A new approach was put forward to identify the damage parameters of the shear modified GTN damage model proposed by Nahshon and Hutchinson (Eur J Mech Solid 27:10–17, 2008) by combining the artificial neural networks algorithm and small punch test. The factorial design method was used to analyze the influence of the parameters on the shape of load-displacement curve of small punch test. The less important parameters were set as empirical value and the significant factors were determined by an artificial neural networks model which was build up based on large amount of simulations of small punch tests with different levels of damage parameters values. The identified parameters were validated by small punch test simulations with different specimen thickness. The results show that the identified parameters of the shear modified GTN damage model are effective to characterize the mechanical behavior as well as the damage evolution and ductile failure of material during the process of small punch test. In addition, the applicability of the identified parameters in the tests with different stress condition were verified.

中文翻译：

---

剪切修正GTN损伤模型材料参数的小冲头试验识别

Nahshon 和 Hutchinson (Eur J Mech Solid 27:10–17, 2008) 提出了一种新的方法来识别剪切修正 GTN 损伤模型的损伤参数，该方法结合人工神经网络算法和小冲试验。采用析因设计方法分析了参数对小冲头试验载荷-位移曲线形状的影响。次要参数设置为经验值，重要因素由人工神经网络模型确定，该模型基于大量不同损伤参数值的小冲压试验模拟而建立。确定的参数通过具有不同试样厚度的小冲头测试模拟进行验证。结果表明，剪切修正GTN损伤模型的识别参数可有效表征材料在小冲头试验过程中的力学行为以及损伤演化和延性破坏。此外，验证了识别参数在不同应力条件下的测试的适用性。