空间网格结构节点螺栓松动损伤定位有限元分析

以单层柱面网壳试验模型为背景进行损伤定位研究，采用ANSYS软件对其螺栓松动损伤进行了有限元分析，利用神经网络的模式分类功能进行了螺栓松动的分步损伤定位，即首先以整体结构为研究对象，采用面向子结构的损伤定位，确定结构是否存在损伤及损伤的大致位置；然后采用面向节点的损伤定位法，确定子结构中螺栓松动损伤的节点位置。研究结果表明：损伤识别的成功率与杆件灵敏度密切相关，为了保证识别结果准确性，其灵敏度达到2%以上才能正确识别，当杆件的灵敏度过低时将无法通过动力测试识别；采用分步损伤定位方法可以显著减少损伤精确定位时训练样本的数量，适合于不完备的模态数据，利用低阶模态数据即可准确识别损伤的子结构与节点位置。

Finite element analysis on damage detection of space grid structure node bolt looseness

Damage location was studied based on single-layer latticed cylindrical shell experiment model. Finite element analysis for the looseness of the bolt-ball was done by FEA software ANSYS using neural network pattern classification. The analysis focused on the whole structure. The sub-structure oriented damage location method was used in order to verify if there’s damage and the general location of the damage. Then the node-location oriented technique was used to confirm the accurate location of the loosened bolt-ball. By finite element analysis and experimental results, damage identification is related to the sensitivity of the bolt-ball. To get the right result, the sensitivity should be more than 2%.The above mentioned method, based on incomplete modal data, can be used for the location of loosened bolt-ball. The above mentioned method, based on incomplete modal data, can be used for the location of the loose of the bolt-ball.