基于CT扫描的花岗岩三维数值试件重构模型及应用

岩石细观结构控制宏观破裂模式,建立一种表征岩石细观结构的数值模型重构方法并能够嵌入到力学模型中计算岩石破坏过程应力分布,对于揭示岩石破坏机理具有重要的理论意义.论文以花岗岩试件为研究对象,借助于先进的CT技术获取试件内部结构切片图像,利用数字图像处理技术实现了基于CT图像的花岗岩试件结构的识别和表征,基于矢量化方法和岩石破裂过程分析系统RFPA3D建立了花岗岩三维数值试件,对花岗岩试件单轴压缩破裂过程进行了数值模拟.通过数值试验结果与物理实验结果对比发现,基于CT扫描的花岗岩数值试件真实表征了岩石细观结构,所以力学模拟得到的破裂模式与室内物理实验得到的破裂模式相似,同时能够解译岩石破裂力学机理,为深入研究岩石、复合材料等力学特征提供了一种可行方法.

混凝土随机损伤本构关系的数值模拟研究

从细观力学层次出发,把混凝土看作是由骨料、砂浆基质及两者之间的界面过渡区组成的三相复合材料,通过引入细观损伤变量及损伤变量的演化规律模拟了混凝土的破坏过程。首先生成骨料位置随机分布且颗粒级配符合Fuller曲线的数值试件,然后对每一相材料分别赋以服从Weibull分布的随机力学参数(弹模、强度和泊松比等)并规定相应的破坏准则,将通过上述途径获得的混凝土数值试件导入有限元软件Abaqus进行数值模拟。与实验结果的对比显示,本文提出的随机损伤数值模型能够较好地模拟混凝土板式试件在单拉、单压下的应力-应变关系曲线以及混凝土中裂纹的萌生、开裂过程。

二维数字试件中骨料模型分析

建立混凝土的细观数字试件,进行数值模拟试验,探讨骨料体积率、骨料级配对混凝土的力学性能的影响具有重要的意义。二维数字试件是采用广泛的一种模型,其中的骨料面积率、剖面级配与三维模型中的骨料体积率和体积级配的关系还没有专门的探讨,然而此项工作对二维数字试件的建立与应用具有很大影响。本文以计算机试验为手段,针对随机的三维骨料投放进行大量的计算机模拟试验,阐述骨料面积率、剖面级配与骨料体积率和三维级配的规律,建立了二维数字试件模型中骨料模型,为二维细观数字试件的研究奠定了基础。

Numerical simulation and experimental study for rheo-forged component using direct and indirect die system

Sand casting and die casting processes have been used widely for complex automotive products such as knuckle, arm, etc.Usually, a part fabricated by casting has limited strength due to manufacturing defects such as the dendrite structure, segregation and porosities.As an attempt to offer a solution to these problems, forging has been used as an alternative process.However, the forging process provides limited formability for complex shape products.Rheo-forging of metal offers not only superior mechanical strength but also requires significantly lower machine loads than solid forming processes.In order to produce semi-solid materials of the desired microstructure, a stirring process is applied during solidification of molten metal.The results of an A356 aluminum alloy sample, which are obtained by experiment and by simulation using DEFORM 3D, are present.

Improved parameter selection method for mesoscopic numerical simulation test of direct tensile failure of rock and concrete

In order to numerically simulate the failure process of rock and concrete under uniaxial tension,an improved method of selecting the mechanical properties of materials was presented for the random mechanic parameter model based on the mesoscopic damage mechanics.The product of strength and elastic modulus of mesoscale representative volume element was considered to be one of the mechanical property parameters of materials and assumed to conform to specified probability distributions to reflect the heterogeneity of mechanical property in materials.With the improved property parameter selection method,a numerical program was developed and the simulation of the failure process of the rock and concrete specimens under static tensile loading condition was carried out.The failure process and complete stress-strain curves of a class of rock and concrete in stable fracture propagation manner under uniaxial tension were obtained.The simulated macroscopic mechanical behavior was compared with the available laboratory experimental observation,and a reasonable agreement was obtained.Verification shows that the improved parameter selection method is suitable for mesoscopic numerical simulation in the failure process of rock and concrete.

Optimization of the MUSCL scheme by dispersion and dissipation

A second-order optimized monotonicity-preserving MUSCL scheme(OMUSCL2) is developed based on the dispersion and dissipation optimization and monotonicity-preserving technique.The new scheme(OMUSCL2) is simple in expression and is easy for use in CFD codes.Compared with the original second-order or third-order MUSCL scheme,the new scheme shows nearly the same CPU cost and higher resolution to shockwaves and small-scale waves.This new scheme has been tested through a set of one-dimensional and two-dimensional tests,including the Shu-Osher problem,the Sod problem,the Lax problem,the two-dimensional double Mach reflection and the RAE2822 transonic airfoil test.All numerical tests show that,compared with the original MUSCL schemes,the new scheme causes fewer dispersion and dissipation errors and produces higher resolution.

Experimental and Numerical Investigation of Centrifugal Pumps with Asymmetric Inflow Conditions

Most of the times pumps operate off best point states.Reasons are changes of operating conditions,modifications,pollution and wearout or erosion.As consequences non-rotational symmetric flows,transient operational conditions,increased risk of cavitation,decrease of efficiency and unpredictable wearout can appear.Especially construction components of centrifugal pumps,in particular intake elbows,contribute to this matter.Intake elbows causes additional losses and secondary flows,hence non-rotational velocity distributions as intake profile to the centrifugal pump.As a result the impeller vanes experience permanent changes of the intake flow angle and with it transient flow conditions in the blade channels.This paper presents the first results of a project,experimentally and numerically investigating the consequences of non-rotational inflow to leading edge flow conditions of a centrifugal pump.Therefore two pumpintake-elbow systems are compared,by only altering the intake elbow geometry:a common single bended 90°elbow and a numerically optimized elbow(improved regarding rotational symmetric inflow conditions and friction coefficient).The experiments are carried out,using time resolved stereoscopic PIV on a full acrylic pump with refractions index matched(RIM)working fluid.This allows transient investigations of the flow field simultaneously for all blade leading edges.Additional CFD results are validated and used to further support the investigation i.e.for comparing an analog pump system with ideal inflow conditions.