About 50 engineers from eight countries attended the workshop on sheet pile wall field testing held by the Centre for Civil Engineering Research and Codes and Delft University of Technology in the Netherlands on 25 and 26 May.
The workshop focused on the results of a sheet pile wall field test carried out at Pernis near Rotterdam between February and October 1999. The main topics of research were: Short- and long-term behaviour of sheet pile walls in soft soils with a high groundwater level.
The consequences of the generation of a plastic hinge in the sheet pile.
To obtain further insight into the problem of possible loss of stiffness in single and double U-piles, due to lack of shear capacity in the interlocks.
The layout of the test site is shown in Figure 1. The site consisted of soft clay and peat with a high groundwater level. The test set-up consisted of four walls installed in an approximate square with sides about 12m long.
The north test wall consisted of10 double Zpiles (Arbed AZ13) to a depth of 18m. The south test wall consisted ofnine double U-piles (Hoesch L607K) where the interlocks were welded before driving. Two rows of bentonite columns were installed at the active side perpendicular to the north and south test walls to enable the trial to be modelled realistically by plane strain analysis. The east and west walls were each formed by 20 single U-piles (British Steel LX32) to a depth of20m.
In the months before the workshop, engineers were invited to predict the behaviour of the test walls. Each participant received extensive site investigation data, laboratory tests results and details of the test to assist in their prediction.
The predictions were divided into two types: Type 1 focused on typical design practice while Type 2 focused on the researchbased aspects of sheet piling. About 50 engineers from eight countries attended. Three types of prediction modelling methods were used:
subgrade reaction models continuum models finite element models.
Overall, the bending moment and displacement predictions were deemed adequate but the pore water pressure and the earth pressure predictions differed greatly from the measured values. A number of predictors questioned the pore water pressure and earth pressure measurements whose accuracy had yet to be verified.
The predicted sheet pile behaviour was dominated by the manner in which the peat layer was treated. Brian Simpson of Arup Geotechnics, who presented one of the more successful predictions obtained using the FREW program, admitted that great uncertainty lay in the modelling of the peat layer.
In addition to the actual test results and predictions, the types of model used were discussed. While finite element models were considered the most accurate, their use was thought to be too time consuming and difficult to set up for the average design engineer.
The subgrade reaction models were easy and quick to formulate but the results are questionable. The continuum model (eg FREW) is regarded as the 'middle-of-the-road model', being easy and quick to set up. Its accuracy lies between the other two models.
A very informative presentation was given by Alex Schmitt ofArbed, Luxembourg, on the new requirements for sheet pile design according to Eurocode 3, part 5. The use of plastic design was highlighted during the presentation by Jorgen Steenfelt (TU Denmark) on Danish design practice. Arjen Kort ofDelft University of Technology highlighted the fact that during the test, good measurements of the effect of rotation of the U-section sheet piles were obtained. Significant reductions in stiffness were noted, even when the U-sections were welded in pairs.
Lively discussion included the various methods used, particularly in relation to accuracy and time. The effects of the peat layer were also debated, with regard to the assumed stiffness and whether the material should have been treated in a drained or undrained manner. The advantages and disadvantages of plastic design were also discussed.
It was clear that geotechnical engineers have a long way to go before the behaviour of sheet pile walls is properly understood, especially in more complex soil conditions. A follow-up workshop was proposed to further discuss some of the issues towards the end of this year.