Finite Element Analysis and Related Numerical Schemes for Ratcheting Simulation

Numerical Schemes;Genetic Algorithm;Ratcheting;ANSYS;Plasticity;Finite element simulation;Kinematic hardening;Radial Return

Rahman, Syed Mizanur ; DR. MANSOOR HAIDER, Committee Member,DR. MURTHY N. GUDDATI, Committee Member,DR. S. RANJI RANJITHAN, Committee Member,DR. TASNIM HASSAN, Committee Chair,Rahman, Syed Mizanur ; DR. MANSOOR HAIDER ; Committee Member ; DR. MURTHY N. GUDDATI ; Committee Member ; DR. S. RANJI RANJITHAN ; Committee Member ; DR. TASNIM HASSAN ; Committee Chair

University:North Carolina State University

关键词: Numerical Schemes;    Genetic Algorithm;    Ratcheting;    ANSYS;    Plasticity;    Finite element simulation;    Kinematic hardening;    Radial Return;   

Others  :  https://repository.lib.ncsu.edu/bitstream/handle/1840.16/5967/etd.pdf?sequence=1&isAllowed=y

美国|英语

来源: null

【 摘 要 】

Towards developing a robust finite element simulation model, this dissertation determines the state-of-the-art of simulation of ratcheting responses of structures. With this objective, the study evaluated seven constitutive models for their ratcheting simulation capabilities for straight pipe and elbow pipe components. Both local (strain ratcheting) and global (e.g., load-deflection, ovalization) responses of these two piping components were considered in the evaluation. The models evaluated are Bilinear, Multilinear and Chaboche, modified Chaboche, Ohno-Wang, modified Ohno-Wang, and Abdel Karim-Ohno. In search of the best numerical scheme for implementation of cyclic plasticity models into finite element programs, Euler and Runge-Kutta (both explicit and implicit) type numerical schemes for solving the nonlinear incremental plasticity equations and approximating consistency condition, and return type algorithms for updating back stresses, were evaluated. This numerical scheme evaluation was conducted at the material level with respect to the simulations of stable hysteresis loop, uniaxial ratcheting and multiaxial ratcheting responses. Implicit radial return scheme was demonstrated to be the best numerical scheme for implementing cyclic plasticity models. Automated parameter determination tools based on genetic algorithm search technique were developed for determining the model parameters of the four advanced constitutive models.In addition, for evaluating constitutive models at the materials level, the strain-driven radial return algorithm was extended, such that both stress and strain increments can be prescribed simultaneously. The study demonstrated that the existing cyclic plasticity models are not capable of simulating straight and elbow pipe ratcheting responses satisfactorily, when the model parameters are determined from material level responses (stable hysteresis loop, and uniaxial and/or biaxial ratcheting responses). For improving the simulation capability of the existing models, this study proposed a semi-inverse approach for refinement of model parameters using both the material level and structural local responses simultaneously. This proposed approach is validated for modified Chaboche model with respect to the straight pipe responses. Finally, limitations of the existing models are identified and recommendations are made for developing a robust model for structural ratcheting simulations.