不确定条件下脆性材料离散元模型参量的计算反求方法研究

脆性材料在民用和军用工程中的广泛应用，脆性材料破坏问题的研究显得尤其重要。离散元法作为脆性材料破坏问题研究的有效工具，对破坏机理研究具有重要意义。离散元模型参量的获取是离散元法成功应用的重要前提。计算反求方法为难以直接获取的模型参量提供了一种有效的获取途径。考虑工程实际中不确定性广泛存在，合理评估不确定性对模型参量识别的影响具有重要的科学意义和应用价值。本项目拟综合考虑离散元建模，不确定性、反问题求解等难点，研究不确定性脆性材料离散元模型参量识别的计算反求技术。首先，基于正交试验设计法建立合理的脆性材料离散元模型；其次，基于区间度量不确定性，开发全局敏感性分析技术，实现参量识别反问题模型的建立；最后，考虑不确定性水平，开发基于区间的不确定性参量高效识别算法，实现参量识别的快速求解。本项目的完成将为脆性材料破坏问题的研究提供更为合理的模型参量区间估计，为离散元方法的工程应用提供技术支撑。

With the wider use of brittle materials in civil and military engineering applications, it is more and more important to study the failure of brittle materials. Discrete element method (DEM), one of the valuable tools for the research on the failure of brittle materials, is useful for the study of failure mechanism of brittle materials. The available parameters of DEM model are very important and necessary for the successful use of DEM. The computational inverse method can provide an effective means for the determination of model parameters which can not be identified directly. In the engineering practice, the uncertainty is widely existed. It is of great scientific and practical important to reasonably estimate the influence of the uncertainty for model parameters identification. In this project, considering comprehensively the difficulties about modeling with DEM, uncertainty and the solution of inverse problem, the computational inverse techniques for the determination of DEM model parameters of brittle materials with uncertainty will be investigated and developed. Firstly, the effective DEM model for brittle material will be constructed though orthogonal experimental design method. Secondly, to construct the inverse analysis model of parameters identification, the uncertainty will be described based on interval and global sensitivity analysis method will be developed. Finally, considering the level of uncertainty, the high efficient algorithm for uncertain parameters based on interval will be developed, which can realize the speedily solving for parameters identification. The completion of this project can provide more compelling interval estimate for model parameters and provide the technological support for the engineering application of the discrete element method.