Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.716042
Title: Constitutive modelling of soils and fibre-reinforced soils
Author: Bower, Thomas
Awarding Body: Cardiff University
Current Institution: Cardiff University
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
This thesis covers two major themes: the first relates to the development of a constitutive soil model, and the second is the development of a model to predict the behaviour of fibre-reinforced soils. The hardening soil (HS) model is an advanced constitutive plasticity model which is applicable to the analysis of many soil types including sands and clays (Schanz et al., 1999; Benz, 2007). This model is explored in depth, and several improvements to the model are proposed. The first improvement is the formulation of a new hardening shear yield surface to replace the previous hardening shear surface and failure surface. The second is the implementation of the model in a robust return mapping scheme. The scheme used is the closest point projection method of Simo and Hughes (2006), which is tailored specifically to this implementation of the model. This constitutive soil model and return mapping scheme is hereinafter referred to as the HS-LC model. The HS-LC model is then used in finite element analyses and compared to published experimental and predicted data obtained from the prior versions of the HS model. It was found that the new HS-LC model was able to reproduce results from both the experimental data and the previous models. The numerical stability of the proposed model was also tested with a step size study, a mesh density study, and investigation of convergence rates for simulations. The second main theme of this thesis is fibre-reinforced soils. The motivation for reinforcing soils is first explored, then a literature review is conducted on different reinforcement types; focussing on fibre-reinforcement. Experimental results from the literature are then discussed, along with several models which predict the behaviour of fibre-reinforced soil. Results from an experimental study (Chatzopoulos, 2015; Wang, 2015) of fibre-reinforced sand are presented and discussed. A series of triaxial compression tests were conducted; in which fibre content, fibre length, and fibre type were varied. It was found that some fibre types added a significant amount of strength to the soil, where other types added little strength. All fibres studied here were manufactured from polypropylene, it was therefore concluded that the geometry had a significant impact on the strength contribution for the composite. This study also found that when the fibre diameter was too large or too fine relative to the sand particle size, then the particles would not bind to the fibre. An additional unique study was also carried out in relation to fibre composites. This was a series of fibre pullout tests, where a fibre was pulled through a prescribed length of compressed, dry soil. In this test it was found that the peak bond strength was linked to the compressive stress acting on the fibre. A novel fibre-soil composite model was also formulated which is based on micro-mechanical relationships between soil and fibre, from findings of the literature review and the experimental study. The proposed model is based on the well known shear lag model (Cox, 1952) and is modified to also include the effects of fibre debonding. The model takes the form of a representative volume element (RVE), which is homogenised using a statistical approach (Bazant and Oh, 1986). The proposed fibre model is then combined with the new HS-LC model using the rule of mixtures. The composite model is then used to predict the behaviour of the tests in the experimental study. Predictions of the triaxial tests closely matched the experimental results in the shear stress response, however, were less accurate for the prediction of volumetric strains. It was concluded that further work is required in the development of this model before it can be considered in routine design.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.716042  DOI: Not available
Keywords: TA Engineering (General). Civil engineering (General)
Share: