, psi and have steel reinforcement eqal to 1 percent of the gross area The yield strength of the reinforcing steel is 6 ksi The clear cover over the reinforcing steel is 3 in The arrangement of the reinforcement is shown in Figres 2 and 3 The vale of bending stiffness sed in the lateral loading analyses was the vale corresponding to the maximm allowable bending moment The maximm allowable bending moment was at a maximm strain of 3 infin, redced by a strength redction factor of 9, divided by a load factor of 14 for dead load, and with the live loading component considered to be zero The maximm allowable bending moment, bending stiffness, moment of inertia, and eqivalent modls of elasticity for the shas considered are shown in Table 6 Lateral Loading Analyses The analysis of the performance of drilled shas sbjected to lateral loading was made sing LPILEl This program can analyze a sha sbjected to general pile-head loading (shear force, axial loading, and bending moment) and distribted lateral loading over a selected section of the sha This program is also capable of sing soils data to generate the lateral load-transfer crves (p-y crves) sed in the analysis The objective for the analysis of drilled sha behavior was to establish the general relationship between the depth of sha reqired to spport a wall of a given height and the level of lateral loading acting on the drilled sha The prpose was to determine if it is possible to overload the drilled shas by sing wall panels that are too long, thereby exerting high levels of lateral loading on the drilled shas The procedre TABLE 6 STRUCTURAL PROPERTIES OF DRILLED SHAFfS Sha Diameter inches Allowable Bending Moment of Moment Stiffness Inertia in-lb lb-in 2 in"* * Calclated from gross area of sha, E = EI/I l2 Modls of Elasticity psi** " m

5 Isenhower et al was to analyze a drilled sha sbjected to for or five levels of lateral loading distribted over the pper section of the sha where the wall is spported In these analyses, the depth of slide and the wall height were assmed to be eqal For each level of loading that did not overstress the sha, the depth of sha reqired to carry that level of loading was eqal to the depth of th second inflection point of defl ection Figre 4 shows the reslts of one sch analysis on a drilled sha spporting a 5--high wall Figre 4 demonstrate that both bending moment and lateral deflection become negligible below the depth of the second inflection point of deflection Aer the reqired sha depths were obtained for several loadings on wall heights varying from 3 to 1, a least-sqares crve was fit throgh the data This related sha depth to wall height and level of distribted lateral load for all analyses in which the shas were not overstressed This relation was developed to allow the ser to estimate the sha depths reqired to spport walls sbjected to loads smaller than the maximm sha capacity The procedre presented above was made for 18-in and 24-in diameter drilled shas constrcted with concrete with 3, psi compressive strength and a 3-in cover over 1 percent 6-ksi steel Larger amonts of reinforcement were examined initially bt were fond to have negligible inflence based on the criteria of a sha depth with two inflection points of deflection In all analyses the Matlock criteria (6) for p-y crves for pile in so clay was sed with an ndrained shear strength vale of 1, psf The so clay p-y criteria was selected instead of criteria for stiff clay so that the initial slope of the p-y crve cold model long-term loading coocliti ns Broms (7) has sggested that the long-term increa in ha deflection de to consolidation and creep of oi l may be calclated by assming sbgrade reactions that are one-half to one-qarter of the initial vales for static loading Undrained shearing strength of highly plastic clay fills typically range from abot 2, to 2,2 psf By sing the so clay p-y criteria and an ndrained shearing strength of 1, psf one may obtain a p-y crve with an initial slope that is abot 43 percent (1/234) of the initial lope of a p-y crve calclated ing the stiff clay above the water table criteria The analysis of the drilled shas prodced two reslts The first reslt is the limiting level of distribted load for each wall height (assmed to be eqal to the depth of slide) below which loading mst be kept to avoid overstressing the drilled shas The second reslt is a relationship for the sha depth reqired for two inflection points of deflection as a fnction of slide depth and distribted load for distribted loads below the maximm allowable load on a sha The maximm allowable distribted loads as a fnction of depth of slide for 18- and 24-in shas are shown in Figre 5 None of the wall panels shown in Tables 3 throgh 5 is strong enogh to carry a load eqal to the load of the limiting soil pressres shown in Figre 5 The loading on the shas, therefore, will always be less than the loading shown in Figre 5, and the depth of the shas can be redced accordingly The corresponding sha depth as fnctions of slide depth and panel horizontal length per sha are shown in Figres 6 and 7 A relationship relating sha depth as a fnction of wall height and level of distribted lateral load was determined for the following cost analysis This relationship is sed to determine the depth of shas where the loading on the sha is limited by the moment-carrying capacity of the wall panels This relationship was developed sing the mltiple regression analysis featre of Lots For an 18-in drilled sha, the total depth of sha (L ), inclding the length of the spporting section behind the wall, is L(in) = D() P(Jb/in) (3) For 24-in shas, the total sha depth is L(in) = D() P(Jb/in) (4) where D is the depth of slide in feet (assmed eqal to the height of wall) and P is the reslting wall loading per nit length over the loaded section Eqations 3 and 4 are accrate to pls or mins 8 in However, when the conservative assmption of an earth pressre coefficient of nity is considered, Eqations 3 and 4 can be sed with little error The two shas at the ends of the wall carry one-half of the load of the interior shas When the depths of the two end shas are estimated, they are sally fond to be within abot 2 of the depth of the interior shas 19 l2 6 6 c 12 c 12 s s c ' J, L L 1 D 8 6 n nch Shcf' ts L8-1nch Shos 3, 3 _, -l O&f"lgctlon - lnctigs Mom12nt - In- Lbs 1' s) 2 3 I! g lo Woll Hetght - Fegt FIGURE 4 Eighteen-in shas spporting 5--high wall FIGURE S Allowable soil pressres acting on wall

6 2 TRANSPORTATION RESEARCH RECORD Wall Height - Feet Nmbers denote pone l w 1 dths 1 n feet II ID COST ANALYSIS Description of Procedre The procedre sed to evalate the most economical layot for rhe drilled shas and wall slabs for a given wall height is the following:, IB 21 LL, UJ g 19 ({) c c I 17 _ c, m 15 _J FIGURE 6 Design chart for 18-in shas Wa 11 Th 1 cknciss nch 6-1nch 8-1nch 1 Figre 6 or Figre 7 is conslted to select a slab length and sha depth 2 Table 3, Table 4, or Table 5 is conslted to select the appropriate wall slab 3 The cost per nit area of the wall is evalated from the total costs of drilled shas pls wall panels divided by the total area of the wall 4 The first three steps are repeated for varios lengths of slab and the nit costs of wall are compared 5 The arrangement with the lowest nit cost is selected This procedre allows the designer to adjst the final details of the design to reflect the costs of the separate components in a slide ppressor wall For example, costs of drilled shas vary by location As a reslt, in some areas, walls sing a smaller nmber of longer shas may be more economical than walls sing a larger nmber of shorter shas Another sorce of variation is the relative cost of varios diameters of shas to carry a given lateral loading In general, the cost of a slide sppressor wall is largely determined by the cost of the drilled shas sed to spport the wall, with the cost of the slabs having a secondary inflence Wal I H"1ght - Feet , Qj 23 Qj LL, UJ _ c 21 ({) c c I 't N 19 _ c, m c 17 3 Nmbers denote panel Widths 1n f'c;;at II 1 Wall Th1ckngss -B- 4-inc:h nch 13 FIGURE 7 Design chart for 24-in shas 8-1nch Ranges for Unit Costs In Texas, the costs of drilled shas and precast concrete slabs are calclated on a per-nit-length basis As an example, the average costs of drilled shas are $24/ for 18-in drilled shas and $45/ for 24-in drilled shas The nit prices for the precast concrete slabs are also qoted by the linear foot, bt sally average abot $2/yd 3 in cost when: calclated on a per-nit-volme basis By sing the cbic yard figre, the designer can rapidly estimate the costs for 4-in, 6-in, and 8- in slabs of varios lengths and heights The above figres were sed to calclate the nit costs for walls spported by 18- and 24-in drilled shas These costs are shown in Tables 7 and 8 The nit costs shown in these tables compare favorably with nit costs for conventional retaining strctres Walls sp- TABLE 7 COST AND SIZE INFORMATION FOR WALL SUPPORTED BY 18-IN SHAFTS Wall Sha Sha Wall Unit I Height Spacing Length Thickness Cost feet inches $/' $ $ $ $ $ $ $ $1 7

7 Isenhower et al TABLE 8 COST AND SIZE INFORMATION FOR WALL SUPPORTED BY 24-IN SHAFTS Wall Sha Sha Wall Unit Height Spacing Length Thickness Cost feet inches $/' $ $ a $ $ $ $15 ll $ ll $1537 ported by 18-in drilled shas ranged in cost from $98/ 2 to $147 /2, and walls spported by 24-in drilled shas ranged from $1511/ 2 to $1769/ 2 for this example Cantilever retaining walls oen cost abot $3/ 2 to $35/ 2 Mechanically stabilized earth walls typically cost abot $25/ 2 to $3/2 Crib-lock retaining walls cost abot $12/ 2 to $15/2 In consideration of these costs, slide sppressor walls can be a practical alternative to conventional retaining strctres CONSTRUCTION CONSIDERATIONS Slide sppressor walls are practical in areas where right-ofway is restricted and the slide srface is not deep-seated As a reslt, the designer shold be aware of possible restrictions on constrction Restrictions commonly encontered are restricted access to the site and clearance problems with overhead electric lines Other considerations are the liing stresses in the wall panels dring placement and orientation so that the panels are facing the correct direction These restrictions shold be considered and proper gidance shold be given contractors as necessary A second consideration is the stability of the shallow soils above the wall Dring constrction of the wall, two precations can be taken to improve the stability of the shallow soils above the wall First, by providing a drain on the p-hill side of the wall-sing a geocomposite drain or backfilling arond the wall with a freely-draining soil-the pore pressres in satrated soils can be redced A second precation is to replace the srficial soils with soil of lower plasticity Soil of this type is less ssceptible to the effects of weathering than the highly plastic clays A third consideration is corrosion protection of the ends of the prestressing cables in the wall panels It is common to cast several solid flat slab panels end-to-end on a casting table and then saw the individal panels apart aer the concrete has set This process leaves the prestressing strands exposed and sbject to corrosion Precations shold be taken to protect the exposed strands One method is to place an expanded foam spacer arond the prestressing strand at the location of the saw ct before concrete is placed Aer the saw ct has been made, the spacer is removed, any excess strand is ct off, and the remaining void is filled with epoxy cement Finally, the designer shold consider the magnitde of the bearing stresses on the wall panel at the point of spport If bearing stresses are excessive, a bearing pad shold be inclded in the design SUMMARY Slide sppressor walls can be sed to repair slope failres in areas where right-of-way is limited The procedre developed by the Texas State Department of Highways and Pblic Transportation for design of slide sppressor walls was presented This design procedre was based on a simplified method of calclating earth pressres acting on a wall embedded in a slope that has failed and was developed considering the behavior of the spporting drilled shas and the load-carrying capacity of standardized prestre ed flat slab The reqired depth of 18-in and 24-in drilled shas to spport wall panels of varios lengths are smmarized in Figres 6 and 7, and the reinforcing strand pattern designations for 4-, 6-, and 8-in-thick walls are smmarized in Tables 3, 4, and 5 An analysis of the estimated costs fond that slide sppressor walls can be cheaper to bild than conventional retaining strctres ACKNOWLEDGMENT The athors want to recognize Bobby L Myers, District Engineer of SDHPT District 1, Paris, Texas, for his encoragement of the work docmented in this paper REFERENCES 1 T G Abram, and S G Wright A S11rvey of Earth Slope Failres and R11media/ Meamres in Texas, Research Report 161-l, Project , enter for Highway Research, The Univer ity oftcxa at Astin, P A Staffer, and S G Wright An Examin a1io11 of Earth Slope Failres in Texas Research Report 353-3F, Project , Center for Transportation Research, The University or Texas at Astill, PC/ Design Handbook Prestressed Concrete Institte, Chicago, S T Wang, and L C Reese Doc111e111atio11 of Compter Program STJFFJ Enso, Inc, Astin, Tex, L C Reese Docmentation of Compter Program LP/LEI Enso, Inc, Astin, Tex, H Matlock Correlations for Design of Laterally Loaded Piles in Sort Clay Proc, 2ntl Annal Offshore Technology Conference, Hoston, 197, Vol 1, pp B B Broms Lateral Resistance of Piles in Cohesive Soil Jornal of the Soil Mechanics and Fondations Division, ASCE, Vol 9, No SM2, Mar, 1964, pp