AC 6-12 CFD Simulations: Difference between revisions

From KBwiki
Jump to navigation Jump to search
m (Dave.Ellacott moved page Silver:AC 6-12 CFD Simulations to AC 6-12 CFD Simulations over redirect)
 
(21 intermediate revisions by 4 users not shown)
Line 1: Line 1:
{{AC|front=AC 6-12|description=AC 6-12 Description|testdata=AC 6-12 Test Data|cfdsimulations=AC 6-12 CFD Simulations|evaluation=AC 6-12 Evaluation|qualityreview=AC 6-12 Quality Review|bestpractice=AC 6-12 Best Practice Advice|relatedACs=AC 6-12 Related ACs}}
{{AC|front=AC 6-12|description=AC 6-12 Description|testdata=AC 6-12 Test Data|cfdsimulations=AC 6-12 CFD Simulations|evaluation=AC 6-12 Evaluation|qualityreview=AC 6-12 Quality Review|bestpractice=AC 6-12 Best Practice Advice|relatedUFRs=AC 6-12 Related ACs}}




Line 20: Line 20:
Numerical results are good in the vicinity of sonic line and in the downstream part considering shock waves, reflected shock waves, wakes (compared to experimental data using pressure distribution or field of Mach number isolines). Behind of the leading edge of the upper side of the profile, results achieved by triangular grid are better for any type of used method. Relatively good results were obtained for laminar viscous flow computed by the combination of finite volume (inviscid part) and finite element (viscous part) methods.
Numerical results are good in the vicinity of sonic line and in the downstream part considering shock waves, reflected shock waves, wakes (compared to experimental data using pressure distribution or field of Mach number isolines). Behind of the leading edge of the upper side of the profile, results achieved by triangular grid are better for any type of used method. Relatively good results were obtained for laminar viscous flow computed by the combination of finite volume (inviscid part) and finite element (viscous part) methods.


Convergence to the steady state was examined by log L<sub>2</sub> – residual. Numerical and experimental results are compared using pressure and Mach number distribution along the profile surface with a very good agreement.
Convergence to the steady state was examined by log L<sub>2</sub> &minus; residual. Numerical and experimental results are compared using pressure and Mach number distribution along the profile surface with a very good agreement.


The Reynolds-Averaged Navier-Stokes (RANS) equations were solved by FLUENT 5 code using the RNG k-eps turbulence model for the Reynolds number Re = 1.5 x 10<sup>6</sup>. The obtained pressure distribution on the profile shows a very good agreement with experiment. The comparison of measured and computed energy losses was made as well.
The Reynolds-Averaged Navier-Stokes (RANS) equations were solved by FLUENT 5 code using the RNG k-eps turbulence model for the Reynolds number Re = 1.5 x 10<sup>6</sup>. The obtained pressure distribution on the profile shows a very good agreement with experiment. The comparison of measured and computed energy losses was made as well.
Line 50: Line 50:
=== Boundary Conditions ===
=== Boundary Conditions ===


Numerical simulations were made for the cascade section one and/or two blade passages. The upstream boundary was approximately one chord before the inlet plane of the cascade. The downstream boundary was situated in the distance 1.5 – 2 chords behind the outlet plane. At the inlet boundary, three quantities were given and one was extrapolated. The pressure was given at the outlet boundary. Periodical boundaries were used along the cascade passage. Wall functions were used as wall conditions in calculations made using the FLUENT code and the reflection principle was used in home-made software.
Numerical simulations were made for the cascade section one and/or two blade passages. The upstream boundary was approximately one chord before the inlet plane of the cascade. The downstream boundary was situated in the distance 1.5 &ndash; 2 chords behind the outlet plane. At the inlet boundary, three quantities were given and one was extrapolated. The pressure was given at the outlet boundary. Periodical boundaries were used along the cascade passage. Wall functions were used as wall conditions in calculations made using the FLUENT code and the reflection principle was used in home-made software.


<br style="mso-special-character: line-break" clear="all" />
<br style="mso-special-character: line-break" clear="all" />
Line 58: Line 58:
Two physical models are used:
Two physical models are used:


a)           CFD1 – RANS with RNG k-ε turbulence model or one-equation Spalart-Allmaras model
a)           CFD1 &ndash; RANS with RNG k-&epsilon; turbulence model or one-equation Spalart-Allmaras model


b)          CFD 2 – Euler equations
b)         CFD2 &ndash; Euler equations


In part a) all needed is given by code FLUENT 5, in part b) used boundary conditions are described above.
In part a) all needed is given by code FLUENT 5, in part b) used boundary conditions are described above.
Line 68: Line 68:
=== Numerical Accuracy ===
=== Numerical Accuracy ===


We compared several results for inviscid flows using high resolution different schemes (TVD, ENO,…) on different grids. We also have comparison of numerical and experimental results. The similar way was used for turbulent RANS computations.
We compared several results for inviscid flows using high resolution different schemes (TVD, ENO,&hellip;) on different grids. We also have comparison of numerical and experimental results. The similar way was used for turbulent RANS computations.


<br style="mso-special-character: line-break" clear="all" />
<br style="mso-special-character: line-break" clear="all" />
Line 74: Line 74:
=== CFD Results ===
=== CFD Results ===


CFD1 Numerical simulation of the viscous flow the steam turbine rotor cascade: isolines of density ρ=f(x,y), pressure distribution on the blade, survey of relevant parameters
CFD1 Numerical simulation of the viscous flow the steam turbine rotor cascade: isolines of density &rho;=f(x,y), pressure distribution on the blade, survey of relevant parameters


The files containing results with measured and evaluated parameters are given in Table EXP-B:
The files containing results with measured and evaluated parameters are given in Table EXP-B:


[http://qnetkb.cfms.org.uk/TA6/AC6-12/I/cfd11.jpg cfd11.jpg] (flow field description by density isolines for M<sub>2is</sub> = 1.198)
[[media:AC6-12_cfd11.jpg|cfd11.jpg]] (flow field description by density isolines for M<sub>2is</sub> = 1.198)


[http://qnetkb.cfms.org.uk/TA6/AC6-12/C/cfd12.dat cfd12.dat] (survey of relevant data β<sub>1</sub>, i, M<sub>1</sub>, M<sub>2is</sub>, Re<sub>2is</sub>, M<sub>2</sub>, ζ, β<sub>2</sub> and pressure distribution on the suction and pressure sides p/p0 vers. x/b)
[[media:AC6-12_cfd12.dat|cfd12.dat]] (survey of relevant data &beta;<sub>1</sub>, i, M<sub>1</sub>, M<sub>2is</sub>, Re<sub>2is</sub>, M<sub>2</sub>, &zeta;, &beta;<sub>2</sub> and pressure distribution on the suction and pressure sides p/p0 vers. x/b)


<u> </u>
<u> </u>


'''CFD2''' Numerical simulation of the inviscid flow the steam turbine rotor cascade: isolines of Mach number M=f(x,y), pressure distribution on the blade, survey of relevant parameters
'''CFD2''' Numerical simulation of the inviscid flow the steam turbine rotor cascade: isolines of Mach number M=f(x,y), pressure distribution on the blade, survey of relevant parameters
Line 88: Line 88:
The files containing results with measured and evaluated parameters are given in Table EXP-B:
The files containing results with measured and evaluated parameters are given in Table EXP-B:


[http://qnetkb.cfms.org.uk/TA6/AC6-12/I/cfd21.jpg cfd21.jpg] (flow field description by Mach number isolines for M<sub>2is</sub> = 1.198)
[[media:AC6-12_cfd21.jpg|cfd21.jpg]] (flow field description by Mach number isolines for M<sub>2is</sub> = 1.198)


[http://qnetkb.cfms.org.uk/TA6/AC6-12/C/cfd22.dat cfd22.dat] (survey of relevant data β<sub>1</sub>, i, M<sub>1</sub>, M<sub>2is</sub>, β<sub>2</sub> and pressure distribution on the suction and pressure sides p/p0 vers. x/b)
[[media:AC6-12_cfd22.dat|cfd22.dat]] (survey of relevant data &beta;<sub>1</sub>, i, M<sub>1</sub>, M<sub>2is</sub>, &beta;<sub>2</sub> and pressure distribution on the suction and pressure sides p/p0 vers. x/b)


<u> </u>
<u> </u>


{| style="width: 496.15pt; margin-left: 3.9pt; border-collapse: collapse; mso-padding-alt: 0cm 0cm 0cm 0cm" width="662"
{| style="width: 496.15pt; margin-left: 3.9pt; border-collapse: collapse; mso-padding-alt: 0cm 0cm 0cm 0cm" width="662"
|+ Table CFD-A Summary description of all test cases
| style="width: 49.65pt; border: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="66" valign="top" |
| style="width: 49.65pt; border: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="66" valign="top" |
Name
Name
Line 109: Line 110:
|-
|-
| style="width: 49.65pt; border: solid windowtext 1.0pt; border-top: none; padding: 0cm 3.5pt 0cm 3.5pt" width="66" valign="top" |
| style="width: 49.65pt; border: solid windowtext 1.0pt; border-top: none; padding: 0cm 3.5pt 0cm 3.5pt" width="66" valign="top" |
<center> </center>
<center> </center>
| style="width: 35.4pt; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="47" valign="top" |
| style="width: 35.4pt; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="47" valign="top" |
<center>Re<sub>2is</sub></center>
<center>Re<sub>2is</sub></center>
Line 117: Line 118:
<center>inlet angle</center>
<center>inlet angle</center>


<center>β<sub>1</sub> (deg)</center>
<center>&beta;<sub>1</sub> (deg)</center>
| style="width: 77.95pt; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="104" valign="top" |
| style="width: 77.95pt; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="104" valign="top" |
<center>total pressure</center>
<center>total pressure</center>
Line 129: Line 130:
<center>detailed data</center>
<center>detailed data</center>


<center>M, ρ=f(x,y)</center>
<center>M, &rho;=f(x,y)</center>
| style="width: 63.8pt; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="85" valign="top" |
| style="width: 63.8pt; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="85" valign="top" |
<center>[../../help/glossary.htm DOAP]s</center>
<center>[[DOAP]]s</center>
|-
|-
| style="width: 49.65pt; border: solid windowtext 1.0pt; border-top: none; padding: 0cm 3.5pt 0cm 3.5pt" width="66" valign="top" |
| style="width: 49.65pt; border: solid windowtext 1.0pt; border-top: none; padding: 0cm 3.5pt 0cm 3.5pt" width="66" valign="top" |
===== CFD 1 =====
'''CFD 1'''


<center>viscous</center>
<center>viscous</center>
Line 148: Line 149:
<center>298.65</center>
<center>298.65</center>
| style="width: 77.95pt; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="104" valign="top" |
| style="width: 77.95pt; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="104" valign="top" |
<center>ρ=f(x,y)</center>
<center>&rho;=f(x,y)</center>


<center> </center>
<center> </center>
| style="width: 63.8pt; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="85" valign="top" |
| style="width: 63.8pt; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="85" valign="top" |
<center>p/p<sub>o</sub><nowiki>=f(x/b),</nowiki></center>
<center>p/p<sub>o</sub><nowiki>=f(x/b),</nowiki></center>


<center>ζ, β<sub>2</sub></center>
<center>&zeta;, &beta;<sub>2</sub></center>
|-
|-
| style="width: 49.65pt; border: solid windowtext 1.0pt; border-top: none; padding: 0cm 3.5pt 0cm 3.5pt" width="66" valign="top" |
| style="width: 49.65pt; border: solid windowtext 1.0pt; border-top: none; padding: 0cm 3.5pt 0cm 3.5pt" width="66" valign="top" |
===== CFD 2 =====
'''CFD 2'''


<center>inviscid</center>
<center>inviscid</center>
Line 175: Line 176:
<center>p/p<sub>o</sub><nowiki>=f(x/b),</nowiki></center>
<center>p/p<sub>o</sub><nowiki>=f(x/b),</nowiki></center>


<center>β<sub>2</sub></center>
<center>&beta;<sub>2</sub></center>
|}
|}


<center>Table CFD-A Summary description of all test cases</center>
 


{| style="margin-left: 67.3pt; border-collapse: collapse; mso-padding-alt: 0cm 0cm 0cm 0cm"
{| style="margin-left: 67.3pt; border-collapse: collapse; mso-padding-alt: 0cm 0cm 0cm 0cm"
|+ Table CFD-B Summary description of all available datafiles, and simulated parameters
| style="width: 2.0cm; border: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="76" valign="top" |
| style="width: 2.0cm; border: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="76" valign="top" |
<center> </center>
<center> </center>
| style="width: 134.65pt; border: solid windowtext 1.0pt; border-left: none; padding: 0cm 3.5pt 0cm 3.5pt" width="180" valign="top" |
| style="width: 134.65pt; border: solid windowtext 1.0pt; border-left: none; padding: 0cm 3.5pt 0cm 3.5pt" width="180" valign="top" |
<center>SP1</center>
<center>SP1</center>


<center>isolines M, ρ=f(x,y)</center>
<center>isolines M, &rho;=f(x,y)</center>
| style="width: 6.0cm; border: solid windowtext 1.0pt; border-left: none; padding: 0cm 3.5pt 0cm 3.5pt" width="227" valign="top" |
| style="width: 6.0cm; border: solid windowtext 1.0pt; border-left: none; padding: 0cm 3.5pt 0cm 3.5pt" width="227" valign="top" |
<center>SP 2</center>
<center>SP 2</center>
Line 195: Line 197:
<center>'''CFD 1'''</center>
<center>'''CFD 1'''</center>
| style="width: 134.65pt; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="180" valign="top" |
| style="width: 134.65pt; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="180" valign="top" |
<center>[http://qnetkb.cfms.org.uk/TA6/AC6-12/I/cfd11.jpg cfd11.jpg]</center>
<center>[[media:AC6-12_cfd11.jpg|cfd11.jpg]]</center>
| style="width: 6.0cm; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="227" valign="top" |
| style="width: 6.0cm; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="227" valign="top" |
<center>[http://qnetkb.cfms.org.uk/TA6/AC6-12/C/cfd12.dat cfd12.dat]</center>
<center>[[media:AC6-12_cfd12.dat|cfd12.dat]]</center>
|-
|-
| style="width: 2.0cm; border: solid windowtext 1.0pt; border-top: none; padding: 0cm 3.5pt 0cm 3.5pt" width="76" valign="top" |
| style="width: 2.0cm; border: solid windowtext 1.0pt; border-top: none; padding: 0cm 3.5pt 0cm 3.5pt" width="76" valign="top" |
<center>'''CFD 2'''</center>
<center>'''CFD 2'''</center>
| style="width: 134.65pt; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="180" valign="top" |
| style="width: 134.65pt; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="180" valign="top" |
<center>[http://qnetkb.cfms.org.uk/TA6/AC6-12/I/cfd21.jpg cfd21.jpg]</center>
<center>[[media:AC6-12_cfd21.jpg|cfd21.jpg]]</center>
| style="width: 6.0cm; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="227" valign="top" |
| style="width: 6.0cm; border-top: none; border-left: none; border-bottom: solid windowtext 1.0pt; border-right: solid windowtext 1.0pt; padding: 0cm 3.5pt 0cm 3.5pt" width="227" valign="top" |
<center>[http://qnetkb.cfms.org.uk/TA6/AC6-12/C/cfd22.dat cfd22.dat]</center>
<center>[[media:AC6-12_cfd22.dat|cfd22.dat]]</center>
|}
|}


<center>Table CFD-B Summary description of all available datafiles, and simulated parameters</center>


<br style="mso-special-character: line-break" clear="all" />
<br style="mso-special-character: line-break" clear="all" />
Line 213: Line 214:
=== References ===
=== References ===


[1]   Feistauer M., Felcman J. Theory and applications of numerical schemes for nonlinear convection-diffusion problems and compressible Navier-Stokes equations, Mathematics of Finite Elements and Applications, 175-194, Wiley, Chichester, 1997
[1]   Feistauer M., Felcman J. Theory and applications of numerical schemes for nonlinear convection-diffusion problems and compressible Navier-Stokes equations, Mathematics of Finite Elements and Applications, 175-194, Wiley, Chichester, 1997


[2]   Fialová M., Hyhlík T., Kozel K., Šafařík P. Numerical analysis data on the transonic flow past the profile cascade SE 1050, Proc. of the Seminar “Topical Problems of Fluid Mechanics”, Institute of Thermomechanics, Prague, 45-48, 2001
[2]   Fialová M., Hyhlík T., Kozel K., &#352;afařík P. Numerical analysis data on the transonic flow past the profile cascade SE 1050, Proc. of the Seminar &ldquo;Topical Problems of Fluid Mechanics&rdquo;, Institute of Thermomechanics, Prague, 45-48, 2001


[3]      Fořt J., Fürst J., Halama J., Kozel K.: Numerical Solution of 2D and 3D Transonic Flows through a Cascade, Proc. of 4th International Symposium on Experimental and Computational Aerothermodynamics of Internal Flows (ed. R. Grundmann), Dresden, Vol.1., 231-240, 1999
[3]     Fořt J., Fürst J., Halama J., Kozel K.: Numerical Solution of 2D and 3D Transonic Flows through a Cascade, Proc. of 4th International Symposium on Experimental and Computational Aerothermodynamics of Internal Flows (ed. R. Grundmann), Dresden, Vol.1., 231-240, 1999


[4]      Fořt J., Huněk M., Kozel K., Lain J., Šejna M., Vavřincová M. Numerical simulation of steady and unsteady flow through plane cascades, Lecture Notes on Physics (eds. S.M. Desphande, S.S. Desai, R. Narasimha), Vol.453, 461-465, Springer, Berlin, 1995
[4]     Fořt J., Huněk M., Kozel K., Lain J., &#352;ejna M., Vavřincová M. Numerical simulation of steady and unsteady flow through plane cascades, Lecture Notes on Physics (eds. S.M. Desphande, S.S. Desai, R. Narasimha), Vol.453, 461-465, Springer, Berlin, 1995


[5]   Matas R., Jůza Z., Riffault T. Some experiences with 2D numerical modelling of transonic profile cascade SE 1050, Proc. of the Colloquium “Fluid Dynamics”, Institute of Thermomechanics, Prague, 85-88, 2000
[5]   Matas R., Jůza Z., Riffault T. Some experiences with 2D numerical modelling of transonic profile cascade SE 1050, Proc. of the Colloquium &ldquo;Fluid Dynamics&rdquo;, Institute of Thermomechanics, Prague, 85-88, 2000


<br style="mso-special-character: line-break" clear="all" /><font size="-2" color="#888888">© copyright ERCOFTAC 2004</font><br />
<br style="mso-special-character: line-break" clear="all" /><font size="-2" color="#888888">© copyright ERCOFTAC 2004</font><br />
Line 228: Line 229:




Contributors: Jaromir Prihoda; Karel Kozel - Czech Academy of Sciences
Contributors: Jaromir Prihoda; Karel Kozel - Czech Academy of Sciences
 
 
{{AC|front=AC 6-12|description=AC 6-12 Description|testdata=AC 6-12 Test Data|cfdsimulations=AC 6-12 CFD Simulations|evaluation=AC 6-12 Evaluation|qualityreview=AC 6-12 Quality Review|bestpractice=AC 6-12 Best Practice Advice|relatedACs=AC 6-12 Related ACs}}




[[Category:Application Challenge]]
{{AC|front=AC 6-12|description=AC 6-12 Description|testdata=AC 6-12 Test Data|cfdsimulations=AC 6-12 CFD Simulations|evaluation=AC 6-12 Evaluation|qualityreview=AC 6-12 Quality Review|bestpractice=AC 6-12 Best Practice Advice|relatedUFRs=AC 6-12 Related ACs}}

Latest revision as of 18:45, 11 February 2017

Front Page

Description

Test Data

CFD Simulations

Evaluation

Best Practice Advice




Steam turbine rotor cascade

Application Challenge 6-12               © copyright ERCOFTAC 2004


CFD Simulations

Overview of CFD Simulations

CFD simulations were carried out for inviscid flows using finite volume methods and grids of quadrilaterall cells (H type) or triangular cells by explicit modern (TVD central and upwind, ENO high order) methods or classical (Ron-Ho-Ni cell-vertex) methods and by implicit methods (TVD-Osher preconditioning and ENO second order implicit). The adaptivity of grids was used in explicit and implicit methods using triangular cells.

Numerical results are good in the vicinity of sonic line and in the downstream part considering shock waves, reflected shock waves, wakes (compared to experimental data using pressure distribution or field of Mach number isolines). Behind of the leading edge of the upper side of the profile, results achieved by triangular grid are better for any type of used method. Relatively good results were obtained for laminar viscous flow computed by the combination of finite volume (inviscid part) and finite element (viscous part) methods.

Convergence to the steady state was examined by log L2 − residual. Numerical and experimental results are compared using pressure and Mach number distribution along the profile surface with a very good agreement.

The Reynolds-Averaged Navier-Stokes (RANS) equations were solved by FLUENT 5 code using the RNG k-eps turbulence model for the Reynolds number Re = 1.5 x 106. The obtained pressure distribution on the profile shows a very good agreement with experiment. The comparison of measured and computed energy losses was made as well.

Some disadvantage of the FLUENT code is that the used numerical procedure has not implemented downstream non-reflection boundary conditions.


Simulation Cases CFD1 and CFD2

Solution strategy

Case CFD1 is turbulent computation, case CFD2 is inviscid computation. Basic equations used in CFD1 and CFD2 computations are considered in conservative form (RANS and Euler equations respectively). For RANS computations we used additionally one- or two-equation models needed to compute turbulent viscosity. Finite volume methods were considered in both cases. For inviscid problem we used only our own software. For viscous computation FLUENT 5 was used.

The fluid flow in the cascade was solved as adiabatic compressible air flow using the commercial code FLUENT 5. Two turbulence models were used: RNG k-eps model and Spalart-Allmaras model. The RNG k-eps model was chosen for the application challenge. Wall functions were used as wall boundary conditions.


Computational Domain

We used as a computational domain one period or two periods:

a)           with profile located inside of the domain and upper-lower boundary of the domain are lines where periodicity conditions are fulfilled.

b)          with periodical boundaries and part of the boundary is also lower-upper part of a profile (profile is not located inside of the domain when one period is considered).


Boundary Conditions

Numerical simulations were made for the cascade section one and/or two blade passages. The upstream boundary was approximately one chord before the inlet plane of the cascade. The downstream boundary was situated in the distance 1.5 – 2 chords behind the outlet plane. At the inlet boundary, three quantities were given and one was extrapolated. The pressure was given at the outlet boundary. Periodical boundaries were used along the cascade passage. Wall functions were used as wall conditions in calculations made using the FLUENT code and the reflection principle was used in home-made software.


Application of Physical Models

Two physical models are used:

a) CFD1 – RANS with RNG k-ε turbulence model or one-equation Spalart-Allmaras model

b) CFD2 – Euler equations

In part a) all needed is given by code FLUENT 5, in part b) used boundary conditions are described above.


Numerical Accuracy

We compared several results for inviscid flows using high resolution different schemes (TVD, ENO,…) on different grids. We also have comparison of numerical and experimental results. The similar way was used for turbulent RANS computations.


CFD Results

CFD1 Numerical simulation of the viscous flow the steam turbine rotor cascade: isolines of density ρ=f(x,y), pressure distribution on the blade, survey of relevant parameters

The files containing results with measured and evaluated parameters are given in Table EXP-B:

cfd11.jpg (flow field description by density isolines for M2is = 1.198)

cfd12.dat (survey of relevant data β1, i, M1, M2is, Re2is, M2, ζ, β2 and pressure distribution on the suction and pressure sides p/p0 vers. x/b)

CFD2 Numerical simulation of the inviscid flow the steam turbine rotor cascade: isolines of Mach number M=f(x,y), pressure distribution on the blade, survey of relevant parameters

The files containing results with measured and evaluated parameters are given in Table EXP-B:

cfd21.jpg (flow field description by Mach number isolines for M2is = 1.198)

cfd22.dat (survey of relevant data β1, i, M1, M2is, β2 and pressure distribution on the suction and pressure sides p/p0 vers. x/b)

Table CFD-A Summary description of all test cases

Name

GNDPs
PDPs
(problem definition parameters)
MPs
(measured parameters)
Re2is
M2is
inlet angle
β1 (deg)
total pressure
po1 (Pa)
total temperature
To1 (K)
detailed data
M, ρ=f(x,y)
DOAPs

CFD 1

viscous
1.48 x 106
1.198
70.7
98071.7
298.65
ρ=f(x,y)
p/po=f(x/b),
ζ, β2

CFD 2

inviscid
-
1.19
70.7
-
-
M=f(x,y)
p/po=f(x/b),
β2


Table CFD-B Summary description of all available datafiles, and simulated parameters
SP1
isolines M, ρ=f(x,y)
SP 2

survey of relevant parameters pressure distribution p/po=f(x/b)

CFD 1
cfd11.jpg
cfd12.dat
CFD 2
cfd21.jpg
cfd22.dat



References

[1] Feistauer M., Felcman J. Theory and applications of numerical schemes for nonlinear convection-diffusion problems and compressible Navier-Stokes equations, Mathematics of Finite Elements and Applications, 175-194, Wiley, Chichester, 1997

[2] Fialová M., Hyhlík T., Kozel K., Šafařík P. Numerical analysis data on the transonic flow past the profile cascade SE 1050, Proc. of the Seminar “Topical Problems of Fluid Mechanics”, Institute of Thermomechanics, Prague, 45-48, 2001

[3] Fořt J., Fürst J., Halama J., Kozel K.: Numerical Solution of 2D and 3D Transonic Flows through a Cascade, Proc. of 4th International Symposium on Experimental and Computational Aerothermodynamics of Internal Flows (ed. R. Grundmann), Dresden, Vol.1., 231-240, 1999

[4] Fořt J., Huněk M., Kozel K., Lain J., Šejna M., Vavřincová M. Numerical simulation of steady and unsteady flow through plane cascades, Lecture Notes on Physics (eds. S.M. Desphande, S.S. Desai, R. Narasimha), Vol.453, 461-465, Springer, Berlin, 1995

[5] Matas R., Jůza Z., Riffault T. Some experiences with 2D numerical modelling of transonic profile cascade SE 1050, Proc. of the Colloquium “Fluid Dynamics”, Institute of Thermomechanics, Prague, 85-88, 2000


© copyright ERCOFTAC 2004



Contributors: Jaromir Prihoda; Karel Kozel - Czech Academy of Sciences


Front Page

Description

Test Data

CFD Simulations

Evaluation

Best Practice Advice