中心刚体-柔性梁耦合系统离散模型的研究

采用数值仿真对由中心刚体、柔性梁组成的刚-柔耦合系统的动力学离散模型进行了研究.考虑到刚柔-耦合系统的控制方程没有精确解析解,只能寻求数值解,最广泛使用的离散方法是有限元,但其广义坐标数目过于庞大,因此本文探讨了采用经典结构动力学中不同边界的模态函数离散动边界下刚柔耦合动力学方程的可行性及各自的优劣,得到刚柔耦合系统的模态缩减规律.     

DYNAMIC ANALYSIS OF SPINNING DEPLOYMENT OF A SOLAR SAIL COMPOSED OF VISCOELASTIC MEMBRANES

近年来, 可用于航天器推进的太阳帆自旋展开技术引起人们广泛关注. 这类太阳帆可视为由中心旋转毂轮、若干柔性绳索、太阳帆薄膜和集中质量等组成的刚柔耦合多体系统.为了对系统中的太阳帆薄膜进行建模, 提出了基于绝对节点坐标方法描述的黏弹性薄板单元, 并对其有效性进行了验证.针对简化的"IKAROS"自旋展开太阳帆系统, 采用结合自然坐标方法与绝对节点坐标方法的绝对坐标方法对其进行建模, 采用广义-α方法对大规模系统动力学方程进行求解.研究了黏弹性太阳帆薄膜自旋展开过程的动力学特性, 讨论了薄膜的黏弹性阻尼对自旋展开过程的影响规律.     

计及热应变的空间曲梁的刚-柔耦合动力学

研究带中心刚体的作大范围运动的空间曲梁的刚-柔耦合动力学.结合混合坐标法和绝对坐标法的特点,取与中心刚体大范围运动有关的变量和柔性梁各单元节点相对中心刚体连体基的位移和斜率作为广义坐标,建立了一种新的柔性梁的刚柔耦合模型.基于精确的应变和位移的关系式,根据Jourdian速度变分原理,建立了带中心刚体柔性曲梁的有限元离散的动力学方程.数值对比了空间曲梁系统和空间直梁系统的刚柔耦合动力学性质,用能量守恒规律验证了文中曲梁模型的合理性.在此基础上,在应变能中计及热应变,研究温度增高引起的曲梁的热膨胀对系统的动力学性态的影响.     

A HIGH-ORDER RIGID-FLEXIBLE COUPLING MODEL OF A ROTATING FLEXIBLE BEAM UNDER LARGE DEFORMATION

对在平面内做大范围转动的中心刚体-柔性梁系统的刚柔耦合建模理论进行了深入研究,建立了系统的高次耦合动力学模型. 该动力学模型考虑了柔性梁横向弯曲变形和纵向伸长变形,且在纵向位移中计及由于横向变形而引起的纵向缩短项,即非线性耦合变形项,并保留了与非线性耦合项相关的一些高阶项,最终得到了系统的高次刚柔耦合动力学方程. 由此得到的动力学方程不仅能适用于柔性梁的小变形问题,也同样适用于大变形问题,弥补了一次近似耦合模型在处理柔性梁大变形问题上的不足. 通过与绝对节点坐标法以及一次近似耦合模型的对比验证了高次耦合模型的正确性.     

两类基于局部标架梁单元的闭锁缓解方法

对于大转动、大变形柔性体的刚柔耦合动力学问题,基于李群SE(3)局部标架(local frame formulation, LFF)的建模方法能够规避刚体运动带来的几何非线性问题,离散数值模型中广义质量矩阵与切线刚度矩阵满足刚体变换的不变性,可明显地提高柔性多体系统动力学问题的计算效率. 有限元方法中,闭锁问题是导致单元收敛性能低下的主要原因, 例如梁单元的剪切以及泊松闭锁.多变量变分原理是缓解梁、板/壳单元闭锁的有效手段. 该方法不仅离散位移场,同时离散应力场或应变场, 可提高应力与应变的计算精度. 本文基于上述局部标架,研究几类梁单元的闭锁处理方法, 包括几何精确梁(geometrically exact beam formulation, GEBF)与绝对节点坐标(absolute nodal coordinate formulation, ANCF)梁单元. 其中, 采用Hu-Washizu三场变分原理缓解几何精确梁单元中的剪切闭锁,采用应变分解法缓解基于局部标架的ANCF全参数梁单元中的泊松闭锁. 数值算例表明,局部标架的梁单元在描述高转速或大变形柔性多体系统时,可消除刚体运动带来的几何非线性, 极大地减少系统质量矩阵和刚度矩阵的更新次数.缓解闭锁后的几类局部标架梁单元收敛性均得到了明显提升.      

径向基点插值法在旋转柔性梁动力学中的应用

将无网格径向基点插值法用于旋转柔性梁的动力学分析. 利用无网格方法对柔性梁的变形场进行离散,考虑梁的纵向拉伸变形和横向弯曲变形,并计入横向弯曲变形引起的纵向缩短,即非线性耦合项,运用第二类拉格朗日方程推导得到系统刚柔耦合动力学方程. 将无网格径向基点插值法的仿真结果有限元法和假设模态法进行比较分析,说明假设模态法的局限性,并表明其作为一种柔性体离散方法在刚柔耦合多体系统动力学的研究中具有可推广性,并讨论了径向基形状参数的影响. 同时运用3 种求解系统动力学方程的方法:纽马克方法、4阶龙格库塔法、亚当姆斯预报校正法,并比较各方法的计算效率, 结果表明纽马克方法最快.      

大变形薄板多体系统的动力学建模

基于非线性弹性理论,考虑剪切应变和横向应变,用绝对节点坐标法建立了大变形矩形薄板的动力学变分方程;为了提高非线性刚度阵的计算效率,根据非线性刚度阵与广义坐标阵的函数关系式,在非线性刚度阵中分离出广义坐标阵,从而避免了每个时间步长的单元刚度阵的积分运算。在此基础上,引入运动学约束关系,建立了大变形薄板系统第一类拉格朗日方程,对重力作用下大变形二连板进行数值仿真。计算结果表明:随着薄板的柔度增大,低频的弯曲变形与高频拉伸变形的耦合愈加显著;此外,系统机械能守恒验证了该模型正确性。     

索网式展开天线网面精度调整的控制方法

针对索网式星载可展开天线调整节点之间强耦合的特点,本文研究了展开天线网面精度调整的方法.首先,基于Lagrange方程建立了网面物理模型,并推导了网面的刚度矩阵.基于刚度矩阵,在仅知道调整索节点位置量的条件下,提出了索网反射面调整的自适应控制方法.仿真结果验证了模型和控制方法的可行性与有效性.     

完备化ANCF薄板单元及在钢板弹簧动力学建模中的应用

绝对节点坐标方法已在多体系统动力学研究中广泛应用, 但常用来描述板壳类结构的薄板单元, 由于梯度不完备而无法直接用于带有初始弯曲参考构型的柔性体变形描述. 为避免全参数板单元建立车辆钢板弹簧模型时存在的严重截面闭锁问题, 拟采用薄板单元用于板簧建模. 为此, 探索了将现有绝对节点坐标薄板单元纳入一般连续介质力学弹性力表达的方法, 采用中面上单位法向量作为单元厚度方向的梯度向量, 从而得到了完备化的薄板单元及其描述初始弯曲构型时消除初应变的方法. 在此基础上通过定义簧片的未变形构型, 在钢板弹簧中引入可控的预应力, 实现对钢板弹簧装配过程的准确模拟. 通过数值算例验证了本方法的正确性. 建立了车辆钢板弹簧模型, 通过建立在簧片上的局部坐标系实现接触点的跨单元搜索, 并采用惩罚函数法和平滑化的库伦摩擦模型施加簧片间的接触力. 引入参考节点的概念建立了整合车身与吊耳及其机构运动关系的刚柔耦合模型.}}      

完备化ANCF薄板单元及在钢板弹簧动力学建模中的应用

绝对节点坐标方法已在多体系统动力学研究中广泛应用,但常用来描述板壳类结构的薄板单元,由于梯度不完备而无法直接用于带有初始弯曲参考构型的柔性体变形描述.为避免全参数板单元建立车辆钢板弹簧模型时存在的严重截面闭锁问题,拟采用薄板单元用于板簧建模.为此,探索了将现有绝对节点坐标薄板单元纳入一般连续介质力学弹性力表达的方法,采用中面上单位法向量作为单元厚度方向的梯度向量,从而得到了完备化的薄板单元及其描述初始弯曲构型时消除初应变的方法.在此基础上通过定义簧片的未变形构型,在钢板弹簧中引入可控的预应力,实现对钢板弹簧装配过程的准确模拟.通过数值算例验证了本方法的正确性.建立了车辆钢板弹簧模型,通过建立在簧片上的局部坐标系实现接触点的跨单元搜索,并采用惩罚函数法和平滑化的库伦摩擦模型施加簧片间的接触力.引入参考节点的概念建立了整合车身与吊耳及其机构运动关系的刚柔耦合模型.     

Deployment dynamics of a simplified spinning IKAROS solar sail via absolute coordinate based method

The spinning solar sail of large scale has been well developed in recent years. Such a solar sail can be considered as a rigid-flexible multibody system mainly composed of a spinning central rigid hub, a number of flexible thin tethers, sail membranes, and tip masses. A simplified interplanetary kite-craft accelerated by radiation of the Sun (IKAROS) model is established in this study by using the absolute-coordinate-based (ACB) method that combines the natural coordinate formulation (NCF) describing the central rigid hub and the absolute nodal coordinate formulation (ANCF) describing flexible parts. The initial configuration of the system in the second-stage deployment is determined through both dynamic and static analyses. The huge set of stiff equations of system dynamics is solved by using the generalized-alpha method, and thus the deployment dynamics of the system can be well understood.     

求解大柔度梁/绳索的ANCF单元及隐式迭代格式

为模拟大柔度梁/绳索结构的变形和大范围运动,基于绝对节点坐标方法ANCF(Absolute nodal coordi-nate formulation)和HHT(Hilber-Hughes-Taylor)积分方法,建立了ANCF单元的隐式动力学迭代格式.得到了简洁的节点等效力向量,且进一步导出了切线刚度矩阵的全部公式,并基于罚方法实现了对不同ANCF单元的铰接和刚性连接.分别对弦振动、双杆摆和T字摆、柔绳进行了数值模拟,数值结果与理论解及已有结果吻合良好,验证了本文方法的正确性.     

Use of B-spline surface to model large-deformation continuum plates: procedure and applications

The absolute nodal coordinate formulation (ANCF) has been used in the analysis of large deformation of flexible multibody systems that encompass belt drive, rotor blade, and cable applications. As demonstrated in the literature, the ANCF finite elements are ideal for isogeometric analysis. The purpose of this investigation is to establish a relationship between the B-splines, which are widely used in the geometric modeling, and the ANCF finite elements in order to construct continuum models of large-deformation geometries. This paper proposes a simplified approach to map the B-spline surfaces into ANCF thin plate elements. Matrix representation of the mapping process is established and examined through numerical examples successfully. The matrix representation of the mapping process is used because of its suitability of computer coding and to minimize the calculation time. The error estimation is carried out by analyzing the gap between the points of each ANCF element and the corresponding points of the portion of the B-spline surface. The Hausdorff distance is used to study the effect of the number of control points, the degree of interpolation, and the knot multiplicity on the mapped geometry. It is found that cubic interpolation is recommended for optimizing the accuracy of mapping the B-spline surface to ANCF thin plate elements. It is found that thin plate element in ANCF missing a number of basis functions which considered a source of error between the two surfaces, as well as it does not allow to converting the ANCF thin plate elements model to B-spline surface. In this investigation, an application example of modeling large-size wind turbine blade with uniform structure is illustrated. The use of the continuum plate elements in modeling flexible blades is more efficient because of the relative scale between the plate thickness and its length and width and the high flexibility of its structure. The numerical results are compared with the results of ANSYS code with a good agreement. The dynamic simulation for mapped surface model shows a numerical convergence, which ensures the ability of using the proposed approach for applications of dynamics for design and computer-aided design.     

On the formulation of the planar ANCF triangular finite elements

In this paper, new planar isoparametric triangular finite elements (FE) based on the absolute nodal coordinate formulation (ANCF) are developed. The proposed ANCF elements have six coordinates per node: two position coordinates that define the absolute position vector of the node and four gradient coordinates that define vectors tangent to coordinate lines (parameters) at the same node. To shed light on the importance of the element geometry and to facilitate the development of some of the new elements presented in this paper, two different parametric definitions of the gradient vectors are used. The first parametrization, called area parameterization, is based on coordinate lines along the sides of the element in the reference configuration, while the second parameterization, called Cartesian parameterization, employs coordinate lines defined along the axes of the structure (body) coordinate system. The fundamental differences between the ANCF parameterizations used in this investigation and the parametrizations used for conventional finite elements are highlighted. The Cartesian parameterization serves as a unique standard for the triangular FE assembly. To this end, a transformation matrix that defines the relationship between the area and the Cartesian parameterizations is introduced for each element in order to allow for the use of standard FE assembly procedure and define the structure (body) inertia and elastic forces. Using Bezier geometry and a linear mapping, cubic displacement fields of the new ANCF triangular elements are systematically developed. Specifically, two new ANCF triangular finite elements are developed in this investigation, namely four-node mixed-coordinate and three-node ANCF triangles. The performance of the proposed new ANCF elements is evaluated by comparison with the conventional linear and quadratic triangular elements as well as previously developed ANCF rectangular and triangular elements. The results obtained in this investigation show that in the case of small and large deformations as well as finite rotations, all the elements considered can produce correct results, which are in a good agreement if appropriate mesh sizes are used.     

Topology optimization of a flexible multibody system with variable-length bodies described by ALE–ANCF

Recent years have witnessed the application of topology optimization to flexible multibody systems (FMBS) so as to enhance their dynamic performances. In this study, an explicit topology optimization approach is proposed for an FMBS with variable-length bodies via the moving morphable components (MMC). Using the arbitrary Lagrangian–Eulerian (ALE) formulation, the thin plate elements of the absolute nodal coordinate formulation (ANCF) are used to describe the platelike bodies with variable length. For the thin plate element of ALE–ANCF, the elastic force and additional inertial force, as well as their Jacobians, are analytically deduced. In order to account for the variable design domain, the sets of equivalent static loads are reanalyzed by introducing the actual and virtual design domains so as to transform the topology optimization problem of dynamic response into a static one. Finally, the novel MMC-based topology optimization method is employed to solve the corresponding static topology optimization problem by explicitly evolving the shapes and orientations of a set of structural components. The effect of the minimum feature size on the optimization of an FMBS is studied. Three numerical examples are presented to validate the accuracy of the thin plate element of ALE–ANCF and the efficiency of the proposed topology optimization approach, respectively.     

Contact dynamics of elasto-plastic thin beams simulated via absolute nodal coordinate formulation

Under the frame of multibody dynamics, the contact dynamics of elasto-plastic spatial thin beams is numerically studied by using the spatial thin beam elements of absolute nodal coordinate formulation (ANCF). The inter-nal force of the elasto-plastic spatial thin beam element is derived under the assumption that the plastic strain of the beam element depends only on its longitudinal deformation. A new body-fixed local coordinate system is introduced into the spatial thin beam element of ANCF for efficient con-tact detection in the contact dynamics simulation. The linear isotropic hardening constitutive law is used to describe the elasto-plastic deformation of beam material, and the classical return mapping algorithm is adopted to evaluate the plastic strains. A multi-zone contact approach of thin beams previ-ously proposed by the authors is also introduced to detect the multiple contact zones of beams accurately, and the penalty method is used to compute the normal contact force of thin beams in contact. Four numerical examples are given to demonstrate the applicability and effectiveness of the pro-posed elasto-plastic spatial thin beam element of ANCF for flexible multibody system dynamics.     

Dynamic simulation of liquid-filled flexible multibody systems via absolute nodal coordinate formulation and SPH method

A new computation method is proposed to study the coupled dynamics of a partially liquid-filled flexible multibody system, where the liquid is modeled by using the Smoothed Particle Hydrodynamics (SPH) method and the flexible bodies are described by using the Absolute Nodal Coordinate Formulation (ANCF). Extra virtual particles are introduced and embedded in the liquid neighboring the rigid or flexible boundaries in order to prevent field particles from penetrating the boundary and force them to follow the deformation of flexible boundary. The interaction forces between the liquid and the flexible multibody system are transmitted by the virtual particles. The domain decomposition is used to improve the efficiency of interaction detection in SPH computation. A predictor-corrector scheme is used to solve the governing equations of liquid discretized by SPH particles. The generalized-alpha method based on sparse matrix storage skill is used to solve a huge set of dynamic equations of the multibody system. The OpenMP+OpenACC based parallel computation skills are embedded in the iteration processes to speed up the computation efficiency. Finally, three numerical examples are given to validate the proposed computation method.