登录 EN

添加临时用户

基于粒子的流体力学与离散元耦合算法开发与应用研究

Development and application of the coupling algorithm between the particle-based fluid dynamics and DEM

作者:丁文涛
  • 学号
    2016******
  • 学位
    硕士
  • 电子邮箱
    180******com
  • 答辩日期
    2018.06.04
  • 导师
    徐文杰
  • 学科名
    土木工程
  • 页码
    104
  • 保密级别
    公开
  • 培养单位
    004 水利系
  • 中文关键词
    光滑粒子流体动力学,格子玻尔兹曼法,离散元,流-固耦合,岩土灾害
  • 英文关键词
    Smooth Particle Fluid Dynamics, Lattice Boltzmann, Discrete Element, Fluid-Solid Coupling, Geotechnical Disasters

摘要

流-固耦合问题是自然界和岩土工程领域常见的现象,比如渗流、滑坡涌浪和泥石流等过程中,均伴随着复杂的流-固耦合作用。由于流-固耦合的机理复杂并且难以通过现场或试验观察和测量,各领域从不同方面开展了很多研究工作,但是仍然存在许多局限。近年来,数值计算技术不断发展,为深入研究流-固耦合问题提供了技术支持。目前最常用的是基于有限体积法(FVM)的计算流体动力学(CFD)和FVM-DEM耦合方法等,但是这类方法难以求解流体与固体之间的复杂接触问题。为了更加精确的研究流-固耦合过程中流体与固体间的相互作用,本文中采用基于粒子的流体动力学和离散元(DEM)方法进行耦合,其中基于粒子的流体动力学方法采用格子玻尔兹曼方法(LBM)和光滑粒子流体动力学方法(SPH)。LBM方法在模拟细观流动以及耦合计算方面具有突出优势,而SPH方法在模拟大规模问题和复杂流体边界问题方面功能强大,同时由于LBM和SPH的离散特性,与DEM方法进行耦合十分高效,因此可以很好实现颗粒材料中复杂的流-固耦合问题分析。为了充分发挥当前开源资源及不同学科领域的优势,本文在研究过程分别采用开源软件Palabos(LBM)和DualSPHyscis(SPH)实现流体动力学的计算分析,而采用Blaze-DEM和Yade实现基于DEM的固体力学计算分析。在此基础上,论文发展了LBM-DEM和SPH-DEM流-固耦合算法,实现了不同软件间的耦合计算分析。同时,基于CUDA开发了高性能计算方法,实现了高效、精确的流-固耦合问题计算,以用于解决颗粒材料的各种复杂流-固耦合过程分析。为了验证本文开发的流-固耦合计算分析模块可行性和可靠性,首先了选取多个经典问题进行计算分析,包括小球落水、子弹入水和楔形体滑坡涌浪问题等,并将数值计算结果与实验结果进行了系统的对比与分析。从分析结果上来看,数值计算得到的包括固体的运动状态、流场形态和波动状态等均与实验结果相吻合,充分验证了所开发的流-固耦合算法及计算模块的可靠性。为了进一步论证,本文开发算法模块在岩体灾害研究中的可行性,针对典型的几类流-固耦合问题开展了研究:1)基于LBM-DEM耦合方法对泥石流问题进行模拟;2)基于SPH-DEM耦合方法对海底滑坡问题和滑坡涌浪问题进行模拟。通过模拟大规模和复杂边界的流-固耦合问题,并对复杂地质灾害过程进行了系统的机理分析和验证,为相关类似灾害研究提供了有力的理论和技术支持。

Fluid-solid interaction is a common phenomenon in nature and geotechnical engineering such as seepage, landslide and mudslides. Because the mechanism of fluid-solid coupling is very complex, and it is difficult to observe and measure through experiment, many researches have been carried out in various fields from various aspects, but there are still many problems and limitations. In recent years, numerical calculation technology has been continuously developed to provide technical support for further research on the flow-solid coupling problem. At present, the most commonly used methods are computational fluid dynamics (CFD) and FVM-DEM coupling method based on the finite volume method (FVM), but such methods are difficult to solve the complex contact problems between the fluid and the solid.In order to solve fluid-solid interaction problem more precisely, particle-based fluid dynamics and discrete element (DEM) methods are used for coupling in this paper. Lattice-based fluid dynamics methods include the lattice Boltzmann method (LBM) and smoothed particle hydrodynamics (SPH). The LBM method has outstanding advantages in simulating meso-flow and coupling problems, while the SPH method is powerful in simulating large-scale problems and complex fluid boundary problems. Because of the discrete nature of LBM and SPH, coupling with the DEM method is very efficient. Therefore, the complex flow-solid coupling problem in granular materials can be well analyzed.In order to take the advantages of current open source resources and different disciplines, this paper uses the open source software Pablabs (LBM) and DualSPHyscis (SPH) to achieve fluid dynamics computational analysis, while using Blaze-DEM and Yade to realize DEM-based solid mechanics calculation and analysis. On this basis, this paper develops the LBM-DEM and SPH-DEM flow-solid coupling algorithm. At the same time, a high-performance calculation method based on CUDA has been developed to achieve highly efficient and accurate computational analysis of the fluid-solid coupling problem, which can be used to solve various complex flow-solid coupling calculations for particulate materials.In order to verify the feasibility and reliability of the fluid-solid coupling module developed in this paper, several classical problems are selected for calculation and analysis, including the ball falling into the water, bullet into the water and the wedge landslide problem, where the numerical calculation results are compared with the experimental results. From the analysis results, the numerical conditions including the solid state of motion, the flow field morphology and the fluctuation state are all in agreement with the experimental results, which fully validates the reliability of the developed fluid-solid coupling algorithm and module.In order to further demonstrate the feasibility of the developed algorithm in the study of geological disasters, a study is conducted on several typical flow-solid coupling problems. First, the debris flow problem is simulated based on the LBM-DEM coupling method. Then, the SPH-DEM coupling method is used to simulate the problem of submarine landslide and landslide surge. By simulating the flow-solid coupling problems of large-scale and complex boundaries, the systematic mechanism analysis and verification of the complex geological disasters process has provided strong theoretical and technical support for similar disaster research.