Janet Lees

Dr Janet Lees - Advanced analyses

Selected Projects

Non-linear Simulations

Reinforced concrete structures are unusual in that they are often designed to be cracked under service loading. However, the presence of cracks results in non-linear structural behaviour which introduces numerous complexities in the modelling of advanced structures. Current work seeks to investigate the potential of conventional and novel simulation methods to accurately model the behaviour of advanced structures. A particular focus has been the performance of existing structures strengthened with carbon fibre reinforced polymer (CFRP) materials. More details of some of the systems under consideration can be found elsewhere. To date, the following work has been undertaken:

A review and comparison of various numerical paradigms for nonlinear simulations (e.g. smeared crack models, discrete crack models, the element-free Galerkin /Meshless method).
The development of new methods such as the Multiscale Element-Free Method (CamEFG). This developmental work will enable more in-depth studies of the crack behaviour and evaluation of the CFRP strengthened structures.
A number of case studies have been used to benchmark various simulation methods. These have included an investigation of the ability of a smeared crack approach to provide a three-dimensional representation of the effectiveness of different strengthening systems and the use of the Multi-scale element free galerkin method to identify localised cracking in half joint (also known as dapped end) beams.

3D modelling of fabric strengthening system

Research Team: Dr Janet Lees, Dr Chris Morley, Dr X-S Yang and Samir Hassan Dirar

Funding/Project Partners: EPSRC, Highways Agency, Tony Gee & Ptnrs, EMPA, Sika Ltd, the Concrete Society

Computational optimisation

Computational optimisation can be a powerful tool for designers to quickly converge on 'optimum' solutions which represent an efficient use of material. Recent work on computational optimisation has focused on the development of new optimisation algorithms and the optimisation of strengthening parameters for precracked concrete structures.

For a given strengthening method, the optimisation of the strengthening parameters is important when designing a system for an optimal or nearly optimal performance. Typical strengthening parameters include the location, spacing, orientation, amount of strengthening material, and, in some systems, the prestress level. The main aims of the optimization can be to increase the load capacity and to reduce the cost. This outcome can be achieved by encoding the parameters into a model representation, which can then be fed into an optimiser (such as Genetic Algorithms, Virtual Ant algorithms, Tabu search). The selected parameter set and the performance of the corresponding strengthened structures are then evaluated using a simulator (e.g. CamEFG).

An optimization procedure which couples virtual ant algorithms/genetic algorithms with the nonlinear element-free method has been developed to simulate the behaviour of CFRP strengthening systems. Such a method can be used to evaluate the possible performance trade offs (if any) of different ways of shear strengthening with the same amount of reinforcing material. An example how virtual ant algorithms can be applied, to find a shorter route to the maxima when considering the optimization of multi-peak function is shown below. Other analytical studies include the comparison of various FE methods, topology optimization, and CFRP strap modelling. This work is being carried out in collaboration with Dr Geoff Parks in the Engineering Design Centre.

Virtual ant optimisation - convergence on maxima

Research Team: Dr Janet Lees, Dr Chris Morley, Dr Geoff Parks and Dr X-S Yang

Funding/Project Partners: EPSRC, Highways Agency, Tony Gee & Ptnrs, EMPA, Sika Ltd, the Concrete Society

	Department of Engineering
	Structures Research Group