|
[Sponsors] |
[ANSYS Meshing] Streamlined vehicle boundary layer problem |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
November 3, 2014, 14:59 |
Streamlined vehicle boundary layer problem
|
#1 |
New Member
Paul Mannion
Join Date: Nov 2014
Posts: 11
Rep Power: 12 |
Hi guys,
I've experienced problem while trying to create a mesh that captures the boundary layer effect on a streamlined vehicle shape. The vehicle is shaped like a simple teardrop with a flat bottom. Only half the vehicle is modelled using symmetry. The problem that occurs is that the front half of the vehicle meshes correctly however as the mesh progresses to the rear of the vehicle, it stops meshing the boundary layer in the same manner and essentially thins out replacing the desired mesh with large tet elements... (An image of the mesh is attached, I created a section cut through the middle to show how the mesh wraps around the shape at the front of the vehicle, it does not do this at the rear of the vehicle however) I've also tried meshing the complete vehicle body shape without using symmetry but this did not yield any results. Any help or advice would be greatly appreciated. I've tried every method of meshing with inflation within Ansys Meshing to find a solution but each yields the same result as above. Regards P |
|
November 4, 2014, 03:52 |
|
#2 |
Senior Member
|
Are you sure that all the surfaces on the car has inflation on them? Looks like they are deminishing to zero layers because the rear is not inflated (?) at all.
|
|
November 4, 2014, 12:52 |
|
#3 |
New Member
Paul Mannion
Join Date: Nov 2014
Posts: 11
Rep Power: 12 |
Hi jrunsten
I have inflation selected on all surfaces of the vehicle body, It breaks down into 3 surface components. The bottom, the side, and there is a minuscule area at the rear edge. I've tried knitting all these surfaces into one surface and applying inflation, and applying inflation to each of these surfaces seperatly. I've also tried eliminating the rear surface so that the only 2 surfaces were the bottom and the side. None of these yielded a different result. If I examine the rear of the vehicle in close detail I find that the program is wrapping the rear of the shape 1 or 2 elements thick, if i select a high number of layers in the inflation menu (+40) then this layer reduces to 1 element thick at the rear, if i select a low number of layers (eg:5)in the inflation menu then the layer at the rear of the vehicle can increase to 3 elements thick I attached 2 images to try and illustrate what is happening at the rear region of the vehicle I apologise if my explanation above is not extremely helpful or intuitive |
|
November 4, 2014, 13:06 |
|
#4 |
Senior Member
|
Under the "Details" of "Mesh", enable "View advanced options" under "Inflation" and try switching collision avoidance from stairstepping to compression instead. That might work in your case.
|
|
November 4, 2014, 14:47 |
|
#5 |
New Member
Paul Mannion
Join Date: Nov 2014
Posts: 11
Rep Power: 12 |
Thanks for the advice, I gave it a go and here are the results
Unfortunately there was not much difference in the results between 'stair stepping' and 'layer compression. An image of the result of each is attached below However I am glad to say that just from experimenting with other options under the same tab I managed to get a far more acceptable mesh made. The change i made was under 'Inflation growth rate type - maximum angle.'' I changed the angle from 90 deg to 180 deg, images are attached below. This new mesh isn't perfect however as highlighted by the last image zoomed up at the rear end of the vehicle shape. Also I noticed the following error appearing under all mesh generations. ''Pre-inflation layer generation is stopped because of face side limitation, proximity collisions and quality'' |
|
November 4, 2014, 15:18 |
|
#6 |
New Member
Paul Mannion
Join Date: Nov 2014
Posts: 11
Rep Power: 12 |
I've managed to make further improvements by changing the 'maximum height over base' to its lowest value of 0.1
I attached 2 images below, one with 'stair stepping' collision avoidance and one with 'none' selected under collision avoidance, I was hoping you could shed some light on these commands? as I do not have a complete understanding on their effect on the mesh. Both mesh images attached below are similar however one does not have collision avoidance and the other does, so I am unsure as to which one is preferred. |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
High aspect ratio mesh problem for flat plate boundary layer | sam1364 | OpenFOAM | 3 | May 14, 2020 12:11 |
[GAMBIT] Problem with boundary layer - high skewness | Muffins | ANSYS Meshing & Geometry | 4 | February 28, 2014 08:34 |
Low Mixing time Problem | Mavier | CFX | 5 | April 29, 2013 01:00 |
Questions about Boundary Layer Thickness and Turbulence Models | famerfamer | STAR-CCM+ | 3 | July 12, 2012 10:47 |
errors | Fahad | Main CFD Forum | 0 | March 23, 2004 14:20 |