Application of Parallel Computing to Joint-Disconnected Multibody Systems

2005 ◽  
Author(s):  
James Critchley ◽  
Michael McCullough
Keyword(s):  
Parallel Computing ◽  
Execution Time ◽  
Multibody Systems ◽  
Mass Matrix ◽  
Interaction Forces ◽  
Kinematic Constraints ◽  
Fully Coupled ◽  
Block Diagonal ◽  
Parallel Resources

In some areas of application, large multibody systems are encountered which are not completely coupled by kinematic constraints (joints). For these “joint-disconnected” systems, the generalized mass matrix is block diagonal and the coordinate accelerations associated with each block may be solved with independent dynamic inversions. In the absence of a complex (fully coupled) mass matrix, coupling through interaction forces becomes the dominant issue in parallel multibody solutions. An implementation which addresses these issues is described and evaluated on two current parallel platforms. The results demonstrate that inexpensive parallel resources can significantly improve the execution time.

Energy-Momentum Conserving Time Integration of Modally Reduced Flexible Multibody Systems

2013 ◽  
Author(s):  
Alexander Humer ◽  
Johannes Gerstmayr
Keyword(s):  
Angular Momentum ◽  
Multibody Systems ◽  
Mass Matrix ◽  
Time Integration ◽  
Flexible Multibody Dynamics ◽  
Integration Scheme ◽  
Integration Schemes ◽  
Flexible Multibody ◽  
Floating Frame ◽  
Energy And Momentum

Many conventional time integration schemes frequently adopted in flexible multibody dynamics fail to retain the fundamental conservation laws of energy and momentum of the continuous time domain. Lack of conservation, however, in particular of angular momentum, may give rise to unexpected, unphysical results. To avoid such problems, a scheme for the consistent integration of modally reduced multibody systems subjected to holonomic constraints is developed in the present paper. As opposed to the conventional approach, in which the floating frame of reference formulation is combined with component mode synthesis for approximating the flexible deformation, an alternative, recently proposed formulation based on absolute coordinates is adopted in the analysis. Owing to the linear relationship between the generalized coordinates and the absolute displacement, the inertia terms in the equations of motion attain a very simple structure. The mass matrix remains independent of the current state of deformation and the velocity dependent term known from the floating frame approach vanishes due to the absence of relative coordinates. These advantageous properties facilitate the construction of an energy and momentum consistent integration scheme. By the mid-point rule, algorithmic conservation of both linear and angular momentum is achieved. In order to consistently integrate the total energy of the system, the discrete derivative needs to be adopted when evaluating the strain energy gradient and the derivative of the algebraic constraint equations.

Parallel Efficiency of Lagrange Multipliers Based Divide and Conquer Algorithm for Dynamics of Multibody Systems

2011 ◽  
Author(s):  
Paweł Malczyk ◽  
Janusz Fra¸czek
Keyword(s):  
Parallel Computing ◽  
Lagrange Multipliers ◽  
Multibody Systems ◽  
Open Loop ◽  
Divide And Conquer ◽  
Dynamics Simulation ◽  
Parallel Computer ◽  
Optimal Order ◽  
Parallel Efficiency ◽  
Divide And Conquer Algorithm

Efficient dynamics simulations of complex multibody systems are essential in many areas of computer aided engineering and design. As parallel computing resources has become more available, researchers began to reformulate existing algorithms or to create new parallel formulations. Recent works on dynamics simulation of multibody systems include sequential recursive algorithms as well as low order, exact or iterative parallel algorithms. The first part of the paper presents an optimal order parallel algorithm for dynamics simulation of open loop chain multibody systems. The proposed method adopts a Featherstone’s divide and conquer scheme by using Lagrange multipliers approach for constraint imposition and dependent set of coordinates for the system state description. In the second part of the paper we investigate parallel efficiency measures of the proposed formulation. The performance comparisons are made on the basis of theoretical floating-point operations count. The main part of the paper is concetrated on experimental investigation performed on parallel computer using OpenMP threads. Numerical experiments confirm good overall efficiency of the formulation in case of modest parallel computing resources available and demonstrate certain computational advantages over sequential versions.

scholarly journals Performance evaluation of MapReduce using full virtualisation on a departmental cloud

2011 ◽  
Vol 21 (2) ◽  
pp. 275-284 ◽  
Author(s):  
Horacio González-Vélez ◽  
Maryam Kontagora
Keyword(s):  
Parallel Computing ◽  
Performance Evaluation ◽  
Execution Time ◽  
Programming Model ◽  
Peak Performance ◽  
Distributed Environment ◽  
Maximum Peak ◽  
Distributed Parallel Computing

Performance evaluation of MapReduce using full virtualisation on a departmental cloudThis work analyses the performance of Hadoop, an implementation of the MapReduce programming model for distributed parallel computing, executing on a virtualisation environment comprised of 1+16 nodes running the VMWare workstation software. A set of experiments using the standard Hadoop benchmarks has been designed in order to determine whether or not significant reductions in the execution time of computations are experienced when using Hadoop on this virtualisation platform on a departmental cloud. Our findings indicate that a significant decrease in computing times is observed under these conditions. They also highlight how overheads and virtualisation in a distributed environment hinder the possibility of achieving the maximum (peak) performance.

scholarly journals Analysis and optimization of a $k$-chain reduction algorithm for distributed computer systems

2016 ◽  
pp. 318-328
Author(s):  
М.Г. Курносов
Keyword(s):  
Growth Rate ◽  
Parallel Computing ◽  
Waiting Time ◽  
Execution Time ◽  
Reduction Algorithm ◽  
Computer Clusters ◽  
Running Time ◽  
Optimal Value ◽  
Logp Model ◽  
Distributed Computer Systems

В модели параллельных вычислений LogP построено аналитическое выражение времени выполнения алгоритма $k$ параллельных цепочек для реализации корневой редукции на распределенных вычислительных системах (ВС). По построенной функциональной зависимости найдено оптимальное значение числа $k$ параллельных цепочек, при котором алгоритм характеризуется минимальным в модели LogP временем выполнения. На основе этого создан алгоритм с оптимальным числом параллельных цепочек. Для сокращения времени ожидания корневым процессом результатов частичных редукций разработан алгоритм с адаптивным числом параллельных цепочек. Зависимость времени выполнения созданных алгоритмов от числа процессов имеет порядок роста $O(\sqrt{P})$, что эффективнее по сравнению с линейным $\Omega(P)$ временем выполнения исходного алгоритма. Алгоритмы реализованы в стандарте MPI и исследованы на вычислительных кластерах с сетями связи стандарта InfiniBand QDR. In the LogP model of parallel computing, an analytical expression of the $k$-chain algorithm's execution time is derived. The optimal value of $k$ in the LogP model is found. A new algorithm based on the optimal value of $k$ is developed. For the reduction of root process's waiting time, an algorithm with an adaptive number of chains is proposed. The dependence of the execution time of the proposed algorithm on the number of processes has a growth rate of O(sqrt(P)), which is more efficient compared to the linear running time of the original $k$-chain algorithm. The proposed algorithms are implemented in the MPI standard and studied on computer clusters with InfiniBand QDR networks.

Efficient Parallel Computer Simulation of the Motion Behaviors of Closed-Loop Multibody Systems

2007 ◽  
Author(s):  
Shanzhong Duan ◽  
Andrew Ries
Keyword(s):  
Parallel Computing ◽  
Multibody Systems ◽  
Closed Loop ◽  
High Efficiency ◽  
Equations Of Motion ◽  
Parallel Computer ◽  
Constraint Forces ◽  
Computing Systems ◽  
Closed Loops

This paper presents an efficient parallelizable algorithm for the computer-aided simulation and numerical analysis of motion behaviors of multibody systems with closed-loops. The method is based on cutting certain user-defined system interbody joints so that a system of independent multibody subchains is formed. These subchains interact with one another through associated unknown constraint forces fc at the cut joints. The increased parallelism is obtainable through cutting joints and the explicit determination of associated constraint forces combined with a sequential O(n) method. Consequently, the sequential O(n) procedure is carried out within each subchain to form and solve the equations of motion while parallel strategies are performed between the subchains to form and solve constraint equations concurrently. For multibody systems with closed-loops, joint separations play both a role of creation of parallelism for computing load distribution and a role of opening a closed-loop for use of the O(n) algorithm. Joint separation strategies provide the flexibility for use of the algorithm so that it can easily accommodate the available number of processors while maintaining high efficiency. The algorithm gives the best performance for the application scenarios for n>>1 and n>>m, where n and m are number of degree of freedom and number of constraints of a multibody system with closed-loops respectively. The algorithm can be applied to both distributed-memory parallel computing systems and shared-memory parallel computing systems.

Sign in / Sign up

Export Citation Format

Share Document