Results for case (not set)

Code: Code_Saturne

Version: 2.0-beta2

Authors: E. Moreau

Method and Numerical Options

It should be noted that all but one calculations have been run with the thermal time-step limitation (IPTLRO=1). In addition, here are the relevant parameters:

Case Download ID Parameters
K-EPSILON download 09141027 IDEUCH=1
IDTVAR=1
    09111415 IDEUCH=2
IDTVAR=1
    09141727 IDEUCH=1
IDTVAR=2
K-OMEGA download 09141121 IDEUCH=2
IDTVAR=1
    09141426 IDEUCH=1
IDTVAR=1
V2F download 09141730 IDEUCH=1
IDTVAR=1
    09141737 IDEUCH=1
IDTVAR=-1
    09161549 IDEUCH=1
IDTVAR=1
no interface

Models

%$k-\varepsilon$%, %$k-\omega$% SST and %$\overline{v^2}-\overline{f}$%

Mesh

Description of the results files

The results file contains all the xmgrace files and png pictures as well as data for inter model/option comparison with V2. Each case results directory also contains temperature and velocity profiles.

Boundary conditions

Colour Boundary
3 Cold wall, %$T=15.6$%
1 Hot wall, %$T=54.7$%
2 or 4 Adiabatic wall, %$\frac{\partial T}{\partial n}=0$%
5 Symmetry

Reference Publications

Discussion

Simulations consisted in comparing three kinds of turbulent models and the effects of the wall function parameter (ideuch) in the models.

Seven simulations have thus been performed in this study : three calculations with the %$k-\varepsilon$% model ( IDEUCH = 1 or 2), with a time step varying with time ( IDTVAR = 1), and with a time and spatial dependent time step ( IDTVAR = 2). Two calculations with the %$k-\omega$% SST model ( IDEUCH = 1 or 2), with IDTVAR = 1. And two calculations with the %$\varphi-\overline{f}$% model, the first one with IDTVAR = 1, and the second one with IDTVAR = -1 (stationary).

The first main result is that the simulations performed with the version 2.0 give very similar results with the previous validated version 1.3. This fact is verified for the three models used, and the conclusions obtained with the version 1.3 could be rewritten here for the version 2.0.

The results are presented in the following figures (1) to (15).

All figures except figure (5) give the temperature and velocity profiles at z = 0 and at several levels y of the domain.

The overall results show that the %$\varphi-\overline{f}$% model and the %$k-\omega$% SST model give results very close to the experience. More precisely:

  • For the %$k-\varepsilon$% model:
    • The temperature gives better results with IDEUCH = 1 while the results are better for the velocity with IDEUCH = 2.
    • The parameter IDTVAR set to 2 (time and space dependent time step) improves these results (temperature and velocity). The case has been used with the time step limitation linked to the local thermal time step ( IPTLRO option).
    • The results are very close to those obtained with the version 1.3 (figures (6) to (9)).
    • Globally this model does not seem to be well adapted to this configuration (as for the version 1.3).

  • For the %$k-\omega$% SST model:
    • The case IDEUCH = 1 gives results very close to experimental data (better than IDEUCH = 2) for the temperatures and velocities.
    • Results obtained with the present and the previous versions can be superposed (figures (10) to (13)).

  • For the %$\varphi-\overline{f}$% model:
    • This model is the closest to the experimental data as for the version 1.3.
    • The stationary case can be strictly superposed to the non stationary case.

Figure (5) shows vertical profiles of velocity and temperature starting from the middle of the upper and lower wall (at z = 0 and x = 0.038m) and extending on 0.04m towards the cavity center. We find again a good agreement between %$k-\omega$% SST or %$\varphi-\overline{f}$% models and experience for the velocity. On the other hand, none of the models used give correct profile near the wall since we have considered completely adiabatic walls. Version 1.3 yields the same conclusion.

It should be pointed out that a simulation without using the xml file have been performed (we have used only the fortrans files) for the %$\varphi-\overline{v^2}$% model ( IDEUCH =1), the same results are obtained whether the intereface is used or not (figures (14) and (15)).

Results with version 2.0. Inter model / options comparison

T horizontal1.png
Figure1: Horizontal profiles of temperature - Lower part of the cavity
T horizontal2.png
Figure2: Horizontal profiles of temperature - Upper part of the cavity
V horizontal1.png
Figure3: Horizontal profiles of velocity- Lower part of the cavity
V horizontal2.png
Figure4: Horizontal profiles of velocity - Upper part of the cavity
Tu verticalexp.png
Figure5: Vertical profiles of temperature and velocity

Comparison between version 1.3 and version 2

compar TH1 KEPS .png
Figure6: Horizontal profiles of temperature - Lower part of the cavity - k-epsilon model - V1.3 / V2 comparison
compar TH2 KEPS.png
Figure7: Horizontal profiles of temperature - Upper part of the cavity - k-epsilon model - V1.3 / V2 comparison
compar VH1 KEPS.png
Figure8: Horizontal profiles of velocity - Lower part of the cavity - k-epsilon model - V1.3 / V2 comparison
compar VH2 KEPS.png
Figure9: Horizontal profiles of velocity - Upper part of the cavity - k-epsilon model - V1.3 / V2 comparison

compar TH1 KOMEGA.png
Figure10: Horizontal profiles of temperature - Lower part of the cavity - k omega SST model - V1.3 / V2 comparison
compar TH2 KOMEGA.png
Figure11: Horizontal profiles of temperature - Upper part of the cavity - k omega SST model - V1.3 / V2 comparison
compar VH1 KOMEGA.png
Figure12: Horizontal profiles of velocity - Lower part of the cavity - k omega SST model - V1.3 / V2 comparison
compar VH2 KOMEGA.png
Figure13: Horizontal profiles of velocity - Upper part of the cavity - k omega SST model - V1.3 / V2 comparison

Influence of using the interface on the results

compar vf2 T.png
Figure14: Horizontal profiles of temperature - Lower part of the cavity - v2f model - GUI ON / OFF comparison
compar vf2 V.png
Figure15: Horizontal profiles of temperature - Lower part of the cavity - v2f model - GUI ON / OFF comparison



Current Tags:
Saturne_Validation_2.01Add my vote for this tag create new tag
, view all tags
Topic revision: r35 - 2014-10-27 - 11:48:22 - AlastairWest
 

Computational Fluid Dynamics and Turbulence Mechanics
@ the University of Manchester
Copyright & by the contributing authors. Unless noted otherwise, all material on this web site is the property of the contributing authors.