Institute of Mechanics Ruhr University Bochum Bochum Germany.
Minim Invasive Ther Allied Technol. 2004 Aug;13(4):228-39. doi: 10.1080/13645700410017362.
The present contribution focuses on the structural modelling of Nitinol devices. Apart from a powerful material modelling which should consider the well-known phenomena pseudo-elasticity, pseudo-plasticity and the shape memory effect a computationally efficient finite element implementation is needed. The cost of the numerical simulation depends crucially on the implementation of the material model at the Gauss point level (local level) and on the finite element technology used. Concerning the first point it is meanwhile state of the art to work with a so-called consistent linearization. This is also pursued in the present paper. On the other hand, the question of finite element technology is hardly discussed in the literature on shape memory alloy (SMA) modelling, although it is one of the crucial points. We discuss here a new one Gauss point concept which avoids the well-known "locking" (artificial stiffening) of standard displacement-based formulations and which is even more efficient than many formulations available in commercial codes.
本文着重于镍钛诺器件的结构建模。除了一个强大的材料建模(应考虑著名的伪弹性、伪塑性和形状记忆效应等现象)外,还需要一个计算效率高的有限元实现。数值模拟的成本关键取决于高斯点级别(局部级别)的材料模型实现以及所使用的有限元技术。关于第一点,目前的技术水平是采用所谓的一致线性化。本文也采用了这种方法。另一方面,有限元技术问题在形状记忆合金(SMA)建模的文献中几乎未被讨论,尽管它是关键要点之一。我们在此讨论一种新的单高斯点概念,它避免了标准基于位移公式中众所周知的“锁定”(人为硬化),并且比商业代码中许多可用的公式更高效。