MATERIAL CHARACTERIZATION BASED ON INSTRUMENTED AND SIMULATED INDENTATION TESTS
MATERIAL CHARACTERIZATION BASED ON INSTRUMENTED AND SIMULATED INDENTATION TESTS
复制标题
DOI：
10.1142/s175882510900006x
复制
发表时间：
2009-03
影响因子：
3.5
通讯作者：
Zishun Liu;Edy Harsono;S. Swaddiwudhipong
中科院分区：
文献类型：
--
作者：
Zishun Liu;Edy Harsono;S. Swaddiwudhipong
本文综述了各种技术来表征材料的载荷-位移数据解释仪表压痕试验。采用尺度分析和无量纲分析方法，总结了压痕曲线特征与材料性能之间的普遍关系。通过广泛的有限元分析的无量纲功能进行了数值校准。从建立的关系的载荷-位移曲线的解释，从而进行迭代求解高阶函数或采用神经网络。在这项研究中，这些技术的优点和缺点突出。讨论了仪器化压痕试验中的摩擦、尺寸效应和反分析算法的唯一性等问题。在这项研究中，一个新的反演算法，通过神经网络模型，以提取力学性能的双伯科维奇和球形压痕试验。所提出的神经网络模型的基础上预测的材料性能与数值输入数据吻合良好。
This paper reviews various techniques to characterize material by interpreting load-displacement data from instrumented indentation tests. Scaling and dimensionless analysis was used to generalize the universal relationships between the characteristics of indentation curves and their material properties. The dimensionless functions were numerically calibrated via extensive finite element analysis. The interpretation of load-displacement curves from the established relationships was thus carried out by either solving higher order functions iteratively or employing neural networks. In this study, the advantages and disadvantages of these techniques are highlighted. Several issues in an instrumented indentation test such as friction, size effect and uniqueness of reverse analysis algorithms are discussed. In this study, a new reverse algorithm via neural network models to extract the mechanical properties by dual Berkovich and spherical indentation tests is introduced. The predicted material properties based on the proposed neural network models agree well with the numerical input data.