This paper presents a remeshing-free, graph-based finite element method (FEM) for simulating ductile and brittle fracture. Since fracture produces new mesh fragments and introduces additional degrees of freedom in the system dynamics, existing FEM based methods suffer from an explosion in computational cost as the system matrix size increases. Our model develops an algorithm for modelling fracture on the graph induced in a volumetric mesh with tetrahedral elements, by its vertices and edges. In order to initialize and propagate fracture, We simply relabel the edges of the graph using a damage variable, and control the diffusion of fracture inside the object by varying the support of a damage kernel. We present the reformulated system dynamics for this relabeled graph that allows us to simulate fracture in a FEM setting, without changing the size of system dynamics matrix of the mesh. This makes our computational method remeshing-free and scalable to high-resolution meshes. The fracture surface has to be reconstructed explicitly for visualization purposes only. We evaluate our algorithm extensively on a variety of brittle and ductile materials and compare its features with other state of the art methods. We simulate standard laboratory experiments from structural mechanics and compare the results visually and quantitatively with real-world experiments performed on physical material samples. We also present results evaluating the performance of our method, and show that our techniques offer stability and speed that is unmatched in current literature.
翻译:本文展示了模拟软质和软质骨折的无透镜、基于图形的限定元素法(FEM ) 。 由于骨折产生新的网状碎片,并在系统动态中引入更多自由度, 现有的基于FEM 的方法随着系统矩阵大小的增大而出现计算成本的爆炸。 我们的模型开发了一种算法,用于模拟带有四面形元素的体积网格中引出的图形骨折。 为了初始化和传播断裂,我们简单地用损坏变量重新标出图表的边缘,并通过改变损坏内核的支撑来控制物体内部骨折的蔓延。 我们展示了这个重新标定的图的系统动态,使我们能够在系统矩阵大小上模拟断裂,而不会改变网状的系统动态矩阵大小。 这使得我们的计算方法能够通过脊椎和边缘进行无损和可缩缩缩缩缩缩缩缩。 为了初始化目的,我们只需对断裂面表面进行明确的重塑,我们用各种当前变形和感动速度的实验,我们用实际实验方法评估了我们当前和感动的物理实验结果,我们用真实的实验方法和感力分析了我们现在的模型的实验结果。