Modeling and computation of shakedown problems for nonlinear hardening materials (Q1209972)

The paper proposes the method for the shakedown analysis for perfectly plastic, nonlinear kinematic hardening, linear unlimited kinematic hardening and cyclic kinematic hardening materials. Although the static shakedown theorem for linear kinematic hardening materials has been extensively elaborated in the literature, in the search for a corresponding theorem in the case of nonlinear kinematic hardening little progress has been made. In the paper first a three-dimensional overlay mode I proposed by the authors in their previous papers is discussed. The material point is composed of a spectrum of microelements. Stresses and strains are defined separately for the material point and microelements. For the latter ones it is sufficient to assume them as linearly elastic, perfectly plastic. The state of microelements defines the state of the material point. As a result, a static shakedown theorem is formulated for both perfectly plastic and linear, unlimited kinematic hardening materials. The FEM approach to shakedown problems is limited to the use of the static theorems since the discretized optimization problem is regular. The influence of kinematic hardening on shakedown limits is discussed. To improve the precision of the shakedown limits, the adaptive mesh refinement is proposed. Here the modified error estimator based on the works of \textit{O. C. Zienkiewicz} and \textit{J. Z. Zhu} [Int. J. Numer. Methods Eng. 24, 337-357 (1987; Zbl 0602.73063)] is presented. Numerical examples prove the efficiency of the proposed approach. Numerical simulation was verified by experimental tests. Results obtained experimentally for cyclic load case precisely coincide with the ones obtained by calculations. The paper constitutes a significant contribution to the shakedown problem. It is written in a clear form, with the details necessary to perform calculations for other problems. It touches practical aspects of the problem rather than theoretical ones. However, the paper must be considered as a certain stage of the research, which will evolve. Alternative solutions and algorithms for respective and more complex problems will certainly appear.