Material parameter optimisation of Ohno-Wang kinematic hardening model using multi objective genetic algorithm

Abstract

Abstract:Ohno-Wang hardening model is an advanced constitutive model to evaluate the cyclic plasticity behaviour of material. This model has capability to simulate uniaxial and biaxial ratcheting response of the material. But, it is required to determine large number of material parameters from several experimental responses in order to simulate this phenomenon. Material parameters for constitutive models are generally determined manually through trial and error approach which is tedious and less accurate. Due to arbitrariness and complexity of cyclic loading, advanced constitutive material models become non-linear and multimodal in functional and parameter space. To overcome this problem, an automated parameter optimisation approach using genetic algorithm has been proposed in the present work to identify Ohno-Wang material parameters of 304LN, stainless steel for uniaxial simulation. Optimisation by this approach has improved the model prediction in uniaxial low cycle and ratcheting fatigue simulations after comparison with the experimental response.

Optimising the multiplicative AF model parameters for AA7075 cyclic plasticity and fatigue simulation

“…As the experimental understanding of materials increases, so has the level of sophistication and complexity of elastoplastic models, leading to increased parameter calculation requirements, as recognised by Grama et al (2015). In order to improve the calculation process of elastoplastic constitutive model parameters, various optimisation techniques have been investigated, with two main optimisation strategies identified, the gradient-based (Mahnken and Stein, 1996, Saleeb et al, 2002, Desai and Chen, 2006 and the genetic algorithm (GA) methodologies (Rahman et al, 2005, Krishna et al, 2009, Badnava et al, 2012, Agius et al, 2017a, Mahmoudi et al, 2011, Farrahi et al, 2014, Rokonuzzaman and Sakai, 2010, Khademi et al, 2015, Cermak et al, 2015, Zhao and Lee, 2002, Khutia and Dey, 2014, Franulović et al, 2009. As highlighted by Furukawa et al (2002), the disadvantage of the gradient-based approach in constitutive model parameter determination lies in the solution divergence, an issue not typically associated with GA methodologies, which are instead associated with poor solution efficiency.…”

Sensitivity and optimisation of the Chaboche plasticity model parameters in strain-life fatigue predictions

“…As the models increase their level of sophistication and complexity, the number of material parameters has increased as well. This has led to the introduction of various material parameter optimisation techniques, in order to ease/streamline the identification process and improve the accuracy of the model predictions [16][17][18][19][20][21][22][23][24][25][26][27]. However, the sensitivity of the strain life fatigue predictions to the model parameters requires a thorough assessment, owing to the complexity of many real-life load spectra.…”