Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Civil-Comp Proceedings
ISSN 1759-3433
CCP: 90
PROCEEDINGS OF THE FIRST INTERNATIONAL CONFERENCE ON PARALLEL, DISTRIBUTED AND GRID COMPUTING FOR ENGINEERING
Edited by:
Paper 41

Solution of Multibody System Dynamics Problems on Client-Server Architectures with Pipe Data Interchange

V.V. Getmanski, A.S. Gorobtsov and A.E. Andreev

Volgograd State Technical University, Russia

Full Bibliographic Reference for this paper
V.V. Getmanski, A.S. Gorobtsov, A.E. Andreev, "Solution of Multibody System Dynamics Problems on Client-Server Architectures with Pipe Data Interchange", in , (Editors), "Proceedings of the First International Conference on Parallel, Distributed and Grid Computing for Engineering", Civil-Comp Press, Stirlingshire, UK, Paper 41, 2009. doi:10.4203/ccp.90.41
Keywords: multibody systems, rigid bodies dynamics, parallel processing, nonlinear structural mechanics, computer aided engineering, differential algebraic equations.

Summary
In many systems there are restrictions on the dimensions of the generated models which can be solved using high-performance computing. The matrix of a system that describes a multibody model for the majority of mechanical problems has a sparse structure which is reduced to a block-diagonal representation for certain sorting of bodies in the model. This allows the use of parallel processing at the calculation stage [1].

The speed up condition is that transfer time does not surpass calculation time. In the case of a complex multibody divided system, calculating the number of transfers increases with increasing task scale. Therefore, it is possible to speak about efficiency of the parallel calculation only for the case of large models on a multiprocessing computer with shared memory, in particular, on multicore processors. The use of multicore architecture enables the acceleration of the calculation of complex multibody models.

Two approaches to model decomposition on weak subsystems are possible [2]. The first is based on graph decomposition and the second approach is based on the analysis of a portrait of a matrix.

The computer realisation of a synchronous parallel solution is based on a client-server architecture with pipe data exchange. In the case of large structures on a multicore architecture, an acceleration is achieved.

The problem of modelling the passage of a blast wave on a multilayered plate with a non-uniform front distribution is considered for effective parallellization. For the inverse dynamics control problem solution, a four legged walking machine with two fingers on feet was simulated [3]. The methods were implemented using the multibody software FRUND [4].

The use of parallel calculations in this case allows the distributed calculation to be carried out. Each set of parameters pays off independently and determines the most stable movement variant. The methods considered enable the solution of new classes of problems and the possibility of an increased calculation speed.

References
1
A.S. Gorobtsov, "Parallel Computing of the Multibody Systems of Large Dimensions", 5th. European Congress on Computational Methods in Applied Sciences and Engineering (ECCOMAS 2008), 2008.
2
D. Negrut, "A Topology Based Approach for Exploiting Sparsity in Multibody Dynamics. The Cartesian Formulation", in R. Serban, F.A. Potra, E.J. Haug, (Editors), "Mechanics of Structures and Machines", 25(3), 1997.
3
A.S. Gorobtsov, "Theoretical Investigations of the Control Movement of the CLAWAR at Statically Unstable Regimes" in M.K. Habib, (Editor), "Bionspiration and Robotics: Walking and Climbing Robots", Vienna, Austria, 95-106, 2007.
4
FRUND Software, http://frund.vstu.ru

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £72 +P&P)