CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (http://www.cfd-online.com/Forums/ansys-meshing/)
-   -   [ANSYS Meshing] Ansys meshing result very high skewness (http://www.cfd-online.com/Forums/ansys-meshing/98428-ansys-meshing-result-very-high-skewness.html)

m5edr March 10, 2012 21:28

Ansys meshing result very high skewness
 
3 Attachment(s)
My Geometry is a wind turbine blade in a huge Air enclosure (1000m depth x 300m height)
I built it in Solidworks then select the enclosures in DesignModeler and started AnsysMeshing,I had tried many mesh types and reduce sizing
BUT the skewness alawys is very high as below:

Element 11088617
skewness Min 2.136 E-10
skewness Max 0.99999498

Any suggestion (I think to try ICEM CFD but iam not professional)

Far March 11, 2012 01:29

Quote:

BUT the skewness alawys is very high as below:
try to spot the max skewness area


Quote:

Any suggestion (I think to try ICEM CFD but iam not professional)
ANSYS meshing is good tool for quick design iteration, However ICEM is the ultimate tool to create high quality meshing and provides high control. It has also advantage of re-usability of blocking (Hexa).

m5edr March 11, 2012 06:10

Quote:

Originally Posted by Far (Post 348741)
try to spot the max skewness area


ANSYS meshing is good tool for quick design iteration, However ICEM is the ultimate tool to create high quality meshing and provides high control. It has also advantage of re-usability of blocking (Hexa).

Thanks Dear Far
1-could you tell me how to spot the max skewness area?

2- When I tried ICEM by import the geometry From DM , the domains (Enclosures) looks different that in AM
I didn't know why , iam still beginner in ICEM

thanks again

PSYMN March 11, 2012 08:44

In ANSYS Meshing, left click on the mesh branch of the tree...

In the resulting details panel, expand the statistics branch... Under Mesh metric, choose "Skewness" from the pulldown.

A histogram will appear below the graphics window.

Left click on the bar of the histogram and its elements will appear.

If the bar you want to see is too small to click easily, click on the "controls" button and adjust the X and Y axis to zoom in on the bad cells.

When you click on a bar, you will see those elements on the screen...

PSYMN March 11, 2012 08:49

Oh yea, what to do with what you find...

Look to see where the skewness is coming from. 9 times out of 10, it is due to over constrained geometry forcing the mesh into awkward configurations...

If the over-constrained geometry is just surface patches (like a sliver surface forcing the mesh to conform) you could try the patch independent tetra or you could try virtual topologies to merge the surface patches together...

If it is between regions, such as between your turbine and the boundary of the rotating region... That second boundary is arbitrary... and you could just increase the size of that disk to allow for more space (flexibility) for the mesh...

If it is something else, post the pic.

Best regards,

Simon

m5edr March 11, 2012 10:43

I reached to bad elements , I will try to repair them
Many thanks

m5edr March 12, 2012 08:52

1 Attachment(s)
Dear simon

I reached to bad elements in My geometry (attached photo)
It is because "Inflation" that used around the blade (this "Inflation" is nessary due to Boundary Layer consideration)

Without Inflation >>> Max Skewness is 0.87 (prefect)
With Inflation>>>> Max Skewness is 0.9999

I tried many types of inflation that result in same High Skewness any change.
Any suggestion,
Thanks

PSYMN March 12, 2012 09:03

You would have to zoom in to get a better idea of what is going on...

My guess is that the inflation at the top of the blade is meeting the inflation from the other side... Since they meet at a trailing edge, you get a sharp angle between them. No amount of smoothing or other "mesh" fixes will solve this problem because when you make one side better, the other side gets worse.

Some people fix this problem by putting a zero thickness baffle behind the trailing edge. This splits the angle (prism grows from both sides) and solves the problem... However, where the thin prism baffle ends, you may have other problems if you can't taper it out. (I know how to taper it out in ICEM CFD, but I don't know how in ANSYS Meshing, maybe someone else knows).

But you could also just try sending it to the solver as it is. The solver can often handle these sorts of poor quality prisms much better then it would handle poor tetras...

m5edr March 12, 2012 11:13

2 Attachment(s)
Many thanks for replay , i am really appreciated that

I attached more closer image , any change in your replay after this images
Note: there is no inflation at the top of the blade (Inflation at both sides of the blade)

thanks again

PSYMN March 12, 2012 11:15

I had expected they were just along the trailing edge...

I guess the wing tip could be a similar problem where the prisms are tilting to miter around the tip...

m5edr March 12, 2012 11:37

Quote:

Originally Posted by PSYMN (Post 348959)
I had expected they were just along the trailing edge...

I guess the wing tip could be a similar problem where the prisms are tilting to miter around the tip...

Sorry
I didn't understand...Is the"wing tip" tutorial or form ??
thanks

mrdelaunay March 16, 2012 13:06

try advanced size function and adjusting your max/min sizes
 
It looks like either your advanced size function is off or your are not setting the correct max/min sizes. Try setting the physics preference to CFD.

PSYMN March 16, 2012 13:17

@m5edr, by "wing tip", I just meant the tip of the airfoil shown in your images... I guess it may not actually be a wing...

m5edr March 16, 2012 22:09

Quote:

Originally Posted by mrdelaunay (Post 349882)
It looks like either your advanced size function is off or your are not setting the correct max/min sizes. Try setting the physics preference to CFD.

On the contrary,
My advance size function is ON and also i tried many size function !!

Any way i started ICEM , May it comes new News

m5edr March 16, 2012 22:13

Quote:

Originally Posted by PSYMN (Post 349885)
@m5edr, by "wing tip", I just meant the tip of the airfoil shown in your images... I guess it may not actually be a wing...

Thanks Simon

I started ICEM , it provide many tools to fix the mesh
Till now I face same problem but i think solution is close and maybe i need your help if i stopped :)

thanks again

Tanjina January 5, 2014 17:11

Very high skewness within two cell zone.
 
1 Attachment(s)
Hi Simon,

I am modeling a 3D object. I found very high skewness in between two cell zone using workbench meshing. Before starting the "run Calculation" , I checked " check case" and it gives me this warning that 696 cell exceed 0.98 skewness. I found a way using Fluent 14.5 to repair the face mesh, but couldn't find any way how can I repair this skewness within the cell zone. Please find the attached image for details.

My model's mesh has high aspect ratio also i.e. 14.3, but Fluent didn't give any warning regarding this.

Any suggestion for lowering the skewness will be highly appreciated. Thanks in advance.

Regards,
Tanjina

sfrankricharrd March 31, 2015 04:11

best way to reduce skewness and its successful, i tried
 
by using inflation> use auto inflation> programmed controlled and i generated my mesh. it shown the skewness value as 0.8999.

then i tried

inflation> use auto inflation> none

then i got the skewness value as 0.822


All times are GMT -4. The time now is 08:22.