A novel architectural Ti composite composed of network-woven structured TiB nanowires in a core-shell structured Ti matrix was fabricated to improve the strength of Ti matrix composites(TMCs),where the shell consists ...A novel architectural Ti composite composed of network-woven structured TiB nanowires in a core-shell structured Ti matrix was fabricated to improve the strength of Ti matrix composites(TMCs),where the shell consists of rich N solute atoms while the core is deficient of N solute atoms through spark plasma sintering of powder mixtures of Ti powder and BN nano-powder.The phase composition,morphology,element distribution,and mechanical properties of prepared samples were analyzed by X-ray diffraction(XRD),scanning electron microscope(SEM),electron probe microanalyzer(EPMA),and electronic universal material testing machine.The results indicate that the TMCs with designed architectures have been successfully achieved,and the as-prepared Ti-2BN(wt.%)composite exhibits an ultimate compressive strength of~1.8 GPa with a strain-to-fracture of~9%,while the Ti-1BN(wt.%)attains an ultimate compressive strength of~1.6 GPa and a strain-to-fracture of~20%.Moreover,the roles of the hybrid reinforcement structures in strengthening the Ti composites were discussed.展开更多
鉴于传统的BP网络的速度慢和局部极小值问题,以及针对基于实验数据训练神经网络存在样本不足的缺陷,文中提出了利用径向基函数(Rad ial Base Function,简记为RBF)神经网络通过有限元方法对含有脱层损伤的复合材料试件进行数值模拟,把前...鉴于传统的BP网络的速度慢和局部极小值问题,以及针对基于实验数据训练神经网络存在样本不足的缺陷,文中提出了利用径向基函数(Rad ial Base Function,简记为RBF)神经网络通过有限元方法对含有脱层损伤的复合材料试件进行数值模拟,把前五阶弯曲模态频率进行修正,以修正后的前五阶弯曲模态频率再经过归一化处理构建训练样本的新思路,将实验模态分析结果经归一化处理后送入训练好的RBF神经网络进行预测,从而实现对编制复合材料梁的脱层损伤定位和损伤程度评估。最后给出了编织复合材料结构损伤大小伤识别及定位的算例,仿真结果表明RBF神经网络速度快,稳定性好,精度高,在复合材料结构损伤监测中具有光明的应用前景和重要的工程应用价值。展开更多
基金supported by the Australian Research Council(No.LP130100913)the Baosteel-Australia Joint Research and Development Centre on the Project(No.BA110014LP)。
文摘A novel architectural Ti composite composed of network-woven structured TiB nanowires in a core-shell structured Ti matrix was fabricated to improve the strength of Ti matrix composites(TMCs),where the shell consists of rich N solute atoms while the core is deficient of N solute atoms through spark plasma sintering of powder mixtures of Ti powder and BN nano-powder.The phase composition,morphology,element distribution,and mechanical properties of prepared samples were analyzed by X-ray diffraction(XRD),scanning electron microscope(SEM),electron probe microanalyzer(EPMA),and electronic universal material testing machine.The results indicate that the TMCs with designed architectures have been successfully achieved,and the as-prepared Ti-2BN(wt.%)composite exhibits an ultimate compressive strength of~1.8 GPa with a strain-to-fracture of~9%,while the Ti-1BN(wt.%)attains an ultimate compressive strength of~1.6 GPa and a strain-to-fracture of~20%.Moreover,the roles of the hybrid reinforcement structures in strengthening the Ti composites were discussed.
文摘鉴于传统的BP网络的速度慢和局部极小值问题,以及针对基于实验数据训练神经网络存在样本不足的缺陷,文中提出了利用径向基函数(Rad ial Base Function,简记为RBF)神经网络通过有限元方法对含有脱层损伤的复合材料试件进行数值模拟,把前五阶弯曲模态频率进行修正,以修正后的前五阶弯曲模态频率再经过归一化处理构建训练样本的新思路,将实验模态分析结果经归一化处理后送入训练好的RBF神经网络进行预测,从而实现对编制复合材料梁的脱层损伤定位和损伤程度评估。最后给出了编织复合材料结构损伤大小伤识别及定位的算例,仿真结果表明RBF神经网络速度快,稳定性好,精度高,在复合材料结构损伤监测中具有光明的应用前景和重要的工程应用价值。
