You are on page 1of 8
The accuracy of two slope stability programs utilising the Bishop and Sarma analysis methods Alan Parrock INTRODUCTION lope stability considerations have long captured the attention of geo-practition- fr, elles (1936) Inlay Introduce hls socalled ‘ordinary method of slices’ (OMS) for the prediction of earth dam sta- bility. Some 20 years later, Bishop (1955) proposed a method that, despite an assumption of interslie forces, 1s stil used twit ucts success tty: Bichop's formulations, howovsr, ear from the constaint that intemal loads and/or extemal forces cannot theoretically be included in the computations without alter: ing the assumptions made in the solution for the factor of safety This facet imposes a major constraint when external andéor intemal loads are pres ent in realfe problems, eg loads atthe top ‘ofthe slope generated by vehicles on road or rails, anchors utilised for stability, 0” geosyn rete for retnfowwement ‘he paper presents nuts of analyses conducted whereby Sarma’s (1979) Method. and Sarma and Bhave's (1974) Rigorous Method as reported and catalogued by Hoek (1987) ae used as base eases for comparison. vt the results those generated by Bishop's Method suitably modified for inclu- sion of interal/external fores/loas. ‘The Prokon suite of geotechnical com- puter programs SLOPBG (1994) utilising Bishop's method of analysis and SLOPNC (2994 uaig she Sarma theory were used for the following evaluations INCORPORATING INTERNAL/EXTERNAL LOADS IN BISHOP’S ANALYSIS METHOD ‘Me formulations given tn appenans 3 dea the methods adapted in SLOPEG for incorpo rating interal/externa loads in Bishop's analysis metho. SARMA‘S ANALYSIS METHOD ‘The Sarma analysis method was described by ‘Morrison and Diktan (1995:76) as follows: ‘The complete multiple wedge method of tod termed Bato’ Spied Meth (85S. Jb rt 1980) ther Galop he method. Tae reece mana or TABR (1985) 2nd STABCM those rly pops by shop Al uber ee et the BUM or he BSH, have Bop igi mations as base a wl here be reed 95 ep Meth ‘calculation was discovered by Dr Sarma at Imperial College and fist published 15, years ago, Sarma (1979), Development, pro fessional aceeptance and commercial use of the method are long overdue. As the method of calculation is easy to under stand, numerically simple and less complex ‘than many of the approximate methods, "mere Is no practical reason fr lack of use. The method of calculation has been described in technical literature both by the authors and others, Hoek (1983). ‘As show by this review of com mercially avallable slope stability program, none of the user trenuy progrant impe ‘ment the complete multiple wedge method with seinforcement. SLOPNC uses Sara's analysis method. Any shape of failure surface may be analyse and Internal and externa loads txvxporated. NO assumptions are made regarding interslce forces and the solution equations satisfy both force and moment equilibrium. The formula: tions contained in appendix 2 detail the ‘methods as advocated by Sarma and used in SUOPNC. CORRECTNESS OF METHODS SLOPBG, utilising BM, usually generates fac tors of sitet that are conservative compared to predictions anae using other theoiies, SLOPNG, ulilning Sarmat Masthod, generat factors of safety that ate usually higher but : The BM usually predicts factors of safety that ate consistently lower (but not as iow as tHe Ordinary Method of Slices (OMS) than ‘ofthe other commonly wed analyte ‘methods. Ths conservatism is deemed not ‘only to be realistic but also to be advisable ‘where often limited testing is conducted and estimates have usually to be made of shear strength parameters that contol stabi. "The inital queetion that mutt be aehad and satisfactorily answered is: ‘Which analy- sis method generates absolutely correct ‘Tambe and Whitman (1969859362), 8 analysis whet bas remained ‘state of the oe evan though i wae advanced in 3 txt ‘ublished some 30 years ago, detail a graph «al analysis method which generates ‘The correct safety factor satisfying statis." This evaluation example, which incorporates a (JO slope 20 tet high, subject co a low net tnd toe drain, war analyred and shown ta fexhihit a factor of safety (FOS) of 1.27 ‘This same slope was input to SLOPNC using Sarma’s Method, and analysed to gen: erate an FOS as per figare 1 of 1.266, te exactly the same asthe Lambe and Whitmen Journal of the South African institution of Civil Engineering, 42(1) TECHNICAL NOTE Journal of the South “Atrioan Institution of Civil Engineering, 42(1), Pages 35-42, Note 7/48 ALAN PARROCK graovates University of Natalin 1973, cia Engineering intaty ‘empryec in ie natal Rosas Department he subsequanty moved to the National institute of Road Research at the pe regsterod 2s @ protessiona engineer In 1978 he joined Brunette Kruger Stooerg and Huge as resident ria and Ovarnbound, He the geotechnical depart Wot BKS in Johannesburg ane Offshore Design Sewvices forthe foundation design ofthe Mossaas 1999 Alan estabished ARO Assacatas with Raimund Mile and tha geotechnical ana project man agement diusiors of ARC Spocalst Frances 35 Table 2 Factors of safety and percentage aitrerance tor analysis by SLOPEG and ‘SLOPNC aie Semone tn —_—_ Methodiene [FOS or cena] FOS ete = Bishop’ [aiterence | soma eerie | Cone San a + ‘ Se Mattes [in 05" | Method Las ee # seh a pou be sleet Se Method {SLOPNC) SORORED AER: SLOPBG) Tacaica fess | vi? fae raueefuvis | wav prise vm rae fis | saves wet peo eco the Sarma Method (ane imi equirium mode of spe tare hed ety —— —— = (9 tna ess an Figure 1 Lambe and Whitman exemple analysed using SLOPNC doce by means Oe eh of cou be argued ha the exact equality Tio ioc and a conitions of sates ae sais eet ang various reabfestuations, analysed using the Sarma Method, have indicated It to be covet (Roure 1994), assumed thatthe Sarma analysis method used in SLOPNC generates correct, EVALUATION OF VARIOUS METHODS In comparing vatlous methods, th Towing facet should be noted. The fo Iations developed by Bishop contain F the factor of safety, on both sides of the ‘matical manipulation will generate an explicit expression for F in terms of the other variables. Thus an iterative proce: dure is necessary fr solution. The accura cy to whlch this iteration fs conducted usualy gene ts small diferences i answers advanced by various researchers. ‘This iteration was dificult in times of ‘manual manipolation, but with the advent of computers is easily accom. prise In order that comparisons between the RM as used in SLOPRG and the Sarma Method adopted in SLOPNC could be made, initially an example from Fredlund and Krahn (1977) (F+K) was revisited by ‘examining the factors of safety generated by various analysis methods fore particu lar slope. (Note the original Imperial sy tem of units was retained such that inac ccracies were eliminated.) Initial evaluation A theoretical 40 foot high. 1 (vertical = 2 (horizontal slope as per figure 2 was analysed 36 Figure 2 Example from Frediund a ‘The slope was characterised by the fo: lowing shear and density parameters = 600 pounesisquare foot o> 2 Y= 120 pounds/eubic foot ‘Table 1 details the factors of safery (TOS) predicted wilising various analyse meth ‘ods as reported by Fk the ditferences between SLOPBG and SLOPNC were farther explozed by taking the example cited by Fak and ind Krahn (a) locating a ine load of 3 000 pounds/foot atthe top of the slope {b) locating a line load of 600 pounds/foot at mic-heignt along tne siope (6) using the seme configuration as in by but with a UDL of 200 ounds/sauare foot located over the Table 1 Factors of safety for example problem utilising various methods of analysis Method? Ordinary | Bishops | Spencss | Janbus | janbus | Mogermier case Method. | Simplited | Method | Simplified | Rigorous | Price otslices | Method Method | Method} Method Avdetaies [x928 | 2000 [207s [aon [200s [acre infiguce 2 rigurez win [cor [izoe uror [arse | azo | i765 Figure2with|ie9s [isis [isso [ise ise [ies plezomeic SLOPBG evaluation The same theoretical problem as analysed by FeK was re-analysed using SLOPBG and SLOPNC. f Satma’s Method is assumed to generate ‘correct answers, factors of safety and the percentages that these factors generated by SLOPRG difer om those as given by SLOPNC are dletaled in table 2 shoulder break point (a) using the same configuration as in {6} but with a geofabric of 300 pounds/foot strength located at midsheght of the slope Table 3 detalls the percentage difference {the calelated FOS batonsen the two methods Joernaal van die Suid-Afrikaanse Instituut van Siviele Ingenieurswese, 42(1) Table 3 Percentage difference in the factors of eafety for analysie by shows the BM in a very ba light ‘SLOPBG and SLOPNC Actual differences between BM and Methodiease FOS for Bishop’ _ [Percentage | POS for Sarma’s the other socalled more correct methods Method (SLOPBG: | difference_| Method (SLOPNC) | YET Farely ff ever) exceed 6, A difer ence of 5 is deemed minimal within the context of slope sability factor of safely predictions Figure 2 witha line load of 1750 save [Tae 3.000 Ibs/ft at crest + piezo line Figure 2 with a line load of 2034 i268 [2060 The BM as utlised in SLOPBG gener 600 Ibs/tt along mid-length of slope ates ansvers that in the majority of cases, Figure 2 with a line load of 600 Tosi | 195 Tae PEO fare within. 5% ofthe true’ Sarma along mid-ength of slope + UDL of SLOPNC method. 200 Ibsft over top postions of slope SLOPBG Isa user-friendly, toelly Ts abowe with horizontal geofabrie | 1.956 sia2 PRR ‘written and inexpensive South Afican product. which. because of it innovative | ______]} mouse input method, typically generates at mic-height Further evaluations all tensile tess on any slice ace ave answers within ten minutes of start-up. ‘liminated. Tiss often difbeul to As the above comparisons show, there is In order to investigate further the die accomplish as any small aspeiis or {hus no reason why should not be use ences benween the evaluations Using abrupt aiscomanuraes tena 0 generate for slope stability analyse incorporating SLOPHG and SLOPNC the example of tensile frees. Although SLOPNC utes, intemal or external loads, Lambe and Whitman (1969:359-362) pre- via a method of linear dlscontnulis, etc situations ate to be analysed, ously detalled was analysed using reshaping of constituent sles, it isnot however, where the factor of safety is SLOPEG and SLOPNC with varius shear always posible to generate» non-nega- marginal (LO to 1.25) when using parameters and inctined loads, Table 4 tive normal force solution SLOPBG, SLOPNC is recommended ws the Gecas te resus ota. prefered anolys method. The use of SLOPNC Is aso tec: Table 4 Percentage ditterence in the factors of safety for analysis by SLOPBG and Seon ereemege citer tery ysis by mended where the prospective fale sur faces obviously not cela Mathod/case POS for Bishops [Percentage] FOS for Sarma’s Method (LOvEG) fatference | Metnoa sLOrNC) | References Tanibe and Whitman 969.559-362) Toa ToT T66 shop, AW 1955, The we ofthe sip cle in with peat surface and toe drain using the simy aatyss oF eat pes 12 slices: c= 90 lost, '= 32° eae, $717. Tame and Whitman(90.59-008) | 1243 Tha 1266 uncon, [M, Lamy BK Sehafoy 208. ‘with phreatic surface and toe drain using STABCGME 0 computer ream fr lope tity Saleen c Sotho ‘asia enfed embanete a ps Tame and Whitman(1969-359-362) 099 207 0920) Department of Civil Engineering. ‘with phreatic surface and toe drain using Fallen. W 1936 Caution ofthe stability IZslices: = 0 Ibs/tt, O32" fear dans. Transat of he 2nd Congress Tame and Whitman (1968:359-362) | LOIS ry Tas on large Dams, 4-45 Washington. with phireatic surface and toe drain using Freund, DG & Krahn, 1977. Comparison of Tastes c = Obsife, O32'plus an Slope stably methods af nae anchor of 1 800 Io lated a ni alope Conan Geena foun, 1429-429, installed at an angle of 22° Feurle 8 1994 Just how accurate ae your Slope tality calealtions? The Stablty of ‘Slopes lecture course presented by SAC Figure 3 details the SLOPBG analysis of CONCLUSION the same problem analysed using Farr ME 1966, Foundations fei si naljexterna loads in the Tum methods for slope sabilty analysis SAICE bility analysis methods on ™ po oane iak Rometodgreacsra istic lower bound sol Lambe, TW & Whitman, RV 1969. Soil tions to problems with _mecharics. Massachusetts Institute of Foure Lambe and Whitman example noniyand 000s using SLOPBG Morrison, IM Dik, $$ 1995, A review of “The Rak compas. Matar of soe nai ney Es rm Limitations of SLOPN ‘sons, although only on a very limited TS Engld ay wring of The pace of rations of SLOPNC scale, have actually taken one of the soll reinfor : In Europe’ oes by the The Sarma method of analysis as ulsed worst-case scenarios forthe generation of ‘RMA Grup un he aun of he 1m SLOPNC ts only complete and satisfies ‘corect answers’ using BM, as ths com: {Re fisttoion of Col Enginees oo Tay all imiting equilibrium conditions once paratively deep-seated slip surface usually 8, pp 677. Thoma lod Journal of the South African Institution of Civil Engineering, 42(1) 7 aon versus state factor of safety in saity Salas of ath dams and embankments Sama, SK 1979, Saiyan of mba nents and slopes ASCE Jounal af Getcha Engine, i081G{ 12) 1S1 T1824 SLOPBG 1994, Analysis of generalised sil Slope wing Bishops Modited eto Computer program by ARQ Ameen and Prokoe Software Consultants (Ti SLOPNC 1984. Non cra slope fare. {Computer progam oy a Ascites end Pook Software Convolsne (hy 13 STABGIM 1985.4 computer progam for iope Stality analy of einfored embankment and slopes, by] M Duncan Bk Low & & Shae igi te ca 2081, STANK 5, compe oa ope Mieocompater setsion by }M Duncan and Ka Sin wong second ection Virgins ech Bhcesburg VA 2061, Depurtnent of Ci Engineering. 38 APPENDIX 1: BISHOP'S FORMULATIONS AS USED IN SLOPBG SLOPBG - Analysis of @ generalised soi slope using Bishop’s Method Introduction SLOPBG isa slope stability computer pro- sg1am Which uses Bishop's Method (1955) ‘of analysis fr the evaluation ofthe sta bility of generalised soll slopes. The ratio ‘of mobilising and resisting moments on tnavidua sles ts usea 10 determine the factor of safety. The lope may be subject cd to line and uniformly distinted imposed loads, it may be reinforced with ‘metal stips or geofabic, it may be clad ‘vith for example a masonty block wall, The foree profile ofthe anchor reinforce. :ment may also be input ‘The slope may consist of materials with differing shear strength properties, efined ether in terms of shear strength, parameters © and 0 of an unerained shear strength profile wsing con its own, ‘Water pressures are accounted for using either a pheeaic line, pore pressure com: tours or a singe pore pressure co-oeffi- Cent (ry, Hotizontal and vertical seismic fore ae aso considered. Tho user is presented witha choice of searching for the eritcal minimuen factor ‘of safety (FOS min} circle ar he or she may input a user defined cicle. Once FOS, rin is located, the user is given the ‘opto of conducting a probabistic analysis by allowing parameters ©”, and the seismic co-ffcients a and, to take on ranges of values within the limits ‘of user defined statistical distributions subject to a user defined range of circle safety subject to the above constraint in the vicinity of FOS min i determine. Theory [A slope is vide into umber of tices and each slice i analysed as to the forces ‘which are acting on it. Implicit in Bishop's Method (BM) i that the inter slice frees ae horizontal. For the i* slice, the folowing formulation (from Duanean eta! 1988) i wed pt W= 0 (oose — sina) _ cose = damp ahaye Ny Where N'= the normal force atthe base of the sie, ane T= (61+ p'liandyF @ Where T’ = the shear strength atthe base of the slice W = welght of te slice 0’ = Mohr-Coutomb strength parameters 1& = pore water pressure a base ofslice = angle of inclination at base of slice 1 = length of ar at base of slice b= wid of slic By = side forcor ‘The factor of safety Fis given by p- Beb = W~ itandt EW sion 8) Where = oose (1 + tang’ tana/A) t) 1 should be noted that the BM ensures ‘hat vertical equilibria conditions are tified, but the formulation sulfors fom he fact that an iteration proce i necessary when evaluating the factor of safety using overall moment equilibrium, as Irom the above equations it can be seen that Fours on both sides ofthe squation, Reinforced slopes ‘when reinforcement is added to a slope, on additional resating force is introduce, ‘which coupled with the location of the reinforcement, introduces an additional resisting moment. Thus ifthe FOS of a slope is defined pee M, o where My = resisting moment and M, = overtuming moment, then the EOS ofa reinforced lope may be expressed as {y= Mo Fe where My ~ resisting moment of the reinforcement ‘This resisting moments generated bya fone and ever an te case of 9 seotete enon lope tis foe Tare sure whist or sya aehor the fore would tend to ein the dee tn ofthe anchor provided ge move tens do nat oct, er factor to ear in mind isthe theo the eas a eto his may be signa (ay 100. The designe shoud ote hs fat ad scale chy the utmatestengh ofthe et forcement to fete the smount of tore External loadings The following formulations adapted from Hare (1966) hove boon sed, Joernaal van die Suid-Afrikaanse Instituut van Siviele Ingenieurswese, 42(1) Line loads rom the figure below Using the same notations as previously, ie as pe the sketch above, we have fora UUDL of widths 2a, intensity, sncined at an angle y tothe cnfae, that 9 = woos y a The following geometric considerations ‘may be cenved D- eO, as) P= Deosy 06) x= Deny an sin 8 From the formulations of Har (1966) the following expressions may be derived for the normal, tangential and shea stresses ov Shee Ea }en Ea) aaa Sen ara) ee sal Grn aaraa| ora-y 2) Now using the Mohr-Coulomb failure condition ices we have * Baty «og ste 3) The normal component of the UDL thus lice n) generates ¢ mobilising force of —r4s Uniformly distributed load orhb D+ RF 0 +e emntdere cose 29) Two cases wil be cataktere oral Yo which ats on te hase ofthe slice under ‘ry surace, ond angential othe urea, Which.actson « _ (@) Normal component € (b} Tangential yr > component ° ‘ . oy In ths exe be vet 1 We, ae wsin « os) (19) pe i se As previously in equations 15 \ to 18 we have \ y bene ey R : De = OF \ 26) or = o,sin ® 03 \ 2 = Doosy @ The term of acts inthe direction fom . x= Dsny ry the centre of the sip ctl in question to Ne x the centre ofthe slice under consider \ tion, and thus doesnot prove any ad \ tional mobilizing or resting momeat The line Hoa acting ata suslace thus only 9) provide an additional mobilising force tn the sic in question equal to ay RD se as . ‘Journal of the South African Institution of Civil Engineering, 42(1) 39 Now from Har (1966:38-62) forthe nor- mal, tangential and shear stresses we have age 2? Hs 2? - a8}? + datz? (a0) 2) and as previously to transform the co- ordinate system oreey co [Now using the Mohr-Coulomb falute condition circles we have cae + 5g 82g 08) The tangential component of the UDL thus generates a mobilising force of oR cose 3) fh acts at the base ofthe slice under consideration Notes: + ‘The above formulations have beet undertaker to mode! the cefects of loads on the stabi of the system. I's realised that they probably violate the conditions imposed for inter- slice forces. However, the sects ofthis wolaton are probably small and will not alter materially the fact that the BM is in itself only an approximation for the actual situation, The imposed load are only Aemed to have an effect if they fall inside that portion of the soil mass which has the potential for failure, Seismic loading Seismic events induce loads on the struc ture which act against stability. Both ver tical and horizontal seismic forces are assumed to act through the centee of masc af exch dice Tension cracks In soils which exhibit a significant cohe sion () intercept itis posible thatthe soll may stand unsupported to some height. This height i the depth to which a tension crack may form inthe slope. The formulation forthe depth vf the tension crack Is given by Lambe and ‘Whitman (1968-342) as, 36) here € = the cohesion inrercepr in tat portion ofthe embankment 40 = L28ne! train an and y= The unit weight of the material Water in tension cracks ‘Water pressure is always assumed to act at right angles to the failure surface. APPENDIX 2: SARMA’S FORMULATIONS AS. USED IN SLOPNC SLOPNC - Non-circular slope failure Introduction SLOPNC sa generalised non-circular slope stability program which uses the pancsertcal slice method as proposed by Sarma (1979) for the prediction of the factor of safety of general shape surfaces As the boundaries ae non-verical,stuc- tural features such as faults or discontin sty planca may be ince. Water pre snes, external loadings and reinforce ment have been Included so as to make the analysis as generalised as possible. Theory Consider a generals! quadslateral with ina sol or rock mase a8 depicted in the figre below: The following geometric considerations as per Hoek (1983) may be derived: op © (Tg ~ XRF + Tay = YRS w By» ACSA ATi ~ XB, Muy ay 2 = XB, ~ XB, 3 a= Arcane, - YE, 1H = MBB YT, JO, XB) 97 ¥8, AAT, XB)) 2W,. = (YW, ~ YB.) 6) ‘Tne submergence contin dictates the force ating on the de oF he gust eral. The wplt water fore acting onthe ‘ase ofthe slice sin all cases given by Ufo yh YH, YO Ys ¥B. 9b, Boos, ‘Cave 4: No submergence of sie, 33 Aotaled in tho figure balons The side Torces are evaluated as PW, = Hern (YW,-YBYF Sec, PW = vw (YW, -¥8,)* Secs, (9) Case 2: Submergence of side | onl as detailed in the ligue beiow: The following forces as per Hoek (1983) may act on the soll or rock segment. Phrestic [surface YTSYW, YT..7 Wy aay The side forces are evaluated as PW, = Ry (QV, ¥T,-¥B)( YB)ICoss) ap PW. = 1IRrwy (YW)..-¥B, 1COs8,) 2) The vertical and horizontal forces applied to surface ofthe slice a a result of sub Joernaal van die Suid-Afrikaanse Instituut van Siviele Ingenieurswese, 42(1) merge of pat ofthe sc are given by WW, = Wey (PW, YTYOT, ATT s- YD) any WH, = 0B (OM YT? a case 3: Submergence of sie i+1 only, as, erated in me Aigure below YPM, YT. s<¥W “Te side forces are evaluated as as PW, = 1121 (OW ¥B)YCOS5) (16) we lsh \ The vertical and horizontal forces applied to surface ofthe slice as a result of sub. rergence of part ofthe slice are given by WA, = A229 Was YT a) Tis XT YD as) WH = Ary PW YTi)® a Case 4: Complete submergence of slice i as detailed in the figure below: ST AT KM @) PW, = 1124 (@YW,-YT,- YB) YT-YB)ICOS8) en PW = Mp YW -YB. YT YB, 06,,) 22) The vertical and horizontal forces applied to suctace ofthe sce as a result of SUD. mergence of part ofthe slice are given by WH = N21 OW TVW YT VAT 9 AT) (23) Water forces on the first and last slices Although the water force PWi acting on the first slice side andthe force PW, acting on the (n+) slice side (whith ‘ould be a tension crack) ate calculated by means ofthe equations listed above, these forces are not normally used in the caleslation ofthe eritical acceleration. ‘The simplest way to incorporate these forces into the analysis sto treat them as external forces With the following compo. 1Y,= PW, Sins, es) TH, = PW, Coss, (26) TY, © PM SI By, en TH, = Py, C08 8 28) Pore pressure contours Im the case where pore pressure contours ave used to charactriee the moisture ca dition, the resultant nett load distribution due tothe water regime is applied to the side ofthe slice in question. The figure below details the approacn adopted, { le face Geosynthetic loadings The presence ofa geosynthetic (9) at a slice houndary is handied as follows. f the gsis flexible, then the resistance gen- erated by the gs will tend to be in direc tion of the shear Boundary along movement, prior to falure, woul take place IFon the other hand the gs isis, lea very stiff gs ora sail nai, then the tensile force within the gs would actin the direction ofthe member, s ilustrated Am the sketch below. The user i provided with this choice i the prog Journal of the South African Institution of Civil Engineering, 42(0) Other external loadings These are assumed to act at the positions which they physlally occupy on the slice in question. These loads may for example be line loads or UDI’ or they ‘may be the position at whieh an anchor Icapplied. Their effet is evaluated by resolving into vertical and horizontal, components, Muttple layers The program posseses the capability of handling multiple layers within one sie. For the evaluation of the Cu parameter in the critical acceleration equations as pre sented hereafter, we have used the follow ing formulation forthe averae fictional component aver the slice in question, based on the figure given below: 2 Cosbs, 8 = cos} a + \_ yet 8 \ tamed 8 \ Caleulation of the critical acceleration he eitical aceleation required to Ding sng equ the slope toa condition ofl bum is given by +S, Sinibar te b.9)-S; Siare-6)) an 8S 8 9)-8; Sinton -8)) 3 "The factor of safety is calculated by rediicing the shear strength parameters simultaneously on ali sing surfaces ‘unl Ue zesullant acceleration a the en tre of mars of the slice Py» QWCoeKb5-«) co) 9, = Of Cael D> Cosbs/Costorei-B'se1-Bin) (35) S,= chd- Pitan 36) 41 Son» Chis y-PH, tends en Ry = cab, |Cose,-Usandy (38) ‘equa tothe resultant acceleration speci- fied. This is done by progressively reduc ing the shear strength parameters in the ‘equations above by ch Gal F o Oy + Cyl F (20) Cas = Cay LF yy bq = Tan (tan ba! F (42) = Tan"an b01 A 3) Oy = TaN GaN Og) When the value of K fora given factor of safety has been found, the fores acting fm the sides and base of each slice are 42 found by progressive solution ofthe fol lowing equations, starting from the vom conto it = 8 Ey 8B K + Ep, as) (E PW) Tan's + cd, 46) MO Mt MOO, 8°66, an ‘The affective normal stresses ating acront the base and sides ofthe slice are cae Iated thus: +U Tang ySine,-cqb Tans) Cos}e) Coeds. =) as) TS, = (N-U)Tanbyrcmb/Cosa, (49) am = (N-U) Coss, | b, (501 05 = (EPUB, on) Pans © Ei PM VO, (52) In the above equations, the effective nor. ‘mal stresses must be postive forthe solu- spon to be vata, Asa final check, Seema (1979) recom mend that moment equilibrium cond. tions must be satisfied for each slice. Referring tothe figure atthe start ofthis ‘addendum and taking moments about te tower tet nana comer, we nave: NIX. bLos{a,*8,,)Cose,-EZ, 3) “Elz ."P SiG /8,.C086] 54) -WUXG,X)sK.WIYG-Yq)-TVIX- 8) 455) sTHY;-¥0)-0 66 tartng rom ane tice atthe toe of ne slope, where 2, = 8, assuming «veka of 1, the moment are Za be calculated of vce versa The values of Z and 2.1 should le within the slice boundary, preferably inthe mide tir Joernaal van die Suid-Afrikaanse Instituut van Siviele Ingenieurswese, 42/1)

You might also like