Optimization of Hydraulic Machinery Bladings by Multilevel CFD Techniques

<p>The numerical design optimization for complex hydraulic machinery bladings requires a high number of design parameters and the use of a precise CFD solver yielding high computational costs. To reduce the CPU time needed, a multilevel CFD method has been developed. First of all, the 3D blade...

Full description

Bibliographic Details
Main Authors: Thum Susanne, Schilling Rudolf
Format: Article
Language:English
Published: Hindawi Limited 2005-01-01
Series:International Journal of Rotating Machinery
Online Access:http://www.hindawi.net/access/get.aspx?journal=ijrm&volume=2005&pii=S1023621X02505029
id doaj-24e95acc532a407986db754f1a80d082
record_format Article
spelling doaj-24e95acc532a407986db754f1a80d0822020-11-24T23:44:50ZengHindawi LimitedInternational Journal of Rotating Machinery1023-621X2005-01-0120052161167Optimization of Hydraulic Machinery Bladings by Multilevel CFD TechniquesThum SusanneSchilling Rudolf<p>The numerical design optimization for complex hydraulic machinery bladings requires a high number of design parameters and the use of a precise CFD solver yielding high computational costs. To reduce the CPU time needed, a multilevel CFD method has been developed. First of all, the 3D blade geometry is parametrized by means of a geometric design tool to reduce the number of design parameters. To keep geometric accuracy, a special B-spline modification technique has been developed. On the first optimization level, a quasi-3D Euler code (EQ3D) is applied. To guarantee a sufficiently accurate result, the code is calibrated by a Navier-Stokes recalculation of the initial design and can be recalibrated after a number of optimization steps by another Navier-Stokes computation. After having got a convergent solution, the optimization process is repeated on the second level using a full 3D Euler code yielding a more accurate flow prediction. Finally, a 3D Navier-Stokes code is applied on the third level to search for the optimum optimorum by means of a fine-tuning of the geometrical parameters. To show the potential of the developed optimization system, the runner blading of a water turbine having a specific speed <math alttext="$n_{q}=41$"><msub><mi>n</mi> <mi>q</mi> </msub> <mo>=</mo> <mn>41</mn> </math> <math alttext="un{1/min}"><mrow><mn>1</mn> <mo>/</mo> <mrow><mtext>min</mtext> </mrow> </mrow> </math> was optimized applying the multilevel approach.</p> http://www.hindawi.net/access/get.aspx?journal=ijrm&volume=2005&pii=S1023621X02505029
collection DOAJ
language English
format Article
sources DOAJ
author Thum Susanne
Schilling Rudolf
spellingShingle Thum Susanne
Schilling Rudolf
Optimization of Hydraulic Machinery Bladings by Multilevel CFD Techniques
International Journal of Rotating Machinery
author_facet Thum Susanne
Schilling Rudolf
author_sort Thum Susanne
title Optimization of Hydraulic Machinery Bladings by Multilevel CFD Techniques
title_short Optimization of Hydraulic Machinery Bladings by Multilevel CFD Techniques
title_full Optimization of Hydraulic Machinery Bladings by Multilevel CFD Techniques
title_fullStr Optimization of Hydraulic Machinery Bladings by Multilevel CFD Techniques
title_full_unstemmed Optimization of Hydraulic Machinery Bladings by Multilevel CFD Techniques
title_sort optimization of hydraulic machinery bladings by multilevel cfd techniques
publisher Hindawi Limited
series International Journal of Rotating Machinery
issn 1023-621X
publishDate 2005-01-01
description <p>The numerical design optimization for complex hydraulic machinery bladings requires a high number of design parameters and the use of a precise CFD solver yielding high computational costs. To reduce the CPU time needed, a multilevel CFD method has been developed. First of all, the 3D blade geometry is parametrized by means of a geometric design tool to reduce the number of design parameters. To keep geometric accuracy, a special B-spline modification technique has been developed. On the first optimization level, a quasi-3D Euler code (EQ3D) is applied. To guarantee a sufficiently accurate result, the code is calibrated by a Navier-Stokes recalculation of the initial design and can be recalibrated after a number of optimization steps by another Navier-Stokes computation. After having got a convergent solution, the optimization process is repeated on the second level using a full 3D Euler code yielding a more accurate flow prediction. Finally, a 3D Navier-Stokes code is applied on the third level to search for the optimum optimorum by means of a fine-tuning of the geometrical parameters. To show the potential of the developed optimization system, the runner blading of a water turbine having a specific speed <math alttext="$n_{q}=41$"><msub><mi>n</mi> <mi>q</mi> </msub> <mo>=</mo> <mn>41</mn> </math> <math alttext="un{1/min}"><mrow><mn>1</mn> <mo>/</mo> <mrow><mtext>min</mtext> </mrow> </mrow> </math> was optimized applying the multilevel approach.</p>
url http://www.hindawi.net/access/get.aspx?journal=ijrm&volume=2005&pii=S1023621X02505029
work_keys_str_mv AT thumsusanne optimizationofhydraulicmachinerybladingsbymultilevelcfdtechniques
AT schillingrudolf optimizationofhydraulicmachinerybladingsbymultilevelcfdtechniques
_version_ 1725498209915109376