Modelling of Peat Compressed under Sand Bodies

Abstract

This study being part of the Weesp Bloemendalerpolder Project, is concentrated on the compression of a peat layer under the weight of a trial sand embankment; examined under two scopes, the laboratory K0-CRS test and the field settlements. Their interconnection is also evaluated. Peat being an organic material with many distinctive characteristics such as time dependency due to the visco-plastic nature, a high water content, fibrosity, etc, required a broad literature study concentrated on its nature, dominant properties and the numerical models that can sufficiently capture such behaviour. Classification laboratory tests (pycnometer and loss on ignition) together with correlations found in the literature that are commonly used in the engineering practice highlight the interdependency of some characteristics in the overall response of the peat, and aided in increasing the trust in the available dataset. Two series of K0-CRS tests were conducted, prior and post the embankment construction, providing an insight into the changes of the material properties due to compression. The laboratory tests are then simulated with the Soft Soil Creep (SCC) model with PLAXIS software with two approaches; using the soil Test Facilities and the Finite Element mesh analysis. Their ability to model the large deformations together with aspects such as samples saturation, parameter transformation and the deviation between the laboratory and fitted values are discussed; with the objective of creating a dataset capable of describing the material in situ. Shifting the focus to the field situation, the settlements are calculated with the SSC model and compared to the monitoring data. The fit illustrates the value and utility of the laboratory test calibration. Alternative settlement calculations are conducted using the abc Isotache model implemented in D-Settlement software, where the settlements produced are fitted to the monitoring data, and the deviation of the parameters is evaluated in comparison to both the former settlement prediction and the laboratory derived values. The water content and permeability are found to play vital role on the layer compressibility. The local character and heterogeneity of the properties are highlighted. Finally the inadequate strength parameter determination by the compression tests is emphasized and the influence on the laboratory tests and the horizontal displacement of the compressed material in the field is illustrated.