Parallel Computation on Multicore Processors Using Explicit Form of the Finite Element Method and C++ Standard Libraries

Abstract

In this paper, the form of modifications of the existing sequential code written in C or C++ programming language for the calculation of various kind of structures using the explicit form of the Finite Element Method (Dynamic Relaxation Method, Explicit Dynamics) in the NEXX system is introduced. The NEXX system is the core of engineering software NEXIS, Scia Engineer, RFEM and RENEX. It has the possibilities of multithreaded running, which can now be supported at the level of native C++ programming language using standard libraries. Thanks to the high degree of abstraction that a contemporary C++ programming language provides, a respective library created in this way can be very generalized for other purposes of usage of parallelism in computational mechanics.

Exploring space separation techniques for 3D elastic waves simulations

“…In low and medium frequency domains, the space-time separation has been successfully validated with standard PGD techniques [33,34]. It consists in introducing sepa rated fonctions depending on the time variable in (18) or (19). Regarding wave propagation problems, however, con vergence difficulties have been reported with the space-time separation (M » 100).…”

“…Yet, even if the spatial domain is as simple as a plate, the computational effort remains pro hibitive if appropriate numerical resources are not available. Different approaches have been developed to tackle large scale simulations with parallel computing strategies [18][19][20]. For instance, Zhang et al presented in [20] a parallel explicit solver based on the scaled boundary finite element method, with an efficient pre-computation approach and element-wise operations.…”

Explicit integration of equations of motion solved on computer cluster

“…This algorithm belongs to the cornerstones of the entire parallel model. The rest of the code related to parallel processing on a single workstation is applied similarly to the algorithm proposed for GPGPU technology [16] for multicore CPU environment [17]. This applies to the integration of respective finite elements and explicit integration of equations of motion applied to all the finite element nodes.…”

A massively parallel explicit solver for elasto-dynamic problems exploiting octree meshes

“…Parallel processing is an attractive technique in high-performance computing to speed up the algorithm by partitioning a serial job into several smaller jobs and distributing them to individual processors for execution [12]. A commonly used parallel processing paradigm is single program, multiple data (SPMD) 1 , i.e., different processors execute the same command on different data sets simultaneously [13].…”