Adaptive Multiple-Scale Meshfree Method for Geo-Mechanics and Earth-Moving Simulation

传统的有限元方法在分析大变形、高梯度、材料分离和多尺度现象时存在许多不足。这些困难部分是由于有限元网格的规则性要求造成的。在农业和建筑业，高生产率的推土设备的设计是非常可取的。了解土壤在开挖、运输和倾倒过程中的运动对设计高生产率的运土设备至关重要。土体在搬运过程中的运动表现出过度的塑性变形和破坏机制。有限元方法尚未成功地应用于土方过程的分析，主要是由于不能有效地模拟大的材料变形和分离，以及在分析土方过程中经常存在的与应变局部化的数值解有关的基本困难。基于网格的数值方法，如有限元方法，在分支问题中引入了一个长度尺度，即网格大小。因此，数值解可以敏感地依赖于网格大小。岩土材料中剪切带形成的多尺度特性也增加了传统有限元方法的复杂性。本研究的目的是开发一种实用的模拟方法，能够预测岩土材料中的大变形、剪切带形成、损伤演化和材料分离，并应用于土方过程。将特别强调开发一种自适应多尺度无网格方法，该方法允许在模拟土壤运动时进行交互和连续的h-和p-模型精化。在土体移动过程中，局部剪切带和损伤机制对土体的整体运动是至关重要的。最终目标是捕捉嵌入在整个土体响应中的细尺度局部剪切带和剪切/拉伸破坏机制。与卡特彼勒的合作将使所采用的本构模型和建议的无网格方法的实验验证方面的研究工作成为可能。本研究的主要目标是：1.在PI以往岩土材料无网格法研究的基础上，发展了适用于土体运动模拟的无网格多尺度格式以及h-和p-自适应方法。建立了岩土材料加速无网格大变形分析的稳定协调节点积分方法。提出了一种分析应变局部化过程中材料失稳的增强正则化方法。与卡特彼勒合作，通过与卡特彼勒提供的实验数据进行比较，验证所采用的材料模型和建议的无网格方法的性能。该建议的多尺度自适应无网格方法是在与卡特彼勒公司正在进行的合作的基础上发展起来的。卡特彼勒承诺帮助验证土壤本构模型和损伤模型以及无网格方法，这将确保建议的无网格方法在实际分析工业土方移动问题中的适用性。

Conventional finite element methods exhibit a number of shortcomings in analyzing problems that involve large deformation, high gradient, material separation, and multiple-scale phenomena. These difficulties are partially due to the regularity requirement of the finite element mesh. In agricultural and construction industries, designs for high productivity earth-moving equipment are highly desirable. The understanding of soil motion during excavation, hauling, and dumping is critical to the design of high productivity earth-moving equipment. Soil motion during the earth-moving process exhibits excessive plastic deformation in conjunction with failure mechanisms. Finite element methods have not been successfully applied to the analysis of earth-movingprimarily due to an inability to effectively model large material distortion and separation.There is also a fundamental difficulty associated with the numerical solution of strainlocalization that often exists in analyzing earth-moving processes. Grid-based numerical methods such as finite element methods introduce a length scale, i.e., the mesh size, in a bifurcation problem. As a result, the numerical solution can be sensitively dependent on the mesh size. The multiple-scale nature of shear-band formation in geotechnical materials also adds considerable complication to conventional finite element approaches.The objective of this research is to develop a practical simulation method capable of predicting large deformation, shear-band formation, damage evolution, and material separation in geotechnical materials with applications to earth-moving processes. Special emphasis will be given to the development of an adaptive multiple-scale meshfree method that allows an interactive and continuous h- and p- model refinement in the simulation of soil motion. The local shear-band and damage mechanisms are critical to overall soil motion in earth moving processes. The ultimate goal is to capture the fine-scale local shear-band and shear/tensile failure mechanisms embedded in the overall soil response. A collaboration with Caterpillar will enable research effort on experimental validation of the employed constitutive model and the proposed meshfree methods. The major objectives of this research are: 1. Based on the PI's previous research progress on meshfree methods for geotechnicalmaterials, develop meshfree multiple-scale formulation and h- and p- adaptivity methods for application to earth-moving simulation.2. Develop a stabilized conforming nodal integration for accelerated meshfree largedeformation analysis of geotechnical materials.3. Develop an enhanced regularization method for analysis of material instability in strainlocalization.4. Collaborate with Caterpillar to verify the performance of the employed material models and the proposed meshfree methods by comparison with experimental data provided by Caterpillar.This proposed multiple-scale adaptive meshfree method extends the development from anongoing collaboration with Caterpillar Inc. The commitment from Caterpillar to aid in validation of the soil constitutive and damage models and meshfree methods will assure the applicability of the proposed meshfree methods for practical use in analyzing industrial earth-moving problems.