|
[Sponsors] |
November 9, 2017, 09:54 |
Bad mesh on easy geometry, Surface Wrapping
|
#1 |
New Member
Mathias
Join Date: Dec 2016
Posts: 25
Rep Power: 9 |
Hi guys,
i am using Surface Wrapper to get a closed Surface of my body. In some regions its not working right. In the picture you can see an underbody of a car. The little steps are orthogonal to the main plate. The height is 1 mm on the thinnest area. Its easy geometry but still hes messing it up. https://www.imagebanana.com/s/871/ULFDkkEs.html I tried to use a finer setting, but it didnt help it anyway. If i am using curvature refinement still nothing happens. Someone an idea how we could get better results? I guess its something wrong with the feature curvs, because after using surface wrapping function the feature curves are moved out of place like in the picture. Settings: Base Size: 0.02 m Target Surface Size: 0.004 m Minimum Surface Size: 0.001 m Surface curvature: 72 points/circle Surface Refinement for the Undertray: Target surface size: 0.002m Minimum Surface size: 0.0008 m |
|
November 10, 2017, 00:32 |
|
#2 |
Senior Member
Join Date: Nov 2010
Location: USA
Posts: 1,232
Rep Power: 24 |
Curvature refinement will not be effective here because the geometry lacks significant curvature.
This does look like an easy geometry - why are you bothering to surface wrap it? Is it that unfixable in surface repair? The best way to get good results on that curve is to define those features as a new part surface and use contact prevention. |
|
November 10, 2017, 10:27 |
|
#3 |
New Member
Mathias
Join Date: Dec 2016
Posts: 25
Rep Power: 9 |
Thanks for your reply!
Why does the geometry lacks on significant curvature? The confusing thing is, that the curvature which is the red feature edge in the corners gets destroyed after wrapping(red line=feature curves). Yes its an easy geometry, but its just a part of a full car model. I have to surface wrap it because i import seperat parts which are not connected to each other. Yes,defining them as seperat part and use contact prevention would be a possible solution, but if possible i dont want to rework the underbody. The solution i am searching for is using better settings for wrapping, or to hear about tricks to get better result. Thanks for your help! |
|
November 10, 2017, 14:29 |
|
#4 |
Senior Member
Join Date: Nov 2010
Location: USA
Posts: 1,232
Rep Power: 24 |
Feature curves are not curvature. A feature curve can be perfectly straight and thus contain no curvature. Curvature is how non-planar/non-linear a surface or curve is. Airfoils have lots of curvature, flat planes have none. The geometry you have there looks like it's pretty flat, and so curvature refinement is not really effective, as you see in your wrap.
For a very flat surface if you want a good wrap your only tools are to lower the target size or to use contact prevention or use a volume control. I would recommend that you split those surfaces and either use a contact prevention or set a lower target size for those features. You can also split those curves out and name them, including them in a curve control, but IMO dealing with surfaces is easier. |
|
November 12, 2017, 13:18 |
|
#5 |
New Member
Mathias
Join Date: Dec 2016
Posts: 25
Rep Power: 9 |
Hi me3840,
thanks again for your reply and the solution approach! I will try it on the next simulation and will give feedback if it was possible to solve this problem with several tools. So other people might can use this to solve their problems |
|
November 14, 2017, 19:41 |
|
#6 |
New Member
Zachary MacChesney
Join Date: Jan 2017
Posts: 6
Rep Power: 9 |
Hi.. I'd go to Regions>Boundaries>'your surface'>Mesh Conditions>Custom Surface Size
Then Mesh Values > Surface Size > Absolute Minimum Size then reduce that value a little.. The setting looks like this: https://www.imagebanana.com/s/big/877/94HmVFvP.PNG |
|
November 15, 2017, 02:01 |
|
#7 |
New Member
Mathias
Join Date: Dec 2016
Posts: 25
Rep Power: 9 |
Hi macchesney!
you use region based meshing? i am using parts based meshing and there its possible as well to reduce the surface sizes like the minimum surface size. but this was not bringing the wished solution. Thanks for reply! |
|
November 16, 2017, 01:42 |
|
#8 |
New Member
Zachary MacChesney
Join Date: Jan 2017
Posts: 6
Rep Power: 9 |
Well I'm going to give you a lazy answer here.. you need to determine whether that feature is critical to the flow.. I'd be inclined to just solve using the mesh you got, and see how your results look.. if they are bad then you can find out why.. just to get some sense of scale.. how big is that feature in comparison to the overall geometry?
|
|
November 20, 2017, 16:01 |
|
#9 |
New Member
Mathias
Join Date: Dec 2016
Posts: 25
Rep Power: 9 |
Hi!
sorry i was in foreign countries and had no time to answer. I solved the mesh before my flight and evaluated the solution already, because the guys which do the constructions need to go on, because we otherwise lose a lot of time. These curves are starting small in their hight(starting with 1 mm and then growing till i think 20 mm, finally in the diffusor ofc they are much higher). The mesh is in generally good, so its a small area where the geometry is not exactly what it should be. For the question with the size of that curves i have postet the picture below. In the middle between the two green big pressure areas on the right side of it, there are 6 lines which presents the curves. https://www.imagebanana.com/s/882/GGF0s38A.html PS. fyi the start of the curves (you see the coarse mesh) is not meshed right, if you follow the curve, the mesh gets betteras you see on the better mesh in the picture |
|
November 20, 2017, 21:32 |
|
#10 |
New Member
Zachary MacChesney
Join Date: Jan 2017
Posts: 6
Rep Power: 9 |
I'd say given that mesh you could probably ignore it.. since it is on the surface deep within the boundary layer.. It would maybe act as additional surface roughness.. if it really bothers you, you could put a volume of refinement around that area..
|
|
November 21, 2017, 07:50 |
|
#11 |
New Member
Mathias
Join Date: Dec 2016
Posts: 25
Rep Power: 9 |
yeah will maybe try it in the future. Thanks for that discussen, i would say that problem is enough discussed and solved!
|
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
decomposePar problem: Cell 0contains face labels out of range | vaina74 | OpenFOAM Pre-Processing | 37 | July 20, 2020 05:38 |
[snappyHexMesh] Bad meshing result on corrugated metal sheet | UebertreibeR | OpenFOAM Meshing & Mesh Conversion | 1 | August 19, 2016 04:55 |
[snappyHexMesh] SnappyHexMesh for internal Flow | vishwa | OpenFOAM Meshing & Mesh Conversion | 24 | June 27, 2016 08:54 |
Cluster ID's not contiguous in compute-nodes domain. ??? | Shogan | FLUENT | 1 | May 28, 2014 15:03 |
unstructured vs. structured grids | Frank Muldoon | Main CFD Forum | 1 | January 5, 1999 10:09 |