Multi-scale modelling of the coupled effects of chemical, mechanical and physical processes on concrete degradation (PhD)

Predicting the performance of concrete structures is complicated by the complexity of the material, which has a heterogeneous microstructure and displays composite behaviour at a series of length scales. In particular, the overall transport and mechanical behaviour of concrete is strongly conditioned by its heterogeneous microstructure, which determines the randomness of the overall transport and mechanical variables.  Permeability, diffusivity, initiation and progression of cracks within concrete are significantly controlled by this randomness.  Multi-scale modelling is an approach which is followed to assess the large-scale durability of concrete. Multi-scale modelling provides a methodology to systematically incorporate detailed information about processes occurring at smaller scales into governing equations at larger scales.  

The principal objective of this research project is to develop a multi-scale modelling framework to incorporate the coupled effects of chemical, mechanical and physical processes on chemical degradation of concrete in order to assess the long-term, large-scale durability of concrete under conditions representative for a near-surface radioactive waste disposal facility based on a multi-scale modelling approach. This project essentially considers PhD research project 1 and 2 as a starting point and runs partly in parallel.

The approach to multi-scale modelling will essentially involve integration of a series of models applicable for different scales, i.e. micro-scale, meso-scale and macro-scale and in essence follows a hierarchical modelling approach. The following key steps are envisaged in realising a multi-scale model:

·          Microscale modelling of unhydrated/hydrated cement paste comprising micro-porosity and solid phases of hydrated cement paste. The starting point for this task will be work package 1 but will elaborate further relations between the degree of chemical degradation, microstructural changes and permeability.

·          Mesoscale modelling of hydrated cement paste, sand and voids (mortar) or mortar, aggregate, interface transition zone and voids – deriving inputs principally from microscale modelling.

·          Macroscopic modelling based on the continuum approach – deriving inputs principally from mesoscale modelling. The general method that will be adopted for upscaling is the homogenization method. A detailed treatment of upscaling at the micro-meso and meso-macro levels will therefore form the central theme of this PhD research. The determination of representative elementary volumes will form a key step for all the scales considered above.

A microstructural model, HYMOSTRUC, is proposed to be used at the microscopic scale. Careful consideration will be given to alternative approaches for mesoscopic scale modelling. For instance,  SEM-images may be used as input for the mesoscopic scale.  As far as the macroscopic modelling of concrete is concerned, a flow and reactive transport code, HP1 (Jacques et al. 2008) is proposed.  This model treats the coupled physical and chemical processes at the macroscopic (continuum) scale.