Pragmatic optimisation methods for determining material constants of viscoplasticity model from isothermal experimental data

Abstract

A procedure to estimate material constants for the unified Chaboche viscoplasticity model from experimental data has been published elsewhere; however several critical assumptions are made to enable this and potential numerical problems can limit the effectiveness of the optimisation. Pragmatic optimisation procedures are therefore required to determine the material properties accurately and efficiently. This is made more complex by the presence of several deformation mechanisms and their interactions. Automation is critical due to the large amounts of data generated in testing. Complications that inhibit this process can arise due to factors such as experimental scatter. In this paper, a general optimisation framework is discussed and investigated using data from isothermal tests on a P91 steel at 600°C. Potential obstacles in the procedure are addressed and solutions (such as pre-optimisation experimental data ‘cleaning’) are suggested. Methods to maximise the amount of confidence a user has in a particular optimised constant set are also discussed.

Keywords:

• Chaboche
• Visco-plasticity model
• Low cycle fatigue
• Optimisation
• P91

The authors greatly appreciate the support of both E.ON and the EPSRC (case study award number EP/J50211X/1) through funding this work. Thanks are also extended to Mr Tom Bus and Mr Brian Webster for their assistance in the experimental work.