Possibilities yield by Eurocode 7 for design of geotechnical structures in urban areas Possibilités ouvertes par Eurocode 7 pour la conception d’ouvrages géotechniques en zones urbaines C. Bauduin BESIX, Brussels, Belgium University of Brussels, Belgium

ABSTRACT The design of geotechnical structures in urban environment is often governed by serviceability limit states of the structure to be build and of adjacent structures. The present paper summarizes the ways proposed by Eurocode 7 to check SLS. The possibilities offered by the code through the Observational Method and the FEM are briefly discussed. This brief overview leads to some topics for discussion. RÉSUMÉ La conception d’ouvrages géotechniques en environnement urbain est bien souvent déterminée par les états de service de la structure considérée, ou par ceux de structures avoisinantes. Le document présent résume les méthodes proposées par Eurocode 7 pour la vérification d’ouvrages aux états de service. Les possibilités ouvertes par le Code au travers des Méthodes Observationnelles et des Eléments Finis sont brièvement discutées. Cette brève revue introduit quelques sujets de discussion. Keywords: Eurocode 7, Serviceability Limit States, Observational Method, FEM 1 INTRODUCTION

-

The Eurocodes, and thus EN 1997-1, are Limit State Design Codes, requiring the geotechnical structure be verified for the Ultimate Limit States (ULS) in the ground and in the structure as well as Service Limit States (SLS). The ULS in the ground and in the structure shall be checked according to one of the three the Design Approach DA1, DA2 or DA3 (EN 1997-1). The National Annex indicates the selected Design Approach(es). The design and construction of geotechnical structures in urban environment, such as deep excavations, heavy loaded foundations…) are often governed by deformation criteria related to Serviceability Limit States (including damage) of the intended or of adjacent structures. Users of EN 1997-1 may object to the code that more information is given concerning the checks against ULS than against SLS. Most of the development work on Eurocode 7 concern the evaluation of ULS design (see e.g. Orr, 2005 and the paper review by T. Orr for this discussion session). In this respect, the usability of EN 1997-1 for the design of geotechnical structures in urban areas is sometimes questioned. The objective of this document is to introduce the discussion regarding:

-

the methods proposed EN 1997-1 for checking against SLS the opportunities yield by the code for designing geotechnical structures in urban environment

2 METHODS FOR CHECKING SERVICEABILITY LIMIT STATES EN 1997-1 allows checking Serviceability States in two ways: - by calculating the design values of the effect of the actions Ed (eg deformation, differential settlement,…) and comparing them with limiting values Cd by using: Ed ≤ Cd (1) - by an alternative method based on comparable experience, by showing that a sufficiently low fraction of the ground strength is mobilized to keep the deformations within acceptable limits Analytical methods, semi-empirical methods and numerical methods may be used to verify the equation (1) and to calculating the fraction of the ground strength that is mobilized in the service state. The design values of actions and of material properties will normally be taken equal to their characteristic values. In cases where deformations are calculated, a combination of upper and lower values of the deformation moduli should be considered, to account for local variations of the

ground properties. In case of uncertainty of the value of the deformation moduli, it is sound engineering practice to check the sensitivity of the design for a range of values. When using eq (1), the limiting values of deformations should be specified as a design requirement for the supported or for the adjacent structures subjected to ground displacements. The code lists a series of items to take into account when establishing limiting values of movements. Informative Annex H of EN 1997-1 gives some indications which can be used as guidelines in absence of limiting values of structural deformations. EN 1997-1 clearly restricts the use of the alternative method to situations where: - Comparable and successful experience supports the use of the method - The structures are conventional - There is no need of the knowledge of the value of the deformation to check the SLS The alternative method is referred to for spread foundations, pile foundations and retaining structures in the corresponding chapters of EN 19971. The code gives however no information on values of what is “sufficiently low fraction”. Frank et al. give some more information on the combinations of actions to be used when applying eq (1) and the alternative method. 3 DISCUSSION AND COMPARISON WITH THE EXISTING PRACTICE Both methods described above were applied long before the existence of Eurocode 7. In this respect, the code is in full continuity with the past experience. The strength of EN 1997-1 is to require clearly the need for checking against both Ultimate and Serviceability limit states, as two different and complementary steps in the design. Too often, existing calculation rules, especially those based on “existing experience” are not clear to which limit state they intend to check. This may lead to unreliable design, in terms of “safety” as in terms as well as in terms of “evaluation of serviceability requirements”. It is important to note that in the present experience with global safety factors, their value, which is mainly based on experience, besides providing “safety”, also implicitly aims to avoid unacceptable deformations: global safety factors often aim also to cover serviceability limit states. The separate checks against ULS and SLS implies

that attention should be paid when setting the values of partial factors for ULS design in the National Annex, to avoid that they become in fact hidden mobilization factors, which is against the spirit of EN 1997-1 For each design situation thus, the ULS shall be checked, even if the final design is governed by serviceability criteria. After defining the general framework through which SLS shall be verified and the specific aspects to which attention has to be paid, the code however remains scarce in giving practical information on e.g. limiting values of movements or the sufficiently low fraction of mobilized strength. It should not be wise, but even misleading, in the Code to cover completely all these aspects, as this will never be universally applicable; the Informative Annexes of EN 1997-1 do not aim to give a complete answer to all questions, but indicate paths which may be elaborated in more details through the National Documents. The National Annex may indeed give the decision of a particular country about the application of the informative annexes of EN 1997-1 in that country. To do that, the National Annex must state, depending on the case, whether the Informative Annex considered shall be regarded as normative at the national level, or shall not be used at the national level. If the National Annex is silent on the use of an Informative Annex, it remains part of the standard. If some guidance on the same subject as that contained in the Informative Annex is required, then a document can be published, separately from the national annex and be referenced in the National annex, as a complementary nonconflicting reference. (Frank et al., 2004). The informative Annex H of EN 1997-1 may be complemented by such a document, containing more precise information on limiting values of displacement if wished nationally. Of course, the code leaves open to establish limiting values for displacements, vibrations etc for each design depending e.g. on the items listed in the code. Relevant literature, existing and validated experience etc may be used to set the appropriate limiting values for a particular design. It should be noted that criteria set on an adjacent structure often need to be translated to criteria to be applied to the structure being designed: e.g. limiting values of (differential) settlements of a building near a deep excavation may need to be translated to limiting values for the horizontal of the retaining wall. This step of course may yield significant uncertainties.

Figure 1. Existing experience was not sufficient to avoid an Ultimate Limit State

4 THE OBSERVATIONAL METHOD EN 1997-1 introduces design by the Observational Method, in which the design is reviewed in a planned manner during the course of the construction and in response to the monitored performance of the structure. Especially when the prediction of geotechnical behaviour is difficult, the Observational method may be appropriate. Observational methods can be applied “ab initio” or as “the best way out”. The use of appropriate monitoring has been used since decades during the construction of geotechnical structures in urban environment, especially where deformation or vibration criteria are set to existing structures which might be affected by the construction of new structures. Traditionally, the monitoring aims to check the design (which usually was performed using “safe” values of the ground parameters) and to warn for possible hindrance or damage to existing structures. In this traditional practice of monitoring during construction the design was not reviewed during the construction, except if the monitored performance indicates (risks of) an undesired behaviour: the monitoring was not intended for optimization of the design during construction. In this respect, one could

state that much experience is available in “checking the design by monitoring”, but that experience has to be gained in the use of the full power of the Observational Method. Recent work (CIRIA, 1999, IREX, 2005, GeoTechNet Project 2001-2005) has been performed in this sense, and could be used as a base to prudently introduce the Observational Method in the design of geotechnical structures in urban areas. Point of special attention in urban areas, is the large number of owners and stakeholders, which not necessarily intent the same objective with the monitoring and do not gave the same motivation for using the Observational Method in its full power. 5 POSSIBILITIES OFFERED BY EN 1997-1 FOR DESIGNING STRUCTURES IN URBAN ENVIRONMENT As explained here-above, EN1997-1 requires for both ULS and SLS checks: - according to the relevant DA for ULS verification and - using one of the two ways for SLS verification Analytical or semi-empirical methods very often are not sufficient for the design of geotechnical structures in urban environment, especially when soil-structure interaction or the effects of ground movements on adjacent structures play an important

Simulation of stress history All loads, geometrical and ground parameters: characteristic value Characteristic stress field for construction stage i Simulation of next construction stage

SLS verification for stage i

ULS verifications for stage i Introduce Overdig Increase qk by γQ / γ G =1.11 Characteristic value of shear strength parameters

Introduce Overdig All parameters characteristic Characteristic stress field for construction stage i+1

Results: Displacements, M and A at (SLS) characteristic stress field

Calculate M & A and multiply by load factor γ G =1.35

Stepwise reduction of shear strength parameters to design value (γ M)

Increase qk by γQ =1.3 Characteristic stress field

Results: Md, Ad for DA1/1 & DA2*

Results: Check failure in the ground Md, Ad for DA1/2 & DA3

Figure 2. Flow chart for FEM simulation of stress history and SLS and ULS check (Bauduin et al 2003)

role in the design. In these situations numerical methods such as FEM are appropriate: as they are developed to model as accurately as possible stresses and deformations of the ground and the structures at service state, they are therefore a powerful tool to check the design of geotechnical structure against SLS. EN 1997-1 allows the use of numerical methods, but is scarce in giving information on the manner to use FEM in the Eurocode 7 framework. Bauduin et al (2003) propose the method illustrated in figure 2. The whole construction process of the geotechnical structure is simulated using characteristic values of actions and ground parameters (thus simulating the most probable stress history), and in the situations where the design needs to be checked against ULS the actions are increased and the shear strength is reduced to their design values using the relevant partial factors; uncertainties on geometrical parameters (e.g. over dig) may be introduced in that step. This method allows to check for ULS (according to the selected Design Approach) and for SLS. Further details and a discussion on the use of FEM for ULS design in the Eurocode framework can be found e.g. in Bauduin et al (2005). In the schema above, it should be noted that the simulation of the stress history using characteristic values of all parameters is usually very close but not necessarily equal to SLS checks. This might play a role e.g. when the characteristic values of ground parameters or load combinations for the FE calculations are selected. Very accurate calculations of deformation may demand for advanced ground model, which are not necessarily needed when the FE calculations aim mainly to check for ULS.

The discussion on the limiting values of movements however remains. The efforts put in the FE model have to be in balance with the accuracy by which these limiting values can be set. As FE codes become users-friendly, sensitivity analyses should be performed when values of parameters or requirements are uncertain. Combination with monitoring, may be in the spirit of the Observational Method, is clearly supported in the framework given by EN 1997-1. 6 TOPICS FOR DISCUSSION This brief, thus incomplete, review focusing on SLS design according to EN 1997-1 leads to several topics for discussion of which some are listed below: 1. Should nationally referred documents be complemented by more information regarding limiting values of movements, or should these requirements left to the designer of each specific project? 2. Should nationally referred documents be complemented by more information regarding limiting values of strength mobilization in service state to clarify what is “sufficiently low fraction”? 3. Which are ways to promote the use the Observational Method for the design of geotechnical structures in the complexity of urban environment? A constructive discussion based on experience in different countries on these points should certainly help to evaluate and to implement the opportunities offered by EN 1997 for improving the design of geotechnical structures in urban environment.

REFERENCES Bauduin, C., De Vos, M., Frank, R. (2003) ULS and SLS design of embedded walls according to Eurocode 7, Proc XIII ECSMGE, Prague, Vol 2 pp 41-46 Bauduin, C., Bakker, K.J., Frank, R. (2005) Use of Finite Element Methods in Geotechnical Ultimate Limit State Design Proc XVI ICSMGE, Osaka pp 2775-2779 CIRIA (1989) The observational method in ground engineering: principles and applications CIRIA report 185 EN 1997-1 (2004) Eurocode 7 part 1, General Rules CEN/TC 250/SC7. European Committee for Standardization, Brussels Frank, R., Bauduin, C., Driscoll, R., Kavvaddas, M., Krebs Ovesen, N, Orr, T., Schuppener, B. (2004). Designer’s Guide to EN 1997-1: Geotechnical Design-General Rules, Thomas Telford ISBN 07277-3154-8 Huybrechts N ed, GeoTechNet Projects (2001-2005) GTC22000-33033 WP3 Innovative design tools in geotechnicsObservational Method and Finite element Method, BBRI http://www.geotechnet.org IREX (2005) La méthode obeservationelle pour le dimensionnement des ouvrages. Presses de l’Ecole Nationale des Ponts et Chaussées Orr, T.L.L. (2005) Design examples for the Eurocode 7 Workshop, Proceedings of International Workshop on Evaluation of Eurocode 7, Dublin March-April 2005, Department of Civil, Structural and Environmental Engineering, Trinity College Dublin