As stated in the sister article that the objective of this study was to explore the low-cycle fatigue and ratcheting failure responses of elbow components through experimental and analytical studies. Low-cycle fatigue and ratcheting damage accumulation in piping components may occur under load reversals induced by earthquakes or thermomechanical operations. Ratcheting damage accumulation can cause failure of components through cracking or plastic buckling. Hence, design by analysis of piping components against ratcheting failure will require simulation of this response with reasonable accuracy. In developing a constitutive model that can simulate ratcheting responses of piping components, a systematic set of elbow experiments involving deformation and strain ratcheting were conducted and reported in the sister article. This article will critically evaluate seven different constitutive models against their elbow response simulation capabilities. The widely used bilinear, multilinear, and Chaboche models in ansys are first evaluated. This is followed by evaluation of the modified Chaboche, Ohno–Wang, modified Ohno–Wang, and Abdel Karim–Ohno models. Results from this simulation study are presented to demonstrate that all the seven models can simulate the elbow force response reasonably. The bilinear and multilinear models can simulate the initial elbow diameter change or strain accumulation, but always simulate shakedown during the subsequent cycles when for some of the cases the experimental trends are ratcheting. Advanced constitutive models like Chaboche, modified Chaboche, Ohno–Wang, modified Ohno–Wang, and Abdel Karim–Ohno can simulate many of the elbow ratcheting responses well, but for some of the strain responses, these models simulate negative ratcheting, which is opposite to the experimental trend. Finally, implications of negative ratcheting simulation are discussed and suggestions are made for improving constitutive models ratcheting response simulation.
Read full abstract