Seismic earth pressures – a technical note

Merrick Taylor Arup Geotechnics, London

his brief technical note summarises the key meth- ing in terms of wall movements – be they settlements or de- ods in codes and the literature and how they are €ections (Serviceability Limit State or SLS). For the seismic Tapplied. It is a synthesis of papers presented by the case we wish to check essentially the same limit states for a author last year at the New Zealand Geotechnical Sympo- “design” earthquake. e design earthquake will have been sium [1] [2] and recent experience on projects in Europe. speci!ed by national or industry code, or a special study e focus is on pseudo-static earth pressures used in typi- to determine the acceptable level of hazard for the struc- cal day-to-day engineering practice, where the transient ture. is increasingly considers dierent levels of seismic dynamic load of an earthquake is simply represented by an hazard to ensure the performance at frequent earthquakes additional static pressure or force resultant on the structure is not disappointing (analogous to an SLS check) as well being designed. e use of more advanced methods to con- as ensuring rare earthquakes do not lead to failure (a ULS sider the dynamic response is only discussed brie€y where check). Here “failure” may be de!ned in terms of unaccept- understanding the dynamic soil-structure interaction may able wall displacements rather than a catastrophic failure be important, such as deriving the eective design inertia. of the wall. A clear understanding of the required perform- ance of the structure in terms of displacements is advisable prior to undertaking design. Modern codes (e.g. Eurocode Seismic design 8 (EC8) [3]) may implement this philosophy implicitly by assuming performance during frequent events will be suf- Philosophy !cient for normal structures if the ULS check performed For static design, the aim of code guidance is to ensure that for rare events is satis!ed with a given safety margin. Fa- catastrophic collapse or failure is avoided (Ultimate Limit cilities supporting hazardous materials and processes (e.g. State or ULS), but also to ensure that the performance of LNG, Nuclear) tend to have more explicit requirements for the structure under the applied load will not be disappoint- the respective hazard levels – the latter will have usually

q x s y

∆

Pwae

qs.H ∆Pae

xc Pwpe

Pwa – P wp Pa W. kh

yc ∆P pe P W. kv p W

cl e NB: Pseudostatic components R have dashed lines & given subscript e, or multiplied by X kh or kv.

Figure 1. Typical loading diagram for a gravity quay wall under seismic loading, assuming active earth pres- sure conditions occur at the ULS.

4 SECED Newsletter Vol. 21 No. 4 October 2009 B P AE k h W

k v W W

W k h W

Q k v W F

D F' P A A. Y AE F

kh / kv = horizontal/vertical seismic coe"cient γratio kh θ = arctan  − ru ± kv

sin (ψ + ϕ − θ) (a) KAE = cos(θ) sin (ψ) sin(ψ − θ − δ)( ! + X)

.! sin ϕ + δ sin ϕ − β − θ where X = if β > ϕ − θ, otherwise: X = ( ) ( ) sin ψ − θ − δ sin ψ + β ( ) ( )

sin (ψ + ϕ − θ) (b) KPE = cos(θ) sin (ψ) sin(ψ + θ)( ! − Y)

.! sin ϕ sin ϕ + β − θ where Y = if θ > β + ϕ, otherwise: Y = ( ) ( ) sin ψ + β sin ψ + θ ( ) ( )

γratio ≈ 1.6 for dynamically pervious back€ll, 2.0 for impervious back€ll, 1.0 for dry. Ref. Matsuzawa et al. [7]

Figure 2. Mononobe–Okabe dynamic earth pressure coecient calculation; (a) active and (b) passive limit condi- tions. Coe€cient K is the ratio of lateral to vertical eective stress at the wall-back!ll interface. NB: "e passive condition has no wall friction δ considered, a#er EN 1998 [3]. been derived from a special study carried out by Engineer- Dynamic earth pressure theories ing Seismologists for these types of structures. e commonly cited methods to assess seismic earth pres- Methodology sures (or “dynamic earth pressures”) in codes and texts e design methodology has been expounded by Steed- make basic assumptions about how the wall and soil inter- man [4] in a paper on seismic design of retaining struc- act, or together referred to as the “wall-soil-system”. ese tures. In addition, useful guidance, particularly in relation fall into the two extremes of system response: perfectly to the performance based design framework may be ob- rigid or no de€ection or displacement; and walls free to tained from PIANC [5] and to a lesser extent from EC8. A displace and/or de€ect until minimum (active) earth pres- typical loading diagram for a quay wall under pseudostatic sures occur. Selection of the appropriate method will de- loading is shown in Figure 1. Inertia loads on the retaining pend on the structure being considered, and the de!nition structure should also be considered in addition to inertia of the ULS for the particular structure. e bulk of this load on the soil and water. paper is geared to describe these two basic methods, their

SECED Newsletter Vol. 21 No. 4 October 2009 5 KP Varying with delta KPE Varying with delta

φ = 30°. Static Case φ = 30°. k h =0.25g Seismic Case 10 10

Log Spiral Mononobe Okabe 9 9 Coulomb Log Spiral 8 Eurocode 7 Annex C 8

7 7

6 6 P 5 PE 5 K K

4 4

3 3

2 2

1 1

0 0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Delta (°) Delta (°)

(a) (b) Figure 3. “Coulomb error” with large angles of wall friction. (a) Static passive earth pressures and (b) Seismic pas- sive earth pressures.

assumptions and limitations. It also presents alternative !ese assumptions and limitations a$ect the applicabil- approaches and guidance from the literature where they ity of the M-O method and raise a number of issues that are required. will be addressed to some extent within the remainder of the paper. !ese include: t the “Coulomb error” for passive earth pressures with Rigid walls free to displace to active-earth wall friction applied; conditions t walls that do not displace su"ciently to form an ac- tive wedge; Mononobe-Okabe method t the appropriate design acceleration coe"cient to ap- !e dynamic earth pressures at limit equilibrium may be ply to the wall back€ll; estimated using the classic “Mononobe-Okabe” (M-O) [6] t the point of action of dynamic thrust; [7] earth pressure method, where the applied inertia re- t cohesive soils; and sults in rotation of the principal stress (angle θ in Figure t water pressure e$ects. 2), enlarging the Coulomb active wedge (or decreasing !e longevity of the M-O method is due in part to its sim- for the passive case) at limiting equilibrium. !e method plicity but also its adaptability as various modi€cations enables calculation of the lateral earth pressure coe"cient have been suggested by researchers over subsequent years K, which is a ratio of horizontal to vertical earth pressure and eventually adopted in code guidance. It is also however at the wall-back€ll interface. In general, M-O provides a much abused and is o%en applied to situations which it was relatively good estimate (cf. model testing), provided the never intended – such as rigid walls, tied back and sti$ can- assumptions are met; chie#y that the wall is able to de#ect tilever embedded walls. Care should therefore be taken in or displace away from the soil to the minimum limit condi- its adoption to avoid gross errors. tion – thus mobilising full shear on the failure plane; sec- ondly, the soil is rigid such that the acceleration applied is Passive earth pressure error uniform; and thirdly, the soil is cohesionless and dry [8]. An important caveat with M-O for passive pressures with An important point is that the original method does not wall friction ratio δ/ϕ > ½ is that an unconservative error make any claim to where the point of action of the result- develops, inherited from Coulomb’s passive earth pressure ant force should be applied to the wall, an important con- equation which assumes a linear failure plane – in reality it sideration for assessing overturning stability. is curved [10]. Because of this, EC8 ignores wall friction en-

6 SECED Newsletter Vol. 21 No. 4 October 2009 tirely. An alternative to remove this slight conservatism for An alternative approach is to consider the modi€cation of both static and seismic case is to adopt a log-spiral shaped the M-O method by Koseki et al. [14], which considers the failure plane [11], refer Figure 3. Note that passive earth punctuated development of multiple wedges: pressure requires a signi€cant amount of wall de#ection to 1. !e initial active wedge occurs under static conditions be mobilised, which may well exceed the ULS criteria for with peak soil strength ( ϕ’peak ) and no inertia applied the wall. Reference to typical wall movements required to (kh = 0). Shearing continues along the pre-de€ned mobilise the active and passive earth pressure conditions is wedge until residual strengths ( ϕ’res ) are developed on provided in Eurocode 7 (EC7) [12], and NAVFAC [13]. A the failure plane. !e earth pressure increases, and this factor should be applied to limit the mobilisation of passive is used for static design. earth pressure for most design situations where the wall is 2. If the active wedge has not developed under static con- embedded in soil. ditions, it may do so when moderate earthquake inertia is applied to the same wedge geometry as 1. !ey rec- Large design inertia ommend a value of 0.2g for typical design as the critical !e M-O method becomes unstable when the sum of value at which the wedge forms (using M-O approach). interia angle θ and back€ll slope angle β exceed the an- Again, lateral earth pressures are considered using a gle of shearing resistance of the soil, ϕ’. In this case, the wedge geometry based on peak strengths, but with a re- square-root term on the denominator of the equation be- duction to residual strengths. comes complex, and cannot be solved. Matsuzawa et al. 3. !e peak earthquake inertia above this critical value [9] proposed a simpli€cation to avoid this problem which causes a secondary larger wedge to develop with peak is adopted by EC8 and is included in Figure 2. However, strengths. !is will cause larger earth pressures to act this results in very large active wedge angles, which may on the wall. be unrealistic in reality. !e recommended approach to !e approach has been referenced by the ISO dra% code avoid this problem is to consider the critical acceleration on performance based design in earthquake geotechnical at which the wall will begin to displace (i.e. factor of safety engineering [15], and Japanese Geocode 21 [16]. !e main = 1), and carry out a performance based assessment of the advantage over M-O is that consideration is made to the retaining structure (more discussion is provided on this as- development of active wedge prior to the earthquake oc- pect in a subsequent section on displacement estimation). curring, and consideration of the e$ect of strain so%ening

Figure 4: Damage to quay wall at Derince, Leman industrial facility. Kocaeli, Turkey Earthquake of 17/8/99. Image courtesy Earthquake Engineering Field Investigation Team (EEFIT), UK [24].

SECED Newsletter Vol. 21 No. 4 October 2009 7 from peak to residual along the pre-de€ned wedge. For tical use. For the undrained case ( ϕ’ = 0), the use of the large events, the peak strength of the soil may be consid- same approach or a trial wedge method may be adopted to ered in the formation of a new active wedge, which assists determine the dynamic earth pressure. !e Japanese Ports in avoiding the problems with M-O and large inertia. and Harbours design manual [23] provides an equation for determining the undrained dynamic earth pressure. Point of action of dynamic thrust Consideration should be given to reduction in the shear From studies in the literature three interacting compo- strength due to cyclic loading and generation of excess pore nents have been identi€ed to a$ect the point of action of pressures concurrent with the application of peak dynamic the dynamic earth pressure resultant (“dynamic thrust”). loading. An alternative is to consider an e$ective stress !ese are: based approach using a modi€ed M-O method, which is t Ground motion frequency discussed in the subsequent section. t Wall-soil system relative #exibility t Global movement of the wall-soil system. Dynamic water pressures Seed and Whitman [8] divided the M-O earth pressure For many retaining structures in terrestrial environments, resultant ( PAE ) into separate static ( PA) and incremental walls are designed with drainage to ensure the build up of dynamic (Δ PAE ) components PAE = PA + Δ PAE , and rec- static water pressures does not occur, however this is not ommended based on model testing, that the incremental possible for walls permanently below the water table – be dynamic component be considered to have a point of ac- they basements, underground structures, or marine struc- tion at 0.6 H measured from the base ( H = wall height). Of- tures such as quay walls. Here the presence of a permanent ten this advice was simpli€ed by adopting an inverted tri- static water table has a signi€cant e$ect on wall stability. angle for the dynamic earth pressure pro€le. Steedman and During earthquakes this has three important e$ects: Zeng [17] in a pseudodynamic analysis of the active wedge 1. !e weight vector in the M-O active wedge (refer Fig- showed the point of action was a function of the height ure 1) is almost halved due to buoyancy, thereby greatly of the wall, the shear wave velocity of the back€ll, and the increasing angle of the weight vector θ. !is e$ect was period of the ground motion; essentially showing that the reported by Matsuzawa et al. [9], who noted that for point of action of Δ PAE is dependent on the proportion of free draining conditions during cyclic loading (referred inertia that a$ects the upper third of the assumed active to as “dynamically pervious”, such that excess pore pres- wedge, where the bulk of the mass resides. !us the 0.6 H of sure generation under cyclic loading is minimal), this Seed and Whitman was at best an upper-bound and would factors the e$ect of horizontal inertia by approximately be conservative for design purposes, particularly in a per- 1.6 (dry unit weight/unit weight of water). For “dynam- formance-based framework. ically constrained” conditions (i.e. €ne grained deposits Veletsos & Younan’s [18] dynamic analysis of €xed base that during dynamic loading will result in essentially walls considers a visco-elastic medium without a pre- undrained behaviour), the factor is around 2.0 to 2.2, de€ned wedge. !eir work provides the point of action depending on the ratio of saturated unit weight to unit of the dynamic thrust for varying wall-soil system #ex- weight of water. It is recommended that care is taken ibilities from Wood’s ~0.6H for perfectly rigid wall [19], to derive the saturated and dry weights from soil phase down to less than ⅓ H for very #exible walls. Richards et al. relationships so that this ratio is correctly estimated as [20] investigated the mode of wall movement and showed the magnitude of the factor can have a signi€cant e$ect that rotation about the base caused the point of action to on the estimated dynamic pressures. drop to the lower ⅓, whilst for a translation mode it was at 2. If the soil remains dynamically pervious – e.g. open or 0.5H, and for rotation about the upper portion of the wall coarse granular €lls that allow the free #ow of water be- 0.67H. tween grains, dynamic water pressures should also be It is perhaps in light of this work from the preceding considered. !e method of Westergaard [25], developed decade that EC8 is the €rst modern code to depart from for free water bodies such as dam reservoirs, is adapted the 0.6 H or inverted triangle convention and recommends for the presence of soil grains – once again reference is applying the point of action of the dynamic component at made to [9] for guidance on this assessment; it is also a mid-height or 0.33H if free to rotate about the toe. useful guide as to whether the material will behave as “dynamically pervious” or “dynamically constrained”. Cohesive soils EC8 ignores the relative e$ect of soil grains on the dy- A number of methods have been adopted to consider co- namic water pressures in the back€ll, and it is true that hesive soils in the literature. Whitman [21] refers to some the e$ect is generally relatively minor. with a degree of scepticism. Recently Anderson et al. [22] 3. If the soil is dynamically constrained, a degree of excess consider soils with signi€cant c’ and ϕ’ using limit equi- pore pressures will be generated during cyclic loading, librium slope stability so%ware to determine the dynamic as with each successive cycle of loading pore pressures earth pressure coe"cient KAE , and provide charts for prac- generated due to volumetric changes on shearing can

8 SECED Newsletter Vol. 21 No. 4 October 2009 ∆σhe = .! khγH

∆P H k H E = γ h

0.58H

∆σhe = .! khγH

Figure 5. Dynamic earth pressure pro"le on rigid walls, after Matthewson et al. [26].

not dissipate in time before the subsequent cycle of an elastically constrained rigid wall: shearing occurs. !e e$ective stress path will thus pro- 2 gressively work its way towards the failure envelope, ΔPE = γ H kh Fp , (1) or phrased in total stress terms, the shear strength de- grades progressively with each subsequent cycle. !e where Fp is a dimensionless thrust factor and a function main concern in this instance is for soils whose steady of the sti$ness of the system, and H the height of the wall. state shear strength – that is, the strength of the soil For typical soils Fp is approximately unity, and is what EC8 a%er large shear strains have taken place – is less than assumes. !is may be thought of as a uniform inertia kh the in-situ strength, resulting in complete collapse of applied to a rigid block of soil of dimensions H×H. !e the deposit (#ow liquefaction). Most catastrophic fail- point of action for this force resultant was determined to ures of quay walls during earthquakes are due to this be 0.58 H, sometimes simpli€ed to 0.6 H. To determine the problem occurring in hydraulic €lls (e.g. Port Island, dynamic earth pressure pro€le, the recommendation of Kobe 1995, Derince Port, Turkey 1999 – refer Figure Matthewson et al. [26] is o%en adopted, where the dynamic 4). Preventing this problem should be the €rst priority earth pressure decreases linearly from the top of the wall at of the designer, principally through the use of ground 1.5 khγH to the base at 0.5 khγH – refer Figure 5. !e static improvement techniques such as stone/vibro columns. earth pressure component is based on at-rest pressures or !e condition where generation of excess pore pressures compaction pressures where the wall has back€ll present. build up progressively but do not necessarily lead to liq- By mere observation it is clear the dynamic load will be uefaction, should also be considered as strength is lost signi€cantly greater than the M-O method, and for many from the deposit, and dynamic earth pressures will be engineering problems will be very conservative. larger. One mitigating e$ect is that earthquake energy is consumed in order to shear the deposit, and by the time Fixed base !exible walls signi€cant pore pressures are generated, the peak load- It may be appreciated that there are many cases when nei- ing cycles may have already occurred. !e guidance to ther of the base assumptions recommended by codes (i.e. consider excess pore pressure ratio ru in the calculation M-O and Wood) are strictly applicable, in which case fur- of angle θ is provided by [10] as a modi€cation to refer- ther analysis beyond the basic code methods may be re- ence [9]. !e estimation of ru for a given deposit and quired. Veletsos and Younan [18] have shown Equation level of earthquake shaking is beyond the scope of this (1) to be valid for rigidly elastic wall-soil conditions, but paper. if the wall-back€ll, and/or wall-base are more #exible, the dynamic earth pressures can be considerably less. !ey provide tabulated Fp factors for varying soil-wall and soil- Elastically constrained rigid walls foundation #exibilities, demonstrating Wood’s solution for rigid systems, through to typical wall #exibilities ( Fp A rigid wall will typically not de#ect su"ciently to develop ≈ 0.5) and down to very #exible systems where Fp ≈ M-O an active or passive failure wedge. !us most codes recom- values of Δ KAE (i.e. KAE - KA, where the latter is based on mend Wood’s [19] solution (e.g. Eurocode 8 (EC8) [3]) for the Coulomb equation). Moreover Psarropoulos et al. [27]

SECED Newsletter Vol. 21 No. 4 October 2009 9 Comparison of Dynamic Earth Pressure Methods

Dynamic Earth Pressure (kPa) 0 10 20 30 40 50 60 70 0

1

2

3

4 Depth (m) Depth 5

6

7

8

Mononobe-Okabe c1929 acting at 0.5H (EN1998-5) Mononobe-Okabe c1929 acting at ~0.6H (Seed and Whitman 1970, FEMA369, upperbound) Simplified Wood 1973 Method using Code acceleration Simplified Wood 1973 using SHAKE acceleration Ostadan & White 1998 Method

Figure 6. Comparison of calculated dynamic earth pressures for a rigid embedded wall, founded on rock. Sand with ϕ’ =35°, Dr =65%. PGA (rock) = 0.15g. PGA (soil) = 0.2g (UBC97), PGA (soil) =0.27g (SHAKE). NB: M-O inter- pretations for comparison only, not strictly applicable for this wall type.

showed their results compared well to FE modelling. US between methods to derive the dynamic earth pressure Army Corps Engineers [28] have adopted this method for pro€le that was carried out for a typical project, including the structural design of L-wall stems. common interpretations of the use of M-O. !is method has been referenced by FEMA 450 [33]. Basement walls and tunnels A displacement based design approach to assess racking For basement walls, Ostadan & White [29] (see also Os- e$ects in box tunnels, by Wang [34], is based on a similar tadan [30]) developed a simpli€ed method, based on 1-D approach – parametric DSSI analyses have been performed site response analysis (such as using SHAKE [31], Oasys to derive a method to estimate the displacement and hence SIREN [32], or a similar program) following which a re- strains induced in the tunnel structure. !is approach sponse spectrum analysis is performed to derive the maxi- avoids the problem of estimating dynamic earth pressures mum dynamic earth pressure. !e €nal pro€le is obtained by either the M-O or Wood methods, which would be in- using a semi-empirical normalised earth pressure curve, appropriate for most tunnel structures. based on a series of dynamic soil structure interaction An alternative pragmatic design approach is to use a (DSSI) analyses. !eir results showed that depending on combination of 1-D site response analysis to obtain free input ground motion, soil and wall properties, the M-O €eld ground displacements, in combination with a pseu- and Wood solutions provided lower and upper bounds re- dostatic soil-structure interaction analysis performed in spectively. !us over-predictions by using Wood may be 2-D €nite element analysis so%ware such as Plaxis, Oasys assessed and reduced through this approach that consid- SAFE, Abaqus, FLAC etc., by a means of prescribed dis- ers the dynamic wall-soil system response to the design placements (e.g. Free et al. [35]). Further discussion on the ground motion inherently. Figure 6 shows a comparison seismic analysis of underground structures is provided by

10 SECED Newsletter Vol. 21 No. 4 October 2009 Hashash et al. [36]. Kontoe et al. [37] note that such simpli- ground motion from PGA may be taken depending on the €ed methods o%en provide reasonable results despite the height of the wall, and the ground motion characteristics inherent simpli€cations. – the latter simpli€ed as a ratio of the design spectra at a period of 1s and at PGA (i.e. relative contribution of long period to short period motion). Design inertia, phase e!ects and ampli"- cation Displacement estimation !e most common method used to reduce the design ac- E$ective acceleration celeration from PGA, is to utilise the Newmark sliding Sarma and Yang [38] attempted to add a theoretical basis block concept [42], originally developed for estimating co- to the common practice of applying a factor of ½ or ⅓ to seismic displacements of non-catastrophic embankment Peak Ground Acceleration (PGA) by considering the ac- failures. Richards and Elms [43] adopted this method for celeration required to generate 95% of the “energy” in an co-seismic sliding of retaining structures in an early appli- earthquake record (measured in Arias Intensity), dubbed cation of the performance based design concept. EC8 con- the A95 parameter. From the results of a study of 135 earth- siders this method inherently in the analysis procedure, by quake recordings, a best €t correlation of A95 = 0.675×PGA allowing a reduction in design inertia kh used in conjunc- was obtained. In contrast, Japanese practice adopts the tion with the M-O method, to 0.5×PGA, implicitly allow- proposal by Noda et al. [39] where kh is determined as fol- ing for small co-seismic displacements to occur during lows: the design earthquake event (typically less than 10cm and therefore negligible for most applications). A simple means

k = PGA ! if PGA ≥ ".#g, to estimate the order of magnitude of the displacements h ! ( ) (2)  PGA if PGA < ".#g. is also provided. However, the meaning of this reduction and the estimate of displacement is ambiguous for embed- !is is based on back-analysed estimates of kh from quay ded walls, and caution is advised. It is recommended that wall performance during earthquakes, and forms an upper- the mode of deformation be considered, and for embedded bound estimate, whilst a mean value is around 0.6×PGA walls the method of Veletsos & Younan [18] based on a (PIANC [5]). Al Atik and Sitar [40] found that a value of shear beam model may be more appropriate. 0.65×PGA provided reasonable estimates for matching the If one wishes to consider a Newmark analysis directly, use of M-O earth pressures centrifuge test results of em- or using one of the many published empirical methods to bedded walls. !e actual value will depend on the ground assess sliding displacements, the implicit reduction fac- motion, wall geometry, soil properties, and whether lique- tor in EC8 should be removed, and partial factors set to faction occurred. None of these empirical approaches con- unity, prior to calculation of the critical acceleration. For siders all of these factors systematically. tilting mode displacements, the method of Steedman and Steedman and Zeng [17] investigated the assumption of Zeng [44] (see also [45]) may be applied where the wall is in€nite sti$ness on dynamic active earth pressures through situated on a rigid founding stratum. For combined tilt- a pseudo-dynamic analysis method. !eir results suggest ing and bearing mode displacements (most applications) phase e$ects on earth pressures are of small signi€cance, a modi€ed method may be developed based on the same but the e$ects of ampli€cation were signi€cant, and clearly concepts. An alternative to the above simple rigid block both these e$ects would be more pronounced with large models is to use a fully dynamic numerical analysis, the walls where peak ground velocity (PGV) rather than PGA details of which is beyond the scope of this paper, but there will control the maximum inertia applied to the wall dur- are many examples in the literature. ing the design earthquake. EC8 is perhaps the €rst code to recommend that walls larger than 10m be considered in a site response analysis Conclusions for design. Arguably this should be performed as a 2-D site response analysis that may consider the change in pro€le !is technical note summarises the common methods to of the ground surface provided by the wall, account for assess dynamic earth pressures and their limitations. It also the relative sti$ness di$erences between wall and back€ll, provides reference to modi€cations and enhancements to and the dynamic response of any structures present that the base design methods in order to account for these limi- may in#uence the wall behaviour. A 1-D analysis cannot tations. Where possible, reference is made to Eurocode 8 capture these important aspects. Recently Anderson et al. to note to what extent these developments have been in- [22] presented the results of a parametric study to con- corporated into modern design. Hopefully this paper pro- sider these e$ects more systematically using QUAD4M – vides a useful reference for understanding the subtleties of an equivalent linear 2-D site response program [41]. !ey dynamic earth pressure evaluation and updates the reader produced a chart showing signi€cant reduction in design on recent developments.

SECED Newsletter Vol. 21 No. 4 October 2009 11 References for Foundation Designs Grounded on a Performance- based Design Concept . March ^``‚, Geocode ^\. JGS [\] TAYLOR M.L. Performance based design of retain- {``\-^``{ ing structures – part \: theory. Proceedings of the %&th [\ƒ] STEEDMAN R.S., & ZENG X. !e in#uence of phase New Zealand Geotechnical Society ())& Symposium . on the calculation of pseudo-static earth pressure on [^] TAYLOR M.L., KONTOE S., & SARMA S. Perform- a retaining wall. Géotechnique , \||`, #$ , No. \, \`_- ance based design of retaining structures – part ^: \\^. case study of a quay wall. Proceedings of the %&th New [\}] VELETSOS, A.S., & YOUNAN, A.H. Dynamic Re- Zealand Geotechnical Society ())& Symposium. sponse of Cantilever Retaining Walls. Journal of Geo- [_] BRITISH STANDARDS INSTITUTION. Eurocode &: technical & Geoenvironmental Engineering . ASCE Design of structures for earthquake resistance — Part \||ƒ, !€ (^):\‚\-\ƒ^ *: Foundations, retaining structures & geotechnical as- [\|] WOOD J.H., Earthquake Induced Soil Pressures on pects. BSI, ^``{, BS EN \||}-. Structures. Report No. EERL ƒ_-`, Earthquake En- [{] STEEDMAN R.S., Seismic design of retaining walls. gineering Research Laboratory, Calif. Inst. Tech., Pa- Proc. Instn Civ. Engrs Geotech. Engng , \||}, €, \^- sadena. \|ƒ_. ^^. [^`] RICHARDS, R., HUANG, C., & FISHMAN, K.L., [] PIANC. Seismic Design Guidelines for Port Structures. (\|||) Seismic Earth Pressure on Retaining Struc- International Navigation Association. Balkema, Rot- tures. Journal of Geotecnical & Geoenvironmental En- terdam, ^``\. gineering . ASCE \|||, !% (|), pp. ƒƒ\-ƒƒ} [‚] MONONOBE, N., & MATSUO, H., On the determi- [^\] WHITMAN, R.V. Seismic design of earth retaining nation of earth pressures during earthquakes, Pro- structures. Proceeding of the (nd International Confer- ceedings of the World Engineering Conference , , \ƒ‚. ence on Recent Advances in Geotechnical Earthquake \|^| Engineering & Soil Dynamics , Missouri, University of [ƒ] OKABE, S. General theory of earth pressure. Jour- Missouri, Rolla, USA. \||\. nal of the Japanese Society of Civil Engineers , Tokyo, [^^] ANDERSON D.G., MARTIN G.R., LAM I.P. & ! (\), \|^‚ WANG J.N. Seismic Analysis & Design of Retaining [}] SEED H.B., & WHITMAN R.V. Design of earth re- Walls, Buried Structures, Slopes, & Embankments . taining structures for dynamic loads. ASCE Special Vol.\. NCHRP Report ‚\\. Transportation Research Conference on Lateral Stresses in the ground & the de- Board ^``}, Washington D.C. sign of earth retaining structures , Cornell, New York, [^_] THE OVERSEAS COASTAL AREA DEVELOP- ASCE \|ƒ`. \`_-\{ƒ MENT INSTITUTE OF JAPAN, Technical standards [|] MATSUZAWA, H., ISHIBASHI, I., & KAWAMURA, & commentaries for port & harbour facilities in Japan. M., Dynamic soil & water pressures of submerged Eds: Goda Y., Tabata T., & Yamamoto S. OCDI ^``^. soils, Journal of Geotechnical Engineering , ASCE [^{] EARTHQUAKE ENGINEERING FIELD INFESTI- \|}, , No. \`, pp \\‚\-\\ƒ‚ GATION TEAM. 1e Kocaeli, Turkey Earthquake of [\`] KRAMER, S.L., Geotechnical Earthquake Engineer- %//&/33, EEFIT photo gallery. [Digital Media]. ^``_. ing . Prentice Hall \||‚, USA [^] WESTERGAARD, H., Water pressure on dams dur- [\\] KUMAR J. Seismic Passive Earth Pressure Coef- ing earthquakes, Transactions of the ASCE , ASCE €cients for Sands. Canadian Geotechnical Journal , \|__, " , pp {\}-{__. ^``\, €" , }ƒ‚-}}\. [^‚] MATTHEWSON, M.B., WOOD, J.H. & BERRILL, [\^] BRITISH STANDARDS INSTITUTION. E urocode J.B. Seismic Design of Bridges – Earth Retaining /: Geotechnical design- Part I general rules. BSI ^``{, Structures, Bulletin of the New Zealand Society of BS EN \||ƒ-\ Earthquake Engineering , \|}`. € (_), ^}`-^|_ [\_] NAVAL FACILITIES ENGINEERING COMMAND. [^ƒ] PSARROPOULOS, P.N., KLONARIS, G., & GAZE- Foundations & Earth Retaining Structures – Design TAS, G., Seismic earth pressures on rigid & #exible Manual . Naval Facilities Engineering Command retaining walls. Soil Dynamics & Earthquake Engi- (NAVFAC), Washington D.C. \|}‚, DMƒ.`^. neering . ^``, !% , ƒ|-}`| [\{] KOSEKI, J., F. TATSUOKA, Y. MUNAF, M. TATEYA- [^}] STROM R.W. & EBELING R.M. Seismic Structural MA & K. KOJIMA. A modi€ed procedure to evalu- Considerations for the Stem & Base of Retaining Walls ate active earth pressure at high seismic loads. Special Subjected to Earthquake Ground Motions. ERDC/ITL Issue Soils & Foundations , \||}, !, ^`|-^\‚. TR-`-_. US Army Corps of Engineers, Engineer [\] INTERNATIONAL ORGANISATION for STAND- Research & Development Center. ^``. Washington ARDISATION. Bases for design of structures — Seis- D.C. mic actions for designing geotechnical works . ISO [^|] OSTADAN, F. & WHITE, WH. Lateral Seismic Earth ^``, ISO ^_{‚| [dra%] Pressure: an Updated Approach, US-Japan SSI Work- [\‚] JAPANESE GEOTECHNICAL SOCIETY. Principles shop \||}, USGS, Menlo Park, CA.

12 SECED Newsletter Vol. 21 No. 4 October 2009 [_`] OSTADAN, F. Seismic Soil Pressures for Building [_}] SARMA, S.K., & YANG K.S. An evaluation of strong Walls: An Updated Approach , Soil Dynamics & Earth- motion records, & a new parameter A|. Earthquake quake Engineering , ^``, !% , ƒ}-ƒ|_. Engineering & Structural Dynamics , \|}ƒ, \, \\|- [_\] SCHNABEL, P.B., LYSMER, J., & SEED, H. B. SHAKE: \_^. A computer program for earthquake response analy- [_|] NODA, S., UWABE, T. & CHIBA, T. Relation between sis of horizontally layered sites . Report EERC ƒ^-\^, seismic coe"cient & ground acceleration for gravity Earthquake Engineering Research Centre, Univer- wall. Report of the Port & Harbour Research Institute, sity of California, Berkeley, CA. \|ƒ^. Japan \|ƒ, # , No. {, ‚ƒ–\\\ [in Japanese]. [_^] HEIDEBRECHT, A.C., HENDERSON, P., NAU- [{`] AL ATIK, L. & SITAR, N. Development of Improved MOSKI, N. & PAPPIN, J.W. Site response e$ects for Procedures for Seismic Design of Buried & Partially structures located in sand sites. Canadian Geotechni- Buried Structures, PEER Report ^``ƒ/`‚, Paci€c cal Journal , !& , \||`, _{^-_{ Earthquake Engineering Research Center ^``ƒ. [__] BUILDING SEISMIC SAFETY COUNCIL. NEHRP [{\] HUDSON M., IDRISS I.M. & BEIKAE M. User's Recommended Provisions for seismic regulations for Manual for QUAD5M: A computer program to evalu- new buildings & other structures (FEMA 5*)). Part (: ate the seismic response of soil structures using 7nite Commentary. National Institute of Building Sciences, element procedures & incorporating a compliant base. Washington D.C. ^``{ University of California at Davis \||{. [_{] WANG J.N. Seismic Design of Tunnels. A Simple [{^] NEWMARK N. M. E$ect of earthquakes on dams & State-of-the-Art Design Approach. Parsons Brincker- embankments. th Rankine lecture. Géotechnique . ho$ Monograph ƒ. \||_, New York, NY. \|‚, % , ^, \_|-\‚` [_] FREE M.W., PAPPIN, J.W., SZE J.W.C., & McGOW- [{_] RICHARDS, R., & ELMS, D., Seismic Behaviour of AN M.J. Seismic design methodology for buried Gravity Retaining Walls, Journal of the Geotechnical structures. Proceedings %5th South East Asian Geo- Engineering Division . ASCE \|ƒ|, % , GT{, {{|-{‚{ technical Conference , ^``\. pp `|-\{. [{{] STEEDMAN, R.S., & ZENG X., Rotation of Large [_‚] HASHASH Y.M.A., HOOK J.J., SCHMIDT B., & Gravity Walls on Rigid Foundations under Seismic YAO J. I.C. Seismic design & analysis of underground Loading, In: Analysis & Design of Retaining Structures structures. Tunnelling & Underground Space Technol- Against Earthquakes . Ed: Prakash S., ASCE \||‚. ogy , ^``\, ' , ^{ƒ-^|_ [{] ZENG X. & STEEDMAN, R.S., Rotating Block Meth- [_ƒ] KONTOE S., ZDRAVKOVIC L., POTTS, D.M., od for Seismic Displacement of Gravity Walls, Jour- MENKITI C.O., Case study of seismic tunnel re- nal of Geotechnical & Geoenvironmental Engineering. sponse. Canadian Geotechnical Journal , ^``}, #%, ASCE ^```, !' (}) ƒ`|-ƒ\ƒ. \ƒ{_-\ƒ‚{.

SECED Newsletter SECED The SECED Newsletter is published quaterly. Contributions SECED, The Society for Earthquake and Civil Engineering are welcome and manuscripts should be sent on a CD or Dynamics, is the UK national section of the International by email. Diagrams, pictures and text should be in sepa- and European Associations for Earthquake Engineering and rate electronic €les. Copy typed on paper is also accept- is an aliated society of the Institution of Civil Engineers. It able. Diagrams should be sharply de€ned and prepared in is also sponsored by the Institution of Mechanical Engineers, a form suitable for direct reproduction. Photographs should the Institution of Structural Engineers, and the Geological be high quality. Colour images are welcome. Diagrams and Society. The Society is also closely associated with the UK photographs are only returned to authors on request. Earthquake Engineering Field Investigation Team. The objective of the Society is to promote co-operation in the Contributions should be sent to the Editor of the Newsletter, advancement of knowledge in the €elds of earthquake Andreas Nielsen. engineering and civil engineering dynamics including Email blast, impact and other vibration problems. For further [email protected] information about SECED contact: Post The Secretary Editor SECED Newsletter SECED c/o The Secretary Institution of Civil Engineers SECED One Great George Street Institution of Civil Engineers London, SW1P 3AA, UK One Great George Street Or visit the SECED website: London, SW1P 3AA http://www.seced.org.uk United Kingdom

SECED Newsletter Vol. 21 No. 4 October 2009 13