Computational Technology Resources - CCP

Keywords: uncertainty quantification, flow simulation, genetic optimization, sampling methods, traveling salesman problem.

Summary

This paper presents a method to reduce the computation time of sampling methods for uncertainty quantification (UQ) in computational fluid dynamics (CFD) for steadystate flow problems. The acceleration method takes advantage of numerical properties of the mentioned procedure for UQ as well as of a finite-volume method based CFD solver. In the latter it is about the characteristic that the flow field is solved by an iterative procedure. Through that the number of iterations until convergence and hence the computing time depends among others on the initially chosen solution. This implies that the solving process takes less time the better the chosen initial solution is. For instance, good initial solutions for flow problems that are connected with UQ are previously calculated flow fields of other sampling points, because there are only small variations in parameters. This, combined with the property of sampling methods that all necessary sample points can be computed independently of each other, is the basic idea of our acceleration method. Thus, there exist favorable and unfavorable orders to calculate the sampling points. Accordingly, the presented technique is a method to optimize the order in which the sampling points should be calculated with respect to the computation time.

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 £45 +P&P)

	Computational & Technology Resources an online resource for computational, engineering & technology publications
	not logged in - login
Front Page Browse CCP CSETS CTR IJRT Other Authors Search Purchase Guide FAQ Contact us	Civil-Comp Proceedings ISSN 1759-3433 CCP: 105 PROCEEDINGS OF THE NINTH INTERNATIONAL CONFERENCE ON ENGINEERING COMPUTATIONAL TECHNOLOGY Edited by: Paper 72 Acceleration of Sampling Methods for Uncertainty Quantification in Computational Fluid Dynamics C.W.T. Thiem and M. Schäfer Institute of Numerical Methods in Mechanical Engineering and Graduate School Computational Engineering, Technische Universität Darmstadt, Germany doi:10.4203/ccp.105.72 purchase the full-text of this paper Full Bibliographic Reference for this paper , "Acceleration of Sampling Methods for Uncertainty Quantification in Computational Fluid Dynamics", in , (Editors), "Proceedings of the Ninth International Conference on Engineering Computational Technology", Civil-Comp Press, Stirlingshire, UK, Paper 72, 2014. doi:10.4203/ccp.105.72 Keywords: uncertainty quantification, flow simulation, genetic optimization, sampling methods, traveling salesman problem. Summary This paper presents a method to reduce the computation time of sampling methods for uncertainty quantification (UQ) in computational fluid dynamics (CFD) for steadystate flow problems. The acceleration method takes advantage of numerical properties of the mentioned procedure for UQ as well as of a finite-volume method based CFD solver. In the latter it is about the characteristic that the flow field is solved by an iterative procedure. Through that the number of iterations until convergence and hence the computing time depends among others on the initially chosen solution. This implies that the solving process takes less time the better the chosen initial solution is. For instance, good initial solutions for flow problems that are connected with UQ are previously calculated flow fields of other sampling points, because there are only small variations in parameters. This, combined with the property of sampling methods that all necessary sample points can be computed independently of each other, is the basic idea of our acceleration method. Thus, there exist favorable and unfavorable orders to calculate the sampling points. Accordingly, the presented technique is a method to optimize the order in which the sampling points should be calculated with respect to the computation time. 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 £45 +P&P)
Back to top	©Civil-Comp Limited 2023 - terms & conditions