You are on page 1of 154
a review of diaphragm walls @ discussion of ‘Diaphragm walls and anchorages’ Published by the Institution of Civil Engineers INSTITUTION OF CIVIL ENGINEERS LONDON 1977 ‘mis volume is the record of a seminar, held on 9-10 September, 1976, to discuss ‘Diaphragm walls and anchorages’ published in 1975 by The Opening Institution of Civil Engineers * 1 General First published 1977 diaphre Reprinted 1978 Discuss ) IT Current conditi anchors Discuss III Practic walls = Discuss 1 WV Practic W Oster Discus: ' Vv Design coneret © ‘he Institution of Civil Engineers, 1977 Ali rights, including translation, reserved, Except for fair copying, no Discuss mare et this publication may be reproduced, stored in a retrieval systen, pert cnsal ited in any form or by any neans electronic, sechanical, photo~ VI Grouna copying, recording or otherwise, without the prior written permission of Procedi the Institution of Civil Engineers 7 iscus: sme Institution of Civil Engineers as a body does not accept responsibi- ree ne oe Ate mae Gr the opinions ewprecsed in the following | WIT Slurzy Pages | AL Lit published and distributed by Thomas Telford Ltd for the Institution of ' > seus: Civil Engineers, 26-34 Old Street, London ECIV 9AD VIII Generai ISBN: 0 7277 0045 6 a Append: wx, 1976, 5 by The copying, no leval system, nical, photo- srmission of responsibi~ » following Ltution of un wv VI vi vila Contents Opening address. M W Leonard General analysis of available methods for designing diaphragm walls and anchored walls. W G K Fleming Discussion Current design practice applied in differing ground conditions for diaphragm walls, contiguous piles and anchorages. B 0 Corbett Discussion Practice in the detail design applications of diaphragn walls and contiguous piling. M Fuchsberger Discussion Practice in the detail design applications of anchorages. H Ostermayer Discussion Design and control of bentonite/clay suspensions and concrete in diaphragm wall construction. F T Hodgson Discussion Ground anchors: installation techniques and testing procedures. G $ Littlejohn Discussion Slurry trench practice for diaphragm walls and cut-offs. AL Little Discussion General discussion Sunming-up. T Hanna Appendix. MA Stroud and D J. Sweeney 10 7 27 33 36 55 62 9 86 93 98 124 131 135 142 M Drafting Co It is gray willing to gatt knovledge, exp foundation engi this i very 5: interest and ir practice not o: The value of 51 person, but ev speaker, oF ont 100% This semis meeting of the of that confer Participants w would like an We were deligh be held to pur then. Opening address MW LBONARD BSc (Chairman, Code of Practice Drafting Committee on Foundations, British Standards Institution) It is gratifying to know that there are so many engineers willing to gather together from all parts of the world to exchange Knowledge, experience and skills on this important aspect of foundation engineering. The conference is over-subscribed and this is very satisfying to the organisers and reflects the continuing interest and importance of these expedients in current civil engineering Practice not only in this country but right throughout the world. The value of such seminars to the individual will vary from person to Person, but even only one new fact, often revealed unexpectedly by a speaker, or one new experience, can increase knowledge by as much as 1002. This seminar stems from 2 suggestion made at the concluding meeting of the 1974 conference. It seemed that the full programme of that conference did not leave enough time for detailed discussion. Participants vere invited to let the organising committee know if they would like an opportunity to follow it up with such a meeting as this. We were delighted with the response that a series of seminars should be held to pursue specific subjects in depth, and this is the first of then. Gener: WGK FLEN 1. ate in Septem made comm for retai given abo assumptio 2, Howe than a sh satisfact 3. This retaining much rese understan greatly i the great between t practical into prac bash practices and in tr tered pro Earth Pre 5. Shee been desi with the design in pile wall 6. Ther applied t is used © General analysis of available methods for designing diaphragm walls and anchored walls WG K FLEMING, BSc PhD NICE (Projects Manager (Technical), Cementation Piling and Foundations Ltd) 1, At the Conference on Diaphragm Walls and Anchorages held in September, 1974, some eleven papers out of twenty six made comment on the method of design which had been employed for retaining walls in specific cases. The information given about design varied from simple statements of the assumptions, to a proposed new empirical design method. 2. However, even the latter paper failed to provoke more than a short discussion on the assumptions which might most satisfactorily represent true earth pressures in practical 3. This does not imply that the theoretical approach to retaining wall problems has been neglected, On the contrary much research has been carried out in recent years and the understanding of soil behaviour in such problems has been greatly improved. If little was said it probably reflects the great difficulty which now exists in bridging the gap between the improved appreciation of the problems and the practical means which can be adopted for translating theory into practice. 4, As has frequently happened in the past, construction practices have developed ahead of the theory to support them and in trying to catch up, research engineers have encoun- tered problems of more than usual complexity. Earth Pressures on Retaining Walls 5. Sheet pile walls of both cantilever and tied types have been designed for many years in this country in accordance with the Code of Practice CP 2 (Ref.1) and the methods of design incorporated in it were adopted by designers when bored pile walls and diaphragm walls became available. 6. There are several earth pressure conditions which may be applied to such walls. Occasionally ‘earth pressure at rest! is used where minimal amounts of movement are required during 3 1_DESIGN METHODS FLEMING the various stages of excavation and propping. Such a require ment may be imposed in order to protect adjacent structures. 7, In cofferdams and trench excavations which can be directly strutted, earth pressures are usually derived from the trapezoidal pressure diagrams suggested by Terzaghi and Peck (Ref. 2), These diagrams are commonly used in the modi- fied form in which they appear in CP 2 and they have also been applied in some cases to the design of retaining walls which are tied back by ground anchors at several level>. 8. For the more usual cases of cantilevered walls or walls with few levels of ties, the pressure normally taken into account on the side of the wall retaining earth, is the active Pressure as based on the Rankine or Coulomb theories. Terzaghi (Ref. 3) has showed that the active thrust on a wall is approximately equal to the Coulomb value for small move- ments away from the soil, corresponding to less than } per~ cent of the wall height, whereas considerably greater move~ ments are necessary to mobilise full passive pressures. 9. In 1965 Rowe and Peaker (Ref. 4) demonstrated that in sands movement of between about 3 percent and 30 percent of wall height is necessary to mobilise full passive pressure depending on the sand density. As a result of this work it was suggested that the passive earth pressure coefficients given in CP 2 were rather too high. Based on limiting wall movement to about 5 percent of wall height, new passive earth pressure coefficients were suggested, showing reductions of up to 35% in the worst case, which corresponds to loose sand with a high angle of wall friction. This work also demonstrated that the development of passive pressure was very much dependent on vertical motion of the wall as well as on its horizontal translation. 10. A further examination of the problems of earth pressure was made by Roscoe in the 1970 Rankine Lecture (Ref. 5). From a considerable body of research work it again appears that the passive pressure coefficients for loose sands given in CP 2 are too optimistic and the large strains involved in mobilising passive pressure in such soils are again confirmed. The vork analysed in this paper goes much further than a criticism of CP 2, and throws light on the considerable com- plexity of the interaction between walls and soil masses Gepending on the type of strain behaviour implicit in the system under consideration. Roscoe points out that although much has been learned, the vork has not yet arrived at a stage which will allow its satisfactory application by Engineers to everyday problems. 11, Analysis of the same problems in regard to cohesive soils, 4 FLEMING particularl: behaviour, + to the pres: that any mor are likely the angle o Design pro. 12 Apart pressure de follows frot in the case in the wall ground corr: stability e either the sure is red these revis brium is re. correspond time unreal wall, make 13. When s treatments 1974 by Jam analysis. soil by a5 as a beam, peaks of sh indicated b the method constants. 14, The Co one objecti Taking into sures, and from curren remains to to be amend 15. Comput advocated t applied to diaphragm & who underst 16. One of effects of FLEMING a require- uctures, 2 be ved from zaghi and in the modi- ye also been alls which 3 or walls an into 3 the active on a wall 311 move- am 4 per- that in arcent of 3 work it Hieients ing wall 3 reductions 20 loose : also as well 1 pressure Be 5)e appears inds given wolved in 1 confirmed. chan a able com in the although lata by esive soils, FLEMING DESIGN METHODS 1 particularly with regard to effective stresses and long term behaviour, does not seem to have attracted much research up to the present time, but this is probably due to the fact that any modifications to existing practice found necessary, are likely to be of lesser importance than in cases where the angle of internal friction of the soil is high. Design procedures for retaining walls 12 Apart from the fundamental problems related to earth pressure determination, the method of wall design which follows from CP 2 has certain disadvantages. For example, in the case of a cantilever wall, ‘working’ bending moments in the wall are deduced from a wall penetration into the ground corresponding to a factor of safety for rotational stability equal to unity, In order to ensure stability, either the active pressure is increased or the passive pres- sure is reduced by a so called ‘Factor of safety’ and ‘rom these revised pressures a new wall penetration for equili- brium is reached. Thus the ‘working’ bending moments do not correspond with the actual installed depth and from time to time unrealistic features, such as high shear forces in the wall, make themselves apparent. 13. When such difficulties arise, alternative calculation treatments similar to that suggested in the Conference in 1974 by James and Jack (Ref. 6) can be used for further analysis. This method which involves representation of the soil by a series of equivalent springs and regards the wall as a beam, can be used to demonstrate that in many cases, peaks of Shear stress in the wall are not so severe as is indicated by the simpler approach. A slight’difficulty in the method is the reasonable choice of equivalent spring constants. 14. The Code of Practice CP 2 is currently under revision, one objective being to bring it into limit state terms. Taking into account the problems in dealing with earth pres~ sures, and the difficulty in making any analysis starting from current methods, compatible with CP 110 (Ref. 7), it remains to be seen how radically day to day design will have to be amended. 15. Computer programmes, which are based on the practices advocated by CP 2, are now readily available and have been applied to the design of both bored pile and cast in place diaphragm walls. They offer a powerful tool to the engineer who understands their basic limitations. 16. One of the main advantages of their use is that the effects of altering some of the variables in the analysis 5 DESIGN METHODS FLEMING can be quickly and economically examined. Long and short term stability may be represented, the various stages in excavation and anchor installation can be followed, the effects of mis-selection of soil properties from frequently inadequate site investigation data can be considered, and the general sensitivity of the problem as a whole in relation to ground conditions and factors of safety may be grasped. 17, In the final analysis however, using any empirical or semi-empirical design method, a good deal devends on the experience and judgment of the engineer. The more he can improve his understanding of the various factors involved, the more satisfactory will be the solution which he can offer. In spite of the problems which have been set out above, there seem to be no cases of failure recorded for dia- phragm or bored pile walls up to the present time, and this may well be due to the conservative approach adopted by designers in general. 18. When large deep excavations are considered, over simpli- fication of the analysis of wall and soil behaviour may lead to an inadequate appreciation of total behaviour and sone possible risk to a completed structure. As described by Burland (Ref. 8) in the 1974 Conference, some form of finite element analysis has been found advantageous in modelling the total behaviour of walls and soil mass. Such methods at the present time are likely to be costly and are only as good as the data which is available to work from, but , they appear to offer an important tool to engineers involved in large projects, where experience on similar projects is : often not available. Problems in retaining wall design 19, There are several lesser problems which make themselves uncomfortably apparent in everyday design. One of these relates to surcharge loading, which is often due to existing foundations within a retained soil mass. Such surcharges may be represented by point loads, line loads or specific loaded areas. In CP 2 a simple graphical procedure is given for working out resulting horizontal surcharges to the wall and an appendix is offered on the calculation of lateral pressure due to a concentrated surcharge, based on elastic theory. Terzaghi (Ref. 3) has suggested the use of equations based on Boussinesq, but in practice one can find consider- able differences in the results obtained by the different methods, particularly when footings rest a short distance above the excavation level within the retained earth. 20. Another problem is related to the use of effective stress methods for examining the stability of walls in clay 6 FLEMING soils, inc will always its worst 1 analysis is than of the 21. Groune difficule t the worse « must of cor which coulc walls havir 22. The or problem in retaining + mass which where the 1 particular but if the must be im within the far from ¢ Deep cireu 23. Deep diaphragm for using 24, The p demonstrat such walls seems unli walls, but that full hence the been const projects © tions, @ depth, anc Design of 25. The « retaining paper to ! been writ! about safe revised e FLEMING and short tages in ed, the frequently ered, and e in relation e grasped. pirical or 8 on the re he can involved, he can set out. rded for dia~ e, and this pted by over simpli- our may lead and some ribed by rm of finite nodelling h methods are only rom, but ars involved cojects is » themselves of these to existing ieharges specific tre is given © the wall lateral n elastic of equations 4d consider~ ifferent distance rth, ective 1s in clay FLEMING DESIGN METHODS I soils, in order to ensure that some adequate factor of safety will always exist. The problem springs from the difficulty in assessing how long an excavation will remain open under its worst loading conditions, and whether an effective stress analysis is appropriate. The absence of soil tests, other than of the triaxial undrained variety, does not heip. 21, Groundwater problems also arise and frequently it is difficult to interpret the data supplied so as to determine the worst conditions likely to arise in practice. Care must of course be taken to avoid water pressure gradients which could bring about 'boiling' conditions at the toes of walls having relatively short embedment. 22, The overall stability of excavations can also be a problem in certain cases, with the possibility of complete retaining walls and anchorage systems lying within a soil mass which is part of a potential ground slip. Clearly where the wall lies totally within such a slip zone, its particular design is not relevant to the general situation, but if the wall intersects the potential slip its strength must be important in relation to the other forces involved within the system. The best way to treat such problems is far from clear at present. Deep circular excavations 23. Deep circular excavations are frequently retained by diaphragm walls, and engineers have debated the necessity for using ring beam walings in such circumstances. 24, The paper presented by Rigden and Rowe in 1974 (Ref.9) dewonstrated that without waling beams and reinforcement, such walls could be built and perform satisfactorily. It seems unlikely that engineers would use such unreinforced walls, but even with reinforcement, there are many who doubt that full continuity of contact will exist between panels ~ hence the need for waling beams. However such walls have been constructed without walings, and the success of such projects seems to have depended on suitable ground condi- tions, a judicious choice of wall width in relation to depth, and on considerable care in construction. Design of anchorages 25. The design of individual ground anchors for tying back retaining walls has been discussed by Ostermayer in his paper to the 1974 Conference (Ref. 10) and a good deal has been written in the codes of practice of various countries, about safe procedures for design and construction. The revised edition of CP 2 which is now in the drafting DESIGN METHODS I FLEMING stages deals with various aspects of ground anchor behaviour. 26. At the present time there appear to be difficulties connected with determination of safe loads for individual anchors in certain ground conditions. Subject to adequate capacity being developed on the individual anchor, to the placing of anchorage load transfer lengths outside zones of potential soil movement resulting from wall deformation, and to reasonably chosen anchor spacings, the technique is very successful. However, the.long term behaviour of anchors in clay soils and the interaction between anchors, soil maes,and retaining wall would seem likely to provide areas for the further improvement of understanding. Movements associated with walls 27, As stated above, there does not appear to be evidence of failures of diaphragm walls in published literature up to the present, and in those cases where movements have been monitored, performance appears to be good by comparison with sheet piling and other support systems. As information builds up it may be possible to improve forecasting of wall behaviour. 28. It is worh noting however that failures have occurred in tied sheet pile walls, and the reasons why these took place. Broms and Stille (Ref.11) have examined a number of cases of failure in soft clays, pointing to causes such as the action of frost leading to lateral expansion of soils behind walls, thus inducing failure of the ground anchors, the driving of bearing piles imto the retained soil mass, and the use of excessively inclined anchors which produced vertical reactions to the wall which the sheet piles could not sustain. These problems could equally be relevant to diaphragm and bored pile walls in certain circumstances. Load bearing diaphragm walls 29. The use of diaphragm walls as earth retaining structures represents one major application, but they also have achieved widespread popularity on the Continent as load bearing units. They can offer a means of carrying both vertical loads and substantial horizontal loads, and can be constructed in varying configurations to suit the particular requirements of the supported structure. 30. Some engineers have in the past been suspicious that the presence of bentonite inhibits the full development of frictional resistance along the sides of such units, but the experience of continental 8 PLEMING practice anc constructed both confire constructior REFERENCES 1. Civil B Retaini 1951. 2, Terzagh Enginee 3. Terzagh of Aner Vo1.119 4. Rowe, P Measure 5. Roscoe, Inflven 1970, v 6. James, walls. Instite 7. Code of Britisk 8, Burland Diaphre Enginec Rigden, reinfor and Anc 1975. 10. Osterm creep Diaphr: Engine: 11, Broms, Sheet of Civ FLEMING behaviour. ulties ividual 1 anchor, outside 11 ngs, the erm behaviour en anchors, provide evidence ture up have been rison with mation g of wall why jing aducing ing piles ively to the se 1 bored ey itineat arrying 8, and t the FLEMING DESIGN METHODS 1 practice and the extensive use of large bored piles constructed by the same basic technique in this country, both confirm that such fears are unjustified where good construction practice has been followed. REFERENCES 1. Civil Engineering Code of fractice No. 2. Earth Retaining Structures. Inst.Structural Engineers, 1951. 2, Terzaghi,K. and Beck, R.B. Soil Mechanics in Engineering Practice, Wiley, New York, 1948 3. Terzaghi, K. Anchored Bulkheads. Transactions of American Society of Civil Engineers, 1954, Vol.119. 4, Rowe, P.W. and Peaker, K. Passive Earth Pressure Measurements. Geotechnique 1965, Vol.15, No.l. 5. Roscoe, K.H. Tenth Rankine Lecture - The Influence of Strains in Soil Mechanics. Geotechnique, 1970, Vol 20, No. 2. 6. James, E.L. and Jack, B.J. A design study of diaphragm walls. Proc. Conf. Diaphragm Walls and Anchorages, Institution of Civil Engineers, London 1975. 7. Gode of Practice CP110, The Structural Use of Concrete. British Standards Institution, London 1972. 8, Burland, J.B. Discussion on Papers 5-9. Proc.Conf. Diaphragm Walls and Anchorages, Institution of Civil Engineers, London 1975. 9. Rigden, W.J. and Rowe, P.W. Model performance of an un~ reinforced diaphragm wall. Proc. Conf. Diaphragm Walls and Anchorages, Institution of Civil Engineers, London, 1975. 10. Ostermayer, H. Construction, carrying behaviour and creep characteristics of ground anchors. Proc.Conf. Diaphragm Walla and Anchorages, Institution of Civil Engineers, London 1975 11. Broms, B. and Stille, H. Failure of Anchored Sheet Pile Walls. Proceedings of American Society of Civil Engineers, 1976, Vol.102, No. GT3. 9 I DESIGN METHODS DISCUSSION DISCUSSION REPORT BY K F SIMM BSc FICE MIHE (Freeman Fox and Partners) Design Methods 1, Fleming, in his introduction, outlined the following currently available design methods: (a) The method given in CP2 and, in particular, the minimum fluid pressure requirement. (b) Rowe's method which takes into account the flexibility of the wall, its anchor system and the soil. It deals with the cantilever and single tied cases but has been little used for diaphragm walls, possibly because it was originally put forward for more flexible sheet pile walls. For stiff cantilever walls it gives bending moments little less than those determined by the Free Earth Support system for which in practice, the classical pressure distributions most nearly apply. (c) Brinch Hansen's method is a limit state design method in which various forms of failure are postulated with the formation of hinges at the points of maximum bending moment. The soil is assumed to be, in parts, in a state of failure and the earth pressure diagrams are assessed from the rotations of the various parts of the wall about the hinges. (4) Other, newer techniques were debated at the 1972 Madrid Conference on Soil Mechanics and Foundation Engineering. One is based on the examination of the velocity fields in the failure zone adjacent to the earth retaining structure. Another is the finite element approach where the elements can be ascribed parameters which represent their actual behaviour and which can be varied with depth. This is a very expensive method. 2. He drew attention to the CIRIA publication A COMPARISON OF QUAY WALL DESIGN METHODS published in 1974 which compared the requirements of various Codes of Practice for the design of single tied sheet pile walls. 3. _ Of the design methods listed above, the first was the most widely used and Simm thought this was mainly for two reasons : 10 DIScUsstoN (a) () 4. The re method and New Methods 5. Rowe s Engineering engineers t observation to this was a tremendou British eng 6. There of a partic importance appreciatio such things of props am: provide a r. different p. 7. Simm s support sys and plastic the wall an support sys It was not until we co its support compatible and enginee excavations and Peck en DISCUSSTON and Partners) following lar, the the sem and the single tied iragm walls, forward for ££ cantilever ’ss than. vrt_ system sssure design method stulated with maximum be, in parts, sure diagrams arious parts the 1972 Foundation ation of the cent to the he finite be ascribed behaviour 3 is a very 4 COMPARISON ich compared c the design it was the 1 for two piscussron DESIGN METHODS 1 (a) Engineers when reporting on results of tests on earth pressures and engineers trying to understand and apply results of the tests do not always distinguish very clearly between Loads on the diaphragm wall or sheeting and the envelope of loads used in the design of the support system, Thus confusion arises in trying to correlate the results of one set of tests with another. () When load measurements are taken on site and the results published, they are not always correlated with the way the excavation was carried out and the way the supports were inserted and these construction problems have a much greater effect on the support loads than any difference in the earth pressure diagram assumed. 4. The result was a lack of confidence in any- design method and engineers tended to rely on the Code of Practice. New Methods of Design 5. Rowe said that advances in the science of Foundation Engineering came mainly from the ability of practical engineers to devise ways of building things and from observations on how they behaved in field conditions. Second to this was the study of physical models in which there was a tremendous interest from Continental but almost nil from British engineers. 6. There was general agreement among speakers that choice of a particular method of analysis was of secondary importance compared with a good site investigation and a proper appreciation of the influence of construction techniques © such things as surcharges, temporary berms, different positions of props and stages of excavation. Computer analysis can provide a rapid indication of the effects of varying the different parameters but it can be expensive. 7. Simm said that any diaphragm wall together with its support system and the ground it retained had both elasticity and plasticilty to some degree. As the excavation progressed the wall and ground moved inwards until restrained by the support system or the ground beneath the base of the excavation. It was not possible to put figures to these parameters but until we could find a simple way of treating the excavation and its support, system as a unit and making the Loads and movement's compatible he did not think that much progress could be made and engineers would have to rely on test results obtained from excavations carried out in a similar way, such as the Terzaghi and Peck envelopes for prop loads in strutted excavations. un DESIGN METHODS Discussion 8. For important excavations, he suggested that it was a fairly simple matter to monitor strut loads in such a way that the contractor was not affected - in this way the number of props could be increased or decreased as necessary. This was done successfully at Humber Bridge. 9. Zeitlen agreed with the idea of in situ measuring and gave an example of an in situ load test which was carried out on a sheet pile wall tied to an anchor wall. 10, After the wall had been constructed, loads were jacked into the ties and the resulting movements of the Sheet pile and anchor walls were measured. By back analysis the equivalent moduli of deformation could be determined. 11. He suggested that this method of testing could easily be adopted in situations where a temporary distortion of the wall top was not important and much useful information could be obtained especially if inclinometers were used to check the wall deflections at the same time. Limit State 12. In order to make foundation design compatible with cPi10 (The Structural Use of Concrete) a limit state method is required but all speakers agreed that the problem was how to assess the characteristic strength of the ground and the pressures applied to the wall at the limit state. 13. Limit states must apply to the wall, support and soil and Rowe doubted if these could occur at’ the same time without gross distortion. He was not against limit state design but thought that because of the amount of distortion necessary to reach the critical state, practical engineers Table 1 , Rate of pulling in mm/min. - Piles castin _ |-—— escastin 1 01 0.01 | 0.001 Water 3.0 26 22 : Clay 31 28 24 7 Mikrosoil 31 28 24 - Bentonite 23 20 18 16 2 DIScussioN 20 — 10 PULLING FORCE,Mp would find i practice sho pointed out with ensurin Skin Frictio 14. Karlsru piles under friction, g Geotechnical 15. Pulling on 8 No. 250 clay. Two p slurry, two slurry made summarised i 16. It can a considerab all the pile to them exce smooth, In the enhanced 'SCUSSION t was a a way that umber of This ing and arried the be | easily on of ormation used to with te method is ras how Cand e. wd soil ime state 4 stortion \gineers DISCUSSION DESIGN METHODS 1 t e 2.20 i Zi Pl] 2 tt 10 Q t 5 | 12m 50 I z= 0 10 20 30 40 0001 01 10 = | HEAVE, mm PULLING vELocity || mm/min 6m TEST PILE B25 cm Fig.1 would find it unacceptable. He thought that codes of practice should guard against unsatisfactory performance and pointed out that in many excavations design was concerned with ensuring that adjacent structures did not move. Skin Friction on Piles cast under Bentonite 14. Karlsrud, conmenting on statements that casting in situ piles under bentonite does not materially affect the skin friction, gave details of tests carried out by the Norwegian Geotechnical Institute in Oslo. , 15. Pulling tests were carried out at various rates of pull on 8 No. 250 mm, diameter piles cast 12m-18m in soft marine clay. Two piles were cast in pure water, two in a bentonite slurry, two in an in situ clay slurry and two in a special slurry made up of microsoil powder. The results are summarised in Figures 1 and 2, 16. It can be seen that casting piles under bentonite caused a considerable decrease in shear resistance. On extraction all the piles had a lime stabilised layer of clay attached to them except those cast under bentonite and these were smooth, In calculating the shear resistance of the piles the enhanced circumference was used. 3B 1_DESIGN METHODS DISCUSSTON 17. Fleming said that tests carried out under the sponsorship of CIRIA, similar to those of Mr Karlsrud had shown the opposite result, He said it was very important to know exactly what the soil conditions were and the construction technique used - in particular the length of time that the borehole remained open and the type of bentonite used. Leonard asked what method had been used to tremie the 250mm diameter piles and if the mid density had been carefully controlled. 18. Jefferies said that the type of’ bentonite used could have a considerable effect on the results of the tests. 19. During back filling with concrete some of the cementitious material may diffuse into or mix with any filter cake on the walls of the excavation. This would produce a mix similar to a clay cement grout and its strength would depend on the type of clay in the cake and the type of cement used. If the cake was from a native clay slurry the mix would generally be rather weak. With a bentonite cake the mix would be much stronger especially if a blast furnace slag cement was used. If the cement penetrated the whole of the cake the skin friction would depend on the shear strength of the mix, the ground, or their interface whichever was the weakest. 20. However, if part of the thickness of the cake remained unaffected by the cement its shear strength would probably control the skin friction, With time the cake would tend to consolidate and its strength could be estimated from tests on the consolidated slurry at the appropriate overburden pressure. Results of such tests on two types of bentonite produced the following figures for the undrained angle of shearing resistance. English, converted sodium, bentonite gcu = 12-18° Wyoming, natural sodium, bentonite deu = 6° 21. Thus the skin friction for an excavation under Wyoming bentonite slurry might be between two and three times less that that for an English bentonite slurry. This is probably because the English bentonites are partially flocculated by chemicals added during the conversion whereas Wyoming bentonite which contains no added salts forms a deflocculated slurry. Anchored Walls 22, Ostermayer agreed with earlier speakers that for 4 BIScUSSION strutted e pressure d wall howev earth pres extent by the anchar 23. He sa Wal1s and published loads of a the excava not creep very well If for exa with a tri Paper 6 by MacFarlane the walls 24. He su to the Cou overall st certain li giving sui experience 25. Rowe it was pos walls by a the anchor early 1950 physical m the Limit movement 0 thought th to consoli 26. Littl United Kin behind wal control mo are assume there are 27. Sever owners abo structures caused by weight of DISCUSSION » sponsorship m the > know, struction that the ased. the 250mm sefully 2d could have 2 cementitious zake on the < similar to on the type If the cake ally be be much = was used. 2 skin 2 mix, the ast. ath ime the ould be 2 the tests rance, t for DISCUSSION strutted excavations the amount and distribution of earth pressure depended on wall displacements. For an anchored wall however, the wall movements and hence the corresponding earth pressure distribution could be chosen to a certain extent by jacking suitable forces into the soil through the anchors. 23. He said that the papers at the Conference on Diaphragm Walls and Anchorages in London in 1974 together with other published case history.studies confirm that the prestress loads of anchors does not usually change very much during the excavation stages provided that the individual anchors do not creep and they confirm that wall displacements correspond very well with the assumptions of earth pressure distributions. If for example the anchors are prestressed in accordance with a triangular pressure distribution as suggested in Paper 6 by James and Jack or in Paper 15 by Littlejohn and MacFarlane the measured wall displacements demonstrate that the walls are rotating about the toe. 24, He suggested that provided the total pressure corresponds to the Coulomb or Rankine active earth pressure and that the overall stability of the wall is satisfactory then within certain limits the pressure distribution may be chosen by giving suitable prestress to the anchors. He asked if experience in the United Kingdom confirmed this statement. 25. Rowe did not agree with Ostermayer's statement that it was possible to control loads on and movements of walls by a judicious choice of the loads stressed into the anchors, Brinch Hansen had made similar claims in the early 1950's but Rowe had demonstrated by the use of a physical model that it was not true, at least as far as the limit state in dense sand was concerned, because of the movement of the sand due to the wall deflection. He thought that similar problems might occur in clay soils due to consolidation. 26. Littlejohn said that it was not current practice in the United Kingdom to use prestressing to control pressures behind walls although it had been used to some extent to control movements. Generally high earth pressure diagrams are assumed mainly in order to minimise movements where there are adjacent structures. 27. Several speakers mentioned the concern of property owners about the effects of ground anchors beneath their structures and it was pointed out that ground movement was caused by the excavation of a hole by virtue of the self weight of the surrounding soil and not by the ground anchor 1s q DESIGN METHODS DISCUSSION although this did compress the ground within the cable length and cause tension behind the anchor. Most anchored walls have very deep anchors and Bassett showed slides of model tests in which the anchors had been progressively shortened until they were inside the ‘active zone’, effectively creating a gravity wall which was safe provided he did not exceed the 3 rule or the allowable bearing pressure. 28, He suggested there might be a good cage for using this short anchor technique to avoid problems of access for anchors beneath adjacent properties or for using thinner diaphragm walls and prestressing them vertically. 29, It was pointed out that the short anchors were more akin to reinforced earth than to a conventional anchor system and Littlejohn said that he would use the reinforced earth principle only in purely frictional materials because so little was known about load transfer in the fixed anchor zones in heterogeneous soils. 30. Referring to the comments by Dr Fleming in his introduction on the design of anchorages, Littlejohn said that every Code of Practice pertaining to anchors had now been translated into English except for the 1976 Austrian Draft Code. None of them include information on how to design anchorages. Circular Excavations 31, In a written contribution, Reid referred to Fleming's comments on the use of diaphragm walls for supporting 7 deep circular excavations without using waling beams and gave an example of one 20m diameter and about 15m deep which fhe had recently completed. The diaphragm walls were excavated through difficult soil to rock and contained only a minimal amount of steel. 16 Current d condition BO CORRETT, 1, Methods 0 ledge, more pov ‘They move back cessive legislati coach is shown construction pre the recovery aft 2. It is unfor construction job Conference on T tinued developm electronic pockt During the samc state design for 3. The bias i both at home an practice, partic 4, The desig the panels of a« cently by Profe: EARTH PRESS! 8. The Code ible with CP 11 their underlying for example the 6. The philo iscussion able length d walls have el tests ned until eating exceed the sing this for hinner e more akin 2 the al materials the fixed 3 introduction avery Code aslated je. None rages. -eming's ng . 1s and leep which ned only II Current design practice applied in differing ground conditions for diaphragm walls, contiguous piles and anchorages BO CORBETT, BSc FICE MIStructE (Geotechnics Division, Ove Arup and Partners) 1. Methods of design are never static; they move forward with increasing know- ledge, more powerful mathematical tools, and confidence following successes. ‘They move backwards when something goes wrong duritig construction, when ex- cessive legislation limits the scope for new ideas and when a new analytical appr- ‘ach is shown not to fit the behaviour. ‘The challenge of increasingly complex construction projects provided an impetus over the quarter of a century following the recovery after World War II up to the oil crisis of 1973. 2, It is unfortunate that the depressed state of the economy has provided few construction jobs to advance techniques in this field during the two years since the Conference on Diaphragm Walls and Anchorages (ref. 1). Nevertheless, the con- tinued development of computerized methods of analysis and the availability of electronic pocket calculators, has led to improvements in the design process. During the same period there has been a more widespread understanding of limit state design for structures (ref. 2). 3. The bias in this note is towards United Kingdom design practice, as applied both at home and overseas. This has been influenced much more by Continental practice, particularly French and Italian, than by American. 4, — The design principles involved in the use of bentonite clay muds to support the panels of a diaphragm wall during excavation have been well summarised re- cently by Professor J.K.T.L. Nash and are not considered further here (ref. 3). EARTH PRESSURES 5. The Code of Practice for Earth Retaining Structures (ref. 4) is not compat ible with CP 110, The result of using the two codes for design without regard to their underlying philosophies can be a wall inadequate for its intended purpose; for example the penetration below excavation level may be insufficient. 6, The philosophy of limit state design differs essentially from the traditional uv 1 ‘II DESIGN APPLICATIONS coRserr way of structural design by limiting stresses. In limit state design, the first con- sideration is of the structure at failure, and partial safety factors are applied, one for material strength and the other for loads and load effects, to determine the structural sections. 7. On the other hand, where design is based on safe stresses and service de- formations, the thought process works in reverse. Starting with a design which is conservative, the components are reduced in size to the minimum acceptable level. . 8. The revised Earth Retaining Structures Code CP 2002 will incorporate the development of advanced theories on earth pressures as far as is practicable, and will be based on a Limit State Concept compatible with that in CP 110. In this the purpose of design is the achievement of acceptable probabilities that the stru~ cture being designed will not become unfit for the use for which it is required i.e. that it will not reach the limit state. All relevant limit states need to be considered in the design so as to ensure an adequate degree of safety and serviceability. The usual approach is to design on the basis of the most critical limit state and then to check that the remaining limit states will not be reached. 9. CP 110 has been generally biased to ensure that a structure which is desig- ned according to its provisions will not be much different from a structure for the same purpose if it had been designed in accordance with CP 114 (ref. 5). For ‘most flexural members (such as diaphragm walls) the serviceability criteria will govern the sizes. 10, Some design offices have done parallel designs by both methods. ‘These in clude government, contractor and consultant but as far as is known there has not yet been any co-ordination of the work. 11, Most earth pressure calculations are carried out either by Coulomb or Rankine theory (refs. 6, 7). 12, Rankine's Theory (1857) assumes that there is a direct relationship between the vertical pressures within the soil mass and the lateral pressures on vertical planes within the soil adjacent to the retaining wall. In other words, it is assum- ed that the presence of the wall introduces no changes in shearing stresses at the surface of contact between the wall and the backfill. 18. The Coulomb Theory dates from 1776, and is based on the concept of a fail- ure wedge which is bounded by the face of the wall and by a surface of failure that passes through the toe of the wall. There are two assumptions. The principal one is that the surface of failure is a plane, and the other that the thrust on the 18 coRRErT wall acts in so. on the wall can 14, Although cohesive soils, ects of seepage efficients of ea 15. A few ba: izontal stress : states, using F where it was ni movements to | Centre (ref. 9) 16, Other the ment of a wall duce toa mini: owable to mobi mobilised sign it active or pa: with respect tc when support t founded on a r: active soil wee 17. The pase mine; British « seen by compa for Earth Reta with =35° an are 7.3 and 10 Frequently the the ground wai sibly through wall is ina re ective stress: 18, Profess: years ago int wall of the ba: By lowering t! conser , the first con are applied, to determine ad service de~ « design which um acceptable weorporate the practicable, CP 110. In this s that the stru- is required i.e. to ensure an is to design on remaining which is desig- tructure for 4 (ref. 5). For ty criteria will ods. These in~ » there has: not oulomb or ionship between es on vertical s, it is assum- stresses at the ncept of a fail- + of failure that he prineipal shrust on the ‘Corser DESIGN APPLICATIONS | IT wall acts in some known direction. With these assumptions, the resultant thrust on the wall can be determined by statics. 14, Although both theories apply to granular soil, they have been modified for cohesive soils, and to allow for wall friction. It is the wall friction, and the eff- ects of seepage pressures that have a critical effect on the magnitude of the co- efficients of earth pressure. 15. A few basement walls have been designed by considering the changes in hor- izontal stress at deformations intermediate between the active and passive states, using Henkel's stress path method (ref. 8). These cases have been those where it was necessary to provide a positive system of restraint to restrict movements to prevent damage to surrounding structures, as at the Barbican Arts Centre (ref. 9). 16, Other than the situation described {n the preceding paragraph, the displace- ment of a wall is invariably sufficient for the pressure on the active side to re duce toa minimum, Rarely is suffictent displacement of the toe of the wall all- owable to mobilise the full passive resistance. ‘The wall friction which can be mobilised significantly affects the horizontal component of the earth pressure, be it active or passive. If the active wedge moves, or is free to move, downwards with respect to the wall then the active horizontal force is reduced. However, when support to the wall is provided by inclined anchors, then unless the toe is, founded on a rigid stratum the wall is likely to move downwards relative to the active soil wedge, and the active horizontal force is increased. 17. The passive resistance mobilised at the toe of the wall is difficult to deter mine; British design practice is more conservative than in France. This can be seen by comparison of the coefficients of passive pressure in the Code of Practice for Earth Retaining Structures with those in Schneebeli (ref. 10). For example with § =35° and maximum allowable wall friction, the comparative parameters are 7.3 and 10.7. Frequently the effect of seepage on the passive resistance is neglected. Where the ground water level is high, then there will be seepage under the wall and pos~ sibly through joints in the wall into the excavation. Ina case where the toe of the wall is ina relatively permeable soil, then the situation can arise where the eff- ective stress on the passive side of the wall is reduced to zero. 18. Professor R.E, Gibson and the writer were faced with this problem twelve years ago in the design of the temporary works for the 20 m. high east perimeter wall of the basement at Fawley Power Station (ref. 11). By lowering the water pressures outside the retaining walls the passive pressures 19 11 DESIGN APPLICATIONS conser were increased and the active decreased and without this control construction would have been impossible. ‘This is an early example of an effective stress an- alysis for calculation of wall stability. 19. Another documented example is the paper on the construction of a dry-dock at Hebburn, Co. Durham (ref. 12). 20. Neither of the two examples quoted above were diaphragm walls, but there is no difference in principle. A similar situation exists for excavations within diaphragm walls for the Chater, Admiralty and Pedder Stations for the Hong Kong Mass Transit Railway, now being built (ref. 13). DEFORMATIONS 21, The allowable deformations around the site can be very important (see para~ graph 15) and had major design implications at Britannic House.and Barbican Arts Centre (refs. 9, 14). Finite element analyses were used to assess the movements (sce paragraph 33), Deformations are due to two causes : a) movements of the wall b) effective stress changes due to seepage. Deformations are controlled by the support system of anchors, struts or perma- nent structure, and by extending the wall downwards to act as a cut-off. STRUT LOADS 22. Whatever design method is used to calculate the strut loads, variations will occur unless the forces are jacked into the system using, for example, Freysin- net flat jacks. ‘These variations are due to : a) method of installation b) sequence of installation ©) movements which have already occurred during installation of the wall or with excavation of material 4) temperature, DESIGN CONSIDERATIONS SPECIFIC TO DIAPHRAGM WALLS. 23, The length of panel in a diaphragm wall is governed by construction consid- erations, which are topics for other sessions. Arching is important, but difficult to quantify at the design stage and in practice the implications on the design are a matter of experience at a particular site, For example, in London Clay (Berbic- an Arts Centre) collapses occurred beneath the guide walls in panels only 4.4 m. long whereas with the barrettes (load bearing diaphragm wall panels) through the Marnes et Caillasses at Centre Pompidou (formerly Centre Beaubourg), Paris, 20 CORBETT there were no ec effective depth 0 Tee shaped pane Barbican Arts C Construction Li 24. Except for widen the sectioy due to the plant, placement of cor Load Transfer 25. Many engir shaft friction. 1 and is not borne a load test on tw the load/deflecti whilst Farmer e in situ piles com bored piles form Fernandez-Rena that the shaft fri ion and concreti 26. With rega test panel in Lo anticipated by # wall panels have France, for exa DESIGN CONSII 27. Bored pil: for the success re adjacent pile ‘This note does of design princi Current practic to calculate the CP 114 (ref. 1€ erally as summ with CP 110 ar corner? ‘onstruction tive stress an~ nof a dry-dock alls, but there ations within the Hong Kong rrtant (see para- 4 Barbican Arts the movements uts or perma- off. variations will ple, Freysin- tion of the uction consid- Mt, but difficult e design are a Clay (Barbie~ s only 4.4m. 3) through the 1g), Paris, CORBETT DESIGN APPLICATIONS TT there were no collapse problems with single panels 11 m. in length (ref.15). ‘The effective depth of the section of a diaphragm wall can be increased by using I and ‘Tee shaped panels. The panels for the theatre of the north and soath walls of the Barbican Arts Centre were designed and constructed in these shapes. 24. Except for the case in the preceding paragraph where it is necessary to widen the section to increase the bending resistance, the construction limitations due to the plant, the size end weight of the reinforcement cages, and the rate of placement of concrete do not, in practice, constrain the design principles. Load Transfer 25, Many engineers are of the opinion that bentonite can substantially reduce shaft friction, However, this opinion appears to be based primarily on intuition and is not borne out by test results. For example,, Burland (ref. 17) found from a load test on two diaphragm wall panels, formed with and without bentonite, that the load/deflection characteristic and ultimate load were virtually identical, whilst Farmer et al (ref. 18) found that the shaft friction developed on bored cast, in situ piles concreted under bentonite was higher than would be anticipated for bored piles formed in the dry. Similar observations have been reported by Fernandez-Renau, Chadeisson, Corbett et al and Cernak, Cernak also observed that the shaft friction mobilised decreased with increasing time between excavat- fon and conereting. (Refs. 19, 20, 21, and 22), 26, With regard to the load carried by the base, Corbett et al recorded for a test panel in London Clay the base was carrying just under half the ultimate load anticipated by the normal semi-empirical design method, However diaphragm wall panels have been generally accepted as load bearing elements (barrettes) in France, for example, at Centre Pompidou (ref. 15) and Tour Maine Montpar- nasse. DESIGN CONSIDERATIONS SPECIFIC TO CONTIGUOUS BORED PILE WALLS 27. Bored piles are used to form a continuous wall in two ways. The first is for the succession of piles to be in contact, or contiguous and the second is whe- re adjacent piles are bored into the adjoining pile, forming a wall of secant piles. ‘This note does not specifically consider secant piles, but from the point of view of design principles the difference is not significant. Current practice is to calculate the earth pressures as for diaphragm walls and to calculate the diameter and reinforcement required by load factor design to CP 114 (ref. 16), The vertical loads which can be carried by the piles are gen- erally as summarised by Skempton (ref, 23). ‘The difficulties of design methods with CP 110 are as for diaphragm walls, a II DESIGN APPLICATIONS CORBETT ANCHORAGES 28. The detail design applications of anchorages form the topic for Session IV, so in this note only the principles are considered. The overwhelming advantage of an anchored wall is that it provides an unrestricted working space within the excavation, The main limitation is that there is little if any scope for retrieving a situation where an anchored wall has been under-designed or the anchors have been poorly installed. In other words, an anchored wall is not a fail-safe system. 29, The principles of design are in 3 stages : a) the pull-out strength of the anchor, which depends on the type of sotl b) the determination of the geometrical arrangement of the anchorages ©) overall stability. ‘The pull-out strength is the smaller of the failure of the strands or the failure of the anchorage in the soil. With regard to the geometry, re-entrant corners require’special consideration. ‘To ensure stability, the fixed anchor zone should be far enough away from the wall to ensure that a slip beneath the toe of the wall will not occur. ‘The most used design method is by Littlejohn, which closely follows French pra~ ctice (refs. 24, 25). 30. It has already been demonstrated in paragraph 16 that steeply inclined ane- hors, imparting a vertical force downwards to the wall, can change the earth pre- ssure distribution so as to increase the horizontal pressures at the top of the wall, which means that increased loads are carried by the system at the highest stiff point, namely at the top level of anchors. Where possible then, it Is best to keep the inclination to the horizontal as small as possible compatible with the plant to be used for installation. Nevertheless, it may be of advantage where the penetration below final dredge level is small to design for high active pressures and provide the whole of the passive resistance on the support system. 31. Corbett and Stroud (ref. 15) mentioned that it was not difficult to envisage that the failure of a small number of anchors in one section could cause a cas~ cade failure and rapid collapse of the whole support system. In some countries the use of waling beams to reduce the risk of such a failure is obligatory. For example, under Swedish regulations, if an anchor fails, then the walings must be capable of redistributing the loads to adjacent anchors, s0 that neither these nor the wall fails. ‘To check on the validity of the regulations Stille and Broms have studied the redistribution of load between adjacent anchors and in the passive zone, following the failure of a single anchor (ref. 26). 22 coRBETT: COMPUTER J 32. The con movements, « Movements 33. Finite e can deal with @ef. 14). Tr ing layout ins tated by other Earth pressur 34. The pro foundations (¢ wall. (Ref. 25 Stability 35. Althoug majority folle ‘TOPICS FOR Diaphragm we 36. The top a) bo @ 2 ot Anchorages 37, The fot a oo by) tt b t ACKNOWLEE 38. The Au publish this r

You might also like