爆炸荷载作用下建筑玻璃的破碎分析
Analysis of Architecture Glass Breaking Under Explosion Load

采用传统的基于侵蚀算法的单元删除法模拟玻璃裂纹的生成时，由于单元删除造成的质量损失及能量不守 恒，体系的控制方程不再成立，导致数值模拟结果准确性不高．为此，基于有限元显式动力分析软件LS-DYNA， 提出了模拟爆炸荷载作用下玻璃裂纹产生的节点分离法，可避免单元的删除，保证体系的控制方程成立．分别采用 节点分离法与单元删除法对某现场爆炸试验中玻璃板的破碎进行数值模拟，对比模拟结果发现，节点分离法得到的 碎片最大抛射速度、最远抛射距离以及玻璃破碎形态与试验吻合得更好，从而验证了节点分离法的准确性．此外， 采用统计学方法对两种方法下数值模拟结果的玻璃碎片尺寸分布、抛射速度分布以及抛射距离分布进行了详细统 计．结果表明：两种方法下，随着爆炸比例距离的减小，生成的小碎片越多，抛射速度越大，抛射距离越远；节点 分离法在模拟玻璃小尺寸碎片方面更具优越性．
When the traditional element deleting method based on erosion algorithm is adopted to simulate the cracking of glass, the mass losses and unbalance of energy makes the governing equations invalid, leading to inaccurate modeling results. In this paper, the node separation method is proposed to simulate the cracking of glass under blast loading based on the explicit finite element dynamic analysis software LS-DYNA, which can avoid element deletion and ensure the validity of the governing equations. A glass panel taking from an explosion test is modeled by using both the node separation and element deleting methods. Comparisons of the simulation results indicate that the maximum ejection velocity, maximum splashing distance and breaking mode getting from the node separation method match better with the test results, which verifies the reliability of the method. In addition, distributions of the glass fragment size, the ejection velocity and the the splashing distance from the numerical results are analyzed by using statistical methods. The results show that more small fragments, larger ejection velocity and farther splashing distance are obtained with decreasing scaled distance, and the node separation method has the superiority to simulate small fragments of glass