|
[Sponsors] |
March 1, 2010, 01:26 |
low pressure at front of buildings
|
#1 |
New Member
CFT
Join Date: Mar 2009
Posts: 15
Rep Power: 17 |
Dear all
I am simulating flow over a small low rise building. When I look at the dynamic pressure at the front of the building center, i get very low velocity and dynamic pressure. However this is contrary to experimental reults. Any suggestions as to why ? I am using a k epsilon rng with wall functions, hexahedral grid, double digit precision. Any help is much appreciated regards CFT |
|
March 1, 2010, 11:55 |
Convergence criteria, wall fuction vs enhanced Wall treatment
|
#2 |
New Member
Join Date: Nov 2009
Posts: 22
Rep Power: 16 |
Hi,
two recommendations: 1) maybe this is a silly question but, are you sure your simulation is converged? I know you might be expert and know it but someone else might find this useful. The default convergence criteria set by Fluent are not sufficient and maybe misleading. I recommend to get rid of the residual checking: solve-> monitors->residuals->convergence criterion: NONE. Create a couple of points in your domain where you monitor a significant quantity, this could be pressure or velocity component in the prevailing direction of your flow. I would consider the simulation converged when the residuals are 10^-4 AND the monitored pressure or velocity components are independent of the iteration. 2) If you are using k-epsilon with STD wall function check the y+ on the walls: 30 <y+<300. If you are using k-epsilon with Enhanced Wall treatment y+<5 and ideally y+~1. Good luck |
|
March 5, 2010, 04:37 |
low pressure at front of building
|
#3 |
New Member
CFT
Join Date: Mar 2009
Posts: 15
Rep Power: 17 |
Hi
I checked my residuals and are all below 1e-6. Also the yplus is in the range of 0<yplus<300 for wall functions. I compare the pressure coefficent and compare with experiemntal results and get less than 0.3. Which cannot be right ? Is it something to do with my reference pressure location ? I place it at 0,0,0 as recommended in manual. Any other suggestions thanks CFT |
|
March 7, 2010, 17:25 |
|
#4 |
New Member
Join Date: Nov 2009
Posts: 22
Rep Power: 16 |
Hi Have you done this?
1) Create a couple of points in your domain where you monitor a significant quantity, this could be pressure or velocity component in the prevailing direction of your flow. I would consider the simulation converged when the residuals are 10^-4 AND the monitored pressure or velocity components are independent of the iteration. 2) this is the correct range of y+ for wall functions: 30<y+<300 instead of 0 <y+<300. |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Pressure BC for combustion chamber | Giuki | FLUENT | 1 | July 19, 2011 11:35 |
measured pressure in the discharge pipe jumps | Jan | Main CFD Forum | 1 | October 30, 2006 07:54 |
Pressure Inlet yields wrong velocities | Ben | FLUENT | 0 | November 21, 2004 01:47 |
Multicomponent fluid | Andrea | CFX | 2 | October 11, 2004 05:12 |
Front tracking | Don Hawken | Main CFD Forum | 1 | December 12, 1999 13:29 |