Pointwise-FLUENT: Negative Cell Volume

dvzd72 · January 24, 2015, 13:25

Hi all,

Currently I have a 737-600 wing meshed in pointwise.

-Surface mesh in an unstructured T-REX mesh.
-Farfield is hemispherical created using 'shape wizard' script.
-Volume is meshed using T-REX block.

As far as I can see the mesh looks reasonable... for a first attempt at pointwise meshing...

HOWEVER

When it is imported into FLUENT it gives me the error of negative cell volume exists. Which is weird because when I examine the mesh in pointwise it says that all my cells are positive.

Hmmm

Thanks in advance

Matt

dvzd72 · January 24, 2015, 13:28

I did try to upload the mesh but it wouldn't let me...

jchawner · January 25, 2015, 14:04

The first thing to check is whether the face normal vectors point into the block.

The other thing to check is whether you are losing data precision through the Fluent export and you have cells that are small but positive in PW but via export you lose enough digits that they appear zero/negative to Fluent.

Hope this helps.

dvzd72 · January 25, 2015, 14:07

Thanks for a speedy reply btw.

I've checked the normal vectors and they all SEEM to be pointing inwards towards the volume... obviously there is a heck of a lot of them so I can't be 100% certain. Is there a way to check this?

Not really sure how I could ensure that PW exports accurately.

Matt

jchawner · January 25, 2015, 14:41

Checking the normal vectors only has to be done on a face by face basis. If your block simply has one (outer) face that's all you need to check. Any inner faces (around voids, for example) would need to be eyeballed too.

You might want to use the Examine command to see whether any of your cell volumes are "really" small (i.e. close to zero).

dvzd72 · January 25, 2015, 16:33

So it turns out some are very small... I'll have a go at making them bigger and re-export the mesh.

*fingers crossed*

pdp.aero · January 27, 2015, 03:42

Hi

I just wanted to make a comment on this. It will work precisely for structured grid, and I suggest you to try this for diagnosing your T-Rex grid. Use handedness instead of volume criteria for detecting the damaged region. To this end, simply after importing your mesh into Fluent, regardless of its version (6.3.26 or higher), and getting a grid check failure, just initialize the solver by going to solve, initialize, initialize...
Then, check the handedness contour to find out where is affected in mesh by very sharp shape of the geometry. Then, modify that region of your mesh in Pointwise, import it again, and run a grid check to know whether you did manage to reduce left-handed cell, negative volume cell, and handedness faces or not. Meanwhile, you shouldn't have face handedness problem when modifying your grid. Continue doing this until passing a grid check.

Additionally, If you are creating a multi-block grid, you don't need to export all the blocks, just export ones you have problem with. Final point, don't forget to use double precision solver for the Fluent, and set the orient of your mesh properly.

Good Luck,
Payam

Marium · April 23, 2020, 13:09

Hello
I am facing the similar problem, I have created volume mesh in pointwise and then imported it in starccm+. You can see the examine results hereorthogonality.PNG Capture.PNG
How I can solve it? It would be very helpful if you give some insightsreceived_590728918202799.jpeg

Marium · April 23, 2020, 13:41

hey,
Have you solved the problem?

dgarlisch · April 24, 2020, 10:32

Try Examine, Component Volume. That metric breaks cells down into component tets for volume calculations. This can sometimes catch bad cell issues.

The min cell size is VERY small. I don't know the extents of your grid or which File, Properties..., Model Size you are using, but I suspect the dynamic range of this grid is too large for the model size.

Also, be sure to use double precision for CAE export.

Can you ZIP the pw file and make it available online? If it is small enough you can attach here (I doubt it will be).

dgarlisch · April 24, 2020, 10:35

I just realized that Examine, Volume IS reporting a negative minimum cell volume! Most likely, you will need to zoom to that cell (press magnifying glass in examine panel) and fix the surface mesh in that area.

Marium · April 24, 2020, 12:56

Hello,
I can not resolve this problem, the imported mesh from pointwise to starccm+ runs for 40 mins and then show this problem1.jpg 2.PNG
I have examined the pw file and yes it has a few neg cells. But how can I solve them? And how to upload the pointwise file to you?
from the error message, do you think it is because of something else?

dgarlisch · April 24, 2020, 14:04

Contact Pointwise Support for information on uploading files.

Marium · May 9, 2020, 10:39

https://drive.google.com/file/d/1b1c...ew?usp=sharing
here;s the mesh, it's have a few neg volume cell. Can you tell me how I can improve them?

dgarlisch · May 11, 2020, 11:02

I took a quick look at the mesh. I found several issues.

cell volume ratios > 60
max included angles of 180
The Symmetry BCs looked wrong to me

This is far too complicated for me to address over this forum. Someone with more structured meshing experience than me will need to pick this up.

OR

View some relevant videos on the Pointwise YouTube channel.

OR

Take advantage of the Pointwise online training course which is FREE until May 31, 2020.

OR

Call Pointwise Support and they can help you.

Good Luck!

Marium · May 14, 2020, 06:51

Hello
I tried to create a block by extrusion in pointwise.
as you can see, these two portions is always extruded crooked like this.
Is there anyway I can smooth them? or without crooked ones?
prob.PNG

January 24, 2015, 13:25	Pointwise-FLUENT: Negative Cell Volume	#1
dvzd72 New Member Matthew Join Date: Jan 2015 Posts: 4 Rep Power: 11	Hi all, Currently I have a 737-600 wing meshed in pointwise. -Surface mesh in an unstructured T-REX mesh. -Farfield is hemispherical created using 'shape wizard' script. -Volume is meshed using T-REX block. As far as I can see the mesh looks reasonable... for a first attempt at pointwise meshing... HOWEVER When it is imported into FLUENT it gives me the error of negative cell volume exists. Which is weird because when I examine the mesh in pointwise it says that all my cells are positive. Hmmm Thanks in advance Matt

January 25, 2015, 14:04		#3
jchawner Senior Member John Chawner Join Date: Mar 2009 Location: Fort Worth, Texas, USA Posts: 275 Rep Power: 18	The first thing to check is whether the face normal vectors point into the block. The other thing to check is whether you are losing data precision through the Fluent export and you have cells that are small but positive in PW but via export you lose enough digits that they appear zero/negative to Fluent. Hope this helps. __________________ John Chawner / jrc@pointwise.com / www.pointwise.com Blog: http://blog.pointwise.com/ on Twitter: @jchawner

January 25, 2015, 14:41		#5
jchawner Senior Member John Chawner Join Date: Mar 2009 Location: Fort Worth, Texas, USA Posts: 275 Rep Power: 18	Checking the normal vectors only has to be done on a face by face basis. If your block simply has one (outer) face that's all you need to check. Any inner faces (around voids, for example) would need to be eyeballed too. You might want to use the Examine command to see whether any of your cell volumes are "really" small (i.e. close to zero). __________________ John Chawner / jrc@pointwise.com / www.pointwise.com Blog: http://blog.pointwise.com/ on Twitter: @jchawner

April 23, 2020, 13:09	non positive volume	#8
Marium Member Marium Mou Join Date: Mar 2020 Posts: 30 Rep Power: 6	Hello I am facing the similar problem, I have created volume mesh in pointwise and then imported it in starccm+. You can see the examine results hereorthogonality.PNG Capture.PNG How I can solve it? It would be very helpful if you give some insightsreceived_590728918202799.jpeg

May 11, 2020, 11:02		#15
dgarlisch Senior Member David Garlisch Join Date: Jan 2013 Location: Fidelity Pointwise, Cadence Design Systems (Fort Worth, Texas Office) Posts: 307 Rep Power: 14	I took a quick look at the mesh. I found several issues. cell volume ratios > 60 max included angles of 180 The Symmetry BCs looked wrong to me This is far too complicated for me to address over this forum. Someone with more structured meshing experience than me will need to pick this up. OR View some relevant videos on the Pointwise YouTube channel. OR Take advantage of the Pointwise online training course which is FREE until May 31, 2020. OR Call Pointwise Support and they can help you. Good Luck!

January 24, 2015, 13:28		#2
dvzd72 New Member Matthew Join Date: Jan 2015 Posts: 4 Rep Power: 11	I did try to upload the mesh but it wouldn't let me...

January 25, 2015, 14:07		#4
dvzd72 New Member Matthew Join Date: Jan 2015 Posts: 4 Rep Power: 11	Thanks for a speedy reply btw. I've checked the normal vectors and they all SEEM to be pointing inwards towards the volume... obviously there is a heck of a lot of them so I can't be 100% certain. Is there a way to check this? Not really sure how I could ensure that PW exports accurately. Matt

January 25, 2015, 16:33		#6
dvzd72 New Member Matthew Join Date: Jan 2015 Posts: 4 Rep Power: 11	So it turns out some are very small... I'll have a go at making them bigger and re-export the mesh. fingers crossed

January 27, 2015, 03:42		#7
pdp.aero Senior Member Pay D. Join Date: Aug 2011 Posts: 166 Blog Entries: 1 Rep Power: 14	Hi I just wanted to make a comment on this. It will work precisely for structured grid, and I suggest you to try this for diagnosing your T-Rex grid. Use handedness instead of volume criteria for detecting the damaged region. To this end, simply after importing your mesh into Fluent, regardless of its version (6.3.26 or higher), and getting a grid check failure, just initialize the solver by going to solve, initialize, initialize... Then, check the handedness contour to find out where is affected in mesh by very sharp shape of the geometry. Then, modify that region of your mesh in Pointwise, import it again, and run a grid check to know whether you did manage to reduce left-handed cell, negative volume cell, and handedness faces or not. Meanwhile, you shouldn't have face handedness problem when modifying your grid. Continue doing this until passing a grid check. Additionally, If you are creating a multi-block grid, you don't need to export all the blocks, just export ones you have problem with. Final point, don't forget to use double precision solver for the Fluent, and set the orient of your mesh properly. Good Luck, Payam

April 23, 2020, 13:41		#9
Marium Member Marium Mou Join Date: Mar 2020 Posts: 30 Rep Power: 6	hey, Have you solved the problem?

April 24, 2020, 10:32		#10
dgarlisch Senior Member David Garlisch Join Date: Jan 2013 Location: Fidelity Pointwise, Cadence Design Systems (Fort Worth, Texas Office) Posts: 307 Rep Power: 14	Try Examine, Component Volume. That metric breaks cells down into component tets for volume calculations. This can sometimes catch bad cell issues. The min cell size is VERY small. I don't know the extents of your grid or which File, Properties..., Model Size you are using, but I suspect the dynamic range of this grid is too large for the model size. Also, be sure to use double precision for CAE export. Can you ZIP the pw file and make it available online? If it is small enough you can attach here (I doubt it will be).

April 24, 2020, 10:35		#11
dgarlisch Senior Member David Garlisch Join Date: Jan 2013 Location: Fidelity Pointwise, Cadence Design Systems (Fort Worth, Texas Office) Posts: 307 Rep Power: 14	I just realized that Examine, Volume IS reporting a negative minimum cell volume! Most likely, you will need to zoom to that cell (press magnifying glass in examine panel) and fix the surface mesh in that area.

April 24, 2020, 12:56		#12
Marium Member Marium Mou Join Date: Mar 2020 Posts: 30 Rep Power: 6	Hello, I can not resolve this problem, the imported mesh from pointwise to starccm+ runs for 40 mins and then show this problem1.jpg 2.PNG I have examined the pw file and yes it has a few neg cells. But how can I solve them? And how to upload the pointwise file to you? from the error message, do you think it is because of something else?

April 24, 2020, 14:04		#13
dgarlisch Senior Member David Garlisch Join Date: Jan 2013 Location: Fidelity Pointwise, Cadence Design Systems (Fort Worth, Texas Office) Posts: 307 Rep Power: 14	Contact Pointwise Support for information on uploading files.

May 9, 2020, 10:39		#14
Marium Member Marium Mou Join Date: Mar 2020 Posts: 30 Rep Power: 6	https://drive.google.com/file/d/1b1c...ew?usp=sharing here;s the mesh, it's have a few neg volume cell. Can you tell me how I can improve them?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Warning: negative cell volume detected!	boboroo	FLUENT	7	February 11, 2018 03:38
Negative cell volume.	techctrl	FLUENT	0	September 25, 2014 08:22
Problem of simulating of small droplet with radius of 2mm	liguifan	OpenFOAM Running, Solving & CFD	5	June 3, 2014 02:53
[blockMesh] non-orthogonal faces and incorrect orientation?	nennbs	OpenFOAM Meshing & Mesh Conversion	7	April 17, 2013 05:42
negative cell volume during dynamic meshing	Tamjid	FLUENT	1	September 5, 2011 02:33

May 14, 2020, 06:51		#16
Marium Member Marium Mou Join Date: Mar 2020 Posts: 30 Rep Power: 6	Hello I tried to create a block by extrusion in pointwise. as you can see, these two portions is always extruded crooked like this. Is there anyway I can smooth them? or without crooked ones? prob.PNG