摘要
基于MonteCarlo方法,对短裂纹的群体萌生、扩展、干涉的物理过程进行可视化再现。细观短裂纹阶段的生长随机性主要表现在裂纹产生的随机性、生长速率的随机性、干涉的随机性和生长方向的随机性,将这些随机信息并行考虑并且有机组合各个阶段的数学或物理模型,通过计算机再现细观短裂纹的群体演化过程。同时很好地解释裂纹数密度分布现象和有主导有效短裂纹的不定性问题。
In order to reappear visually the process of short cracks growth in computer, a new method based on the development of computer technique and the Monte Carlo method was proposed. This method divided the whole process of short cracks evolution into three parts. First, the metal resistance failure model was used to simulate the stochastic initiation of cracks. Secondly, a two-stage grow model was used to calculate the growth rate of single short crack. Finally, the link model and the relax zone model were adopted to reflect the interactive effects of short cracks which were much more important in the short cracks period than in the long cracks period. The new method takes three random factors into account with the same importance: initiation, propagation and multi-cracks interactivity. All the models can be combined to simulate the growth of short cracks affected by the closure effect and the microstrueture of the material. The method can be used to simulate the behaviors of short cracks in all kinds of material. The method proposed in this paper considered much more factors that affect the growth of the interactive short cracks than previous methods. It avoided the unilateral view in the research of the behaviors of short cracks. This method can be used to reappear visually and realistically the process of the multi-micro cracks' evolution. Furthermore, using this method can successfully reappear the uncertainty of the main crack, the distribution of the density of the micro cracks. According to the comparison with the results from experiments and calculations, in the number of cycles VS length of micro-cracks, the new method is feasible in analyzing the short cracks' behavior of the fatigue of materials.
出处
《机械强度》
EI
CAS
CSCD
北大核心
2005年第5期703-707,共5页
Journal of Mechanical Strength
关键词
细观多裂纹
群体干涉
数值仿真
随机性
蒙特卡洛法
Multi-micro cracks
Collective interaction
Simulation
Random
Monte Carlo method