TECHNICAL NOTE Effect of berms on sheet-pile wall behaviour M. GEORGIADIS  and C. ANAGNOSTOPOULOS  KEYWORDS: model tests; numerical modelling & analysis; sands; sheet piles & cofferdams. INTRODUCTION The use of berms located in front of temporary sheet-pile walls (Fig. 1) can often prevent excessive wall movements and minimize excavation costs. Vertical soil stresses are increased, resulting in larger passive earth pressures on the embedded part of the wall and improved wall stability. Berms can either be left in place throughout the excavation period or, more often, be removed in small sections as permanent works progress and struts are installed. Current design practice is highly empirical and based on limit equilibrium analysis, while wall movements are kept within acceptable limits by the use of appropriate safety factors. In the approximate methods which are usually employed, the berm is treated as either an effective surcharge applied at the ®nal excavation level or as an increase in the original ground level on the passive side of the wall (Fleming et al., 1985). Both the magnitude of the effective surcharge and the effective excavation level are determined empirically, taking into ac- count the berm geometry. Similarly, based on limit equilibrium, Construction Industry Research and Information Association (1993) gives simple rules for adjustments to the passive earth pressures that act in front of the wall, taking into account the berm geometry. A graphical method (Culmann line) for calculating the passive soil resistance offered by a berm is described in the Naval Facilities Engineer- ing Command (1982) design manual. Available research on the effect of berms on reducing wall de¯ections is rather scarce. Cole & Burland (1972) and Burland et al. (1979) have reported numerical results and ®eld measurements demonstrating the bene®cial effects of the use of berms for propped walls in overconsolidated clay. Clough & Denby (1977) used ®nite-element ana- lyses to quantify the effect of the size of the berm on limiting deformations of temporary walls in soft to medium clay. This paper presents results of ®nite-element analyses and an experimental investigation of canti- lever sheet-pile walls in sand, demonstrating the effect of berms on the behaviour of the wall. Based on these results, a procedure for assessing the effect of berm geometry, excavation depth, wall stiffness and sand density on the reduction of wall movements and bending moments is proposed. NUMERICAL ANALYSIS Parametric studies of typical sheet-pile wall sup- ported excavations were performed using the ®nite- element computer program PLAXIS (Vermeer, 1996). This program employs up to 200 15-noded triangular elements, as well as interface elements between the soil and the wall. To model the exca- vation process, the appropriate elements are disac- tivated and the resulting out-of-balance forces are applied to the mesh. The effect of excavation depth, berm height, berm width and relative sheet-pile wall/soil stiff- ness on horizontal wall displacements was analysed using the geometry shown in Fig. 1. Excavation was performed to a maximum depth of 3 : 0m by progressively removing rows of 0 : 5 m thickness in Georgiadis, M. & Anagnostopoulos, C. (1998). Ge Âotechnique 48, No. 4, 569±574 569 Manuscript received 8 May 1997; revised manuscript accepted 30 October 1997. Discussion on this paper closes 6 November 1998; for further details see p. ii.  Aristotle University of Thessaloniki. H H 1 H 0 d b y b E I L = 5 m Fig. 1. Numerically analysed sheet-pile wall Géotechnique 1998.48:569-574.