ABSTRACT: Numerous analytical approaches exist for the prediction of the settlement improvement offered by the installation of granular columns in weak or marginal soil deposits. In this paper, an appraisal of some of the more popular settlement prediction methods applicable to vibro-replacement design is carried out. The settlement improvement factors calculated using the different design methods are compared with both the ‘bottom feed’ and ‘top feed’ data from a previously published databas e of settlement improvement factors for stone columns in soft clays and silts. The calculated improvement factors are plotted as a function of the main variables in vibro-replacement design, namely the area-replacement ratio (a measure of the amount of in- situ soil replaced with stone), and the modular ratio, which relates the stiffness of the column to that of the in-situ soil. In addition, finite element analyses have been carried out using PLAXIS 2D for comparison with both the field data and the design method predictions. Load settlement behaviour (primary settlement) has been analysed using the Hardening Soil Model to model the behaviour of both the column material and the treated soil. The majority of the design methods appear to predict similar settlement improvement factors. Purely elastic methods overestimate the settlement improvement factor for large modular ratios while it appears that better matches between finite element and analytical predictions are found at higher modular ratios, as the analytical methods assume a significant bulging mechanism which is more prevalent in soft soils. KEY WORDS: Stone Columns; Settlement Improvement Factor; Analytical Design Methods; Finite Element Analyses. 1 INTRODUCTION The use of Vibro Stone Columns (VSCs) as a ground improvement technique has gained popularity in recent years. It is now widely accepted that stone columns reduce settlement [1, 2], improve bearing capacity [3], and accelerate consolidation [4, 5, 6]. In addition, VSCs provide a suitable economic alternative to piled foundations, while the construction time associated with VSCs is also significantly shorter than that associated with piling and other alternative foundation solutions. The vibro-replacement process and equipment have been described in detail by [7] and [8]. Vibro stone columns can be constructed using either a top feed method or a bottom feed method using either wet or dry jetting processes. In the case of the top feed method (e.g. Figure 1), stone is tipped into a hole formed by a vibrating poker, whereas for the bottom feed method, stone is added through a delivery tube along the side of the poker and exits at the poker tip. In both cases, compaction is carried out in stages from the base of the hole upwards. Air is used to aid construction and maintain stability of the hole for the dry method whereas water is used for the same purpose where the wet method is concerned. Aggregate of size 40-75mm is used for the top feed method whereas for the bottom feed method, the size of the aggregate used ranges from 15-45mm [9]. The wet top feed method is suited to soft cohesive soil deposits where the ground water level is high while the dry top feed method is mainly used for firmer soil deposits with lower ground water levels. The dry bottom feed method is now the preferred construction technique in softer cohesive soil deposits (its use has largely replaced the wet top feed method [10] since its development in the 1970s) and enables columns to be constructed in soils with low undrained shear strengths, c u << 15-20kPa [10]. Use of the wet method has weaned in recent years with the disposal of ‘flush’ becoming an ever -increasing problem. Figure 1. Schematic of wet top feed method - Raju et al. [11] McCabe et al. [10] compared predicted settlement improvement factors using Priebe’s [12, basic settlement improvement factor, n 0 ] method for a column friction angle, φ’ c , of 40 o with measured settlement improvement factors for the different methods of column construction. In that research, bottom feed columns tended to behave better than predicted, whereas for top feed columns, the opposite was generally the case. Their comparisons indicated that a design friction angle, φ’ c , of 40 o is a conservative assumption for the bottom feed system; however, for the top feed system, φ’ c = 40 o may lead to over-predicted settlement improvement factors. The purpose of this study is to review and evaluate a selection of the more popular analytical settlement prediction methods. The predictions are compared with the field data from [10]. A numerical study is also conducted for Appraisal of Current Settlement Prediction Methods Applicable to Vibro Replacement Design B.G. Sexton 1 , M.M. Killeen 1 , B.A. McCabe 1 1 College of Engineering and Informatics, National University of Ireland, Galway, Ireland email: b.sexton1@nuigalway.ie, michealkilleen@gmail.com, bryan.mccabe@nuigalway.ie