Search

This paper reports, for the first time, the computational performance of SequenceL for mesoscale simulations of large numbers of particles in a microfluidic device via the lattice-Boltzmann method. The performance of SequenceL simulations was assessed against the optimized serial and parallelized (via OpenMP directives) FORTRAN90 simulations. At present, OpenMP directives were not included in interparticle and particle-wall repulsive (steric) interaction calculations due to difficulties that arose from inter-iteration dependencies between consecutive iterations of the do-loops. SequenceL simulations, on the other hand, relied on built-in automatic parallelism. Under these conditions, numerical simulations revealed that the parallelized FORTRAN90 outran the performance of SequenceL by a factor of 2.5 or more when the number of particles was 100 or less. SequenceL, however, outran the performance of the parallelized FORTRAN90 by a factor of 1.3 when the number of particles was 300. Our results show that when the number of particles increased by 30-fold, the computational time of SequenceL simulations increased linearly by a factor of 1.5, as compared to a 3.2-fold increase in serial and a 7.7-fold increase in parallelized FORTRAN90 simulations. Considering SequenceL's efficient built-in parallelism that led to a relatively small increase in computational time with increased number of particles, it could be a promising programming language for computationally-efficient mesoscale simulations of large numbers of particles in microfluidic experiments. (C) 2016 Elsevier B.V. All rights reserved.

Significant improvements in the computational performance of the lattice-Boltzmann (LB) model, coded in FORTRAN90, were achieved through application of enhancement techniques. Applied techniques include optimization of array memory layouts, data structure simplification, random number generation outside the simulation thread(s), code parallelization via OpenMP, and intra- and inter-timestep task pipelining. Effectiveness of these optimization techniques was measured on three benchmark problems: (i) transient flow of multiple particles in a Newtonian fluid in a heterogeneous fractured porous domain, (ii) thermal fluctuation of the fluid at the sub-micron scale and the resultant Brownian motion of a particle, and (iii) non-Newtonian fluid flow in a smooth-walled channel. Application of the aforementioned optimization techniques resulted in an average 21 performance improvement, which could significantly enhance practical uses of the LB models in diverse applications, focusing on the fate and transport of nano-size or micron-size particles in non-Newtonian fluids., Comment: 15 pages, 7 figures