CFD-based analysis and two-level aerodynamic optimization on graphics processing units

Kampolis, IC; Trompoukis, XS; Asouti, VG; Giannakoglou, KC

dc.contributor.author	Kampolis, IC	en
dc.contributor.author	Trompoukis, XS	en
dc.contributor.author	Asouti, VG	en
dc.contributor.author	Giannakoglou, KC	en
dc.date.accessioned	2014-03-01T01:32:59Z
dc.date.available	2014-03-01T01:32:59Z
dc.date.issued	2010	en
dc.identifier.issn	0045-7825	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/20264
dc.subject	Aerodynamic shape optimization	en
dc.subject	Computational fluid dynamics	en
dc.subject	Evolutionary algorithms	en
dc.subject	Graphics processing units	en
dc.subject.classification	Engineering, Multidisciplinary	en
dc.subject.classification	Mathematics, Interdisciplinary Applications	en
dc.subject.classification	Mechanics	en
dc.subject.other	2D flow	en
dc.subject.other	Aerodynamic optimization	en
dc.subject.other	Aerodynamic shape optimization	en
dc.subject.other	Candidate solution	en
dc.subject.other	Convergence criterion	en
dc.subject.other	Different speed	en
dc.subject.other	Double precision	en
dc.subject.other	Evaluation tool	en
dc.subject.other	Flow solver	en
dc.subject.other	GPU implementation	en
dc.subject.other	Graphics Processing Unit	en
dc.subject.other	Hierarchical optimization	en
dc.subject.other	Low level	en
dc.subject.other	Mixed precision	en
dc.subject.other	Optimal solutions	en
dc.subject.other	Parallel efficiency	en
dc.subject.other	Prediction accuracy	en
dc.subject.other	Single precision	en
dc.subject.other	Speed-ups	en
dc.subject.other	Unstructured grid	en
dc.subject.other	Aerodynamics	en
dc.subject.other	Computational efficiency	en
dc.subject.other	Computational fluid dynamics	en
dc.subject.other	Computer graphics equipment	en
dc.subject.other	Evolutionary algorithms	en
dc.subject.other	Fluid dynamics	en
dc.subject.other	Germanium	en
dc.subject.other	Image coding	en
dc.subject.other	Knowledge based systems	en
dc.subject.other	Mathematical operators	en
dc.subject.other	Navier Stokes equations	en
dc.subject.other	Program processors	en
dc.subject.other	Three dimensional	en
dc.subject.other	Viscous flow	en
dc.subject.other	Shape optimization	en
dc.title	CFD-based analysis and two-level aerodynamic optimization on graphics processing units	en
heal.type	journalArticle	en
heal.identifier.primary	10.1016/j.cma.2009.11.001	en
heal.identifier.secondary	http://dx.doi.org/10.1016/j.cma.2009.11.001	en
heal.language	English	en
heal.publicationDate	2010	en
heal.abstract	This paper presents the porting of 2D and 3D Navier-Stokes equations solvers for unstructured grids, from the CPU to the graphics processing unit (GPU; NVIDIA's Ge-Force GTX 280 and 285), using the CUDA language. The performance of the GPU implementations, with single, double or mixed precision arithmetic operations. is compared to that of the CPU code. Issues regarding the optimal handling of the unstructured grid topology on the CPU, particularly for vertex-centered CFD algorithms, are discussed. Restructuring the existing codes was necessary in order to maximize the parallel efficiency of the GPU implementations. The mixed precision implementation, in which the left-hand-side operators are computed with single precision, was shown to bridge the gap between the single and double precision speed-ups. Based on the different speed-ups and prediction accuracy of the aforementioned GPU implementations of the Navier-Stokes equations solver, a hierarchical optimization method which is suitable for GPUs is proposed and demonstrated in inviscid and turbulent 2D flow problems. The search for the optimal solution(s) splits into two levels, both relying upon evolutionary algorithms (EAs) though with different evaluation tools each. The low level EA uses the very fast single precision GPU implementation with relaxed convergence criteria for the inexpensive evaluation of candidate solutions. Promising solutions are regularly broadcast to the high level EA which uses the mixed precision GPU implementation of the same flow solver. Single- and two-objective aerodynamic shape optimization problems are solved using the developed software. (C) 2009 Elsevier B.V. All rights reserved.	en
heal.publisher	ELSEVIER SCIENCE SA	en
heal.journalName	Computer Methods in Applied Mechanics and Engineering	en
dc.identifier.doi	10.1016/j.cma.2009.11.001	en
dc.identifier.isi	ISI:000274572600023	en
dc.identifier.volume	199	en
dc.identifier.issue	9-12	en
dc.identifier.spage	712	en
dc.identifier.epage	722	en