On the thermodynamically consistent modeling of distortional hardening: A novel generalized framework
On the thermodynamically consistent modeling of distortional hardening: A novel generalized framework
复制标题
DOI：
10.1016/j.ijplas.2014.05.008
复制
发表时间：
2014-12
影响因子：
9.8
通讯作者：
Baodong Shi;A. Bartels;J. Mosler
中科院分区：
文献类型：
--
作者：
Baodong Shi;A. Bartels;J. Mosler
材料在宏观尺度上的塑性变形的许多重要物理效应不能仅仅通过各向同性和运动硬化来真实地描述。例如，多晶体织构的演变在宏观上导致变形的屈服面。本文讨论了适当的硬化模型，这样的变形。更准确地说，阐述了一种新的通用框架的有限应变塑性模型。据作者所知，这是第一个结合了以下特点的：（1）材料相容性的证明;（2）畸变硬化分解为动态硬化（由于目前活跃的位错）和潜在硬化（由于目前不活跃的位错）;（3）在加载方向和相反方向的屈服面曲率的差异。这个模型的基石是一个新的塑性潜力的发展方程畸变硬化。虽然这种类型的硬化是通过四阶张量作为内部变量来表征的，但上述势的结构非常简单。尽管最终模型相当复杂，但它只需要很少的模型参数。对于这些参数，反过来，物理上健全的边界的基础上的凸性条件的屈服面可以推导出来。三个不同的例子证明了新框架的预测能力。
Many important physical effects of materials undergoing plasticity at the macroscale cannot be captured realistically by isotropic and kinematic hardening only. For instance, the evolution of the texture in polycrystals results macroscopically in a distorted yield surface. This paper deals with adequate hardening models for such a distortion. To be more precise, a novel general frame for finite strain plasticity models is elaborated. To the best knowledge of the authors, it is the first one combining the following features: (1) proof of thermodynamical consistency; (2) decomposition of distortional hardening into dynamic hardening (due to currently active dislocations) and latent hardening (due to currently inactive dislocations); (3) difference of the yield surface’s curvature in loading direction and in the opposite direction. The cornerstone of this model is a new plastic potential for the evolution equations governing distortional hardening. Although this type of hardening is characterized through a fourth-order tensor as internal variable, the structure of the aforementioned potential is surprisingly simple. Even though the final model is rather complex, it requires only few model parameters. For these parameters, in turn, physically sound bounds based on the convexity condition of the yield surface can be derived. Three different examples demonstrate the predictive capabilities of the novel framework.