[amcook3]

I am working on a 2-D axisymmetric simulation where I have meshed using interior partition lines to make every domain structured. During the boundary condition phase, I continue to have the same error message regarding only a few of the boundaries.

Error: Face processing failed
*Inconsistent domain normal
*Incompatible BC types used on opposite sides of a baffle or connection
*Non-manifold connectivity

I have checked the domain normal(All positive K) as well as the non-manifold connectivity(I think) which leaves the BC types used on opposite sides as the problem. It is only a few random boundaries that cause this issue after some painful debugging(less than 10 out of the 150 boundaries).

Some times when I leave them set as unspecified it imports, sometimes it doesn't. After struggling for over a month trying to search and figure out the issue, I need some help.

If anyone has and incite to try and help fix the issue, it would be greatly appreciated.

Again, this is a 2-d axisymmetric problem with boundary conditions causing the error when exporting to CAE.

I have attached the overview of the drawing as well as an example of two boundaries that are causing the issue.

[jchawner]

Aaron, because you say you've struggled for over a month I suggest you email our Technical Support at support@pointwise.com. Provide your Customer ID and description of your problem.

[amcook3]

Originally Posted by jchawner

Aaron, because you say you've struggled for over a month I suggest you email our Technical Support at support@pointwise.com. Provide your Customer ID and description of your problem.

I tried that in the beginning. I’m currently working on my masters thesis, and the only student to use Pointwise. The university won’t update the maintance package to make it current. It expired April 2017. Therefore, tech support wouldn’t help.

I feel like it’s something simple I’m overlooking since it’s only 2D and only a few BCs that cause the export error, but I’m still struggling with it.

I appreciate the thought.

[jchawner]

That's unfortunate. I don't know what version of the software you're using but it may be old enough that there have been changes to the Fluent export.

I have a vague recollection that Fluent didn't like axisymmetric geometry that had a true singularity along the axis of rotation (i.e. it required a small "tube" down the center). Is that still the case?

I'm not a Fluent user but "non-manifold connectivity" means it can't figure out how to reassemble the mesh into a consistent whole.

Have you set this up for 2D Fluent or 3D?

[amcook3]

Originally Posted by jchawner

That's unfortunate. I don't know what version of the software you're using but it may be old enough that there have been changes to the Fluent export.

I have a vague recollection that Fluent didn't like axisymmetric geometry that had a true singularity along the axis of rotation (i.e. it required a small "tube" down the center). Is that still the case?

I'm not a Fluent user but "non-manifold connectivity" means it can't figure out how to reassemble the mesh into a consistent whole.

Have you set this up for 2D Fluent or 3D?

In fluent It’s 2D Axisymmetric. I’ve ran some simpler geometries without issues so I’m not 100% sure on if it’s a fluent issue or not. The pointwise version is 18.2. I’m almost positive. But I know for sure it’s V18. When I check the manifold in pointwise it gives the blue connection color and seems to not have any issues but I’m not an expert. It’s been more of learning on the fly as I go.

As far as fluent, to my knowledge it just requires the x axis as the line of rotation.

I did get one export to not fail, but only if I left all BC as unspecified. It auto sets as wall in fluent, but They can be changed after. The only down side is you can’t set interior conditions once in fluent, so all my partition lines will have wall conditions instead of through/interior. Not 100 percent on how that will skew the results given both sides of the partition use the same material in fluent.

In fluent, I can set the thickness to zero, but there might have some influence from contact resistance. This week I’m going to build a cas from the file that did export with all unspecified and see how fluent handles that.

[jchawner]

Wait. The messages you quoted are from Pointwise during export? Or from Fluent during Import?

[dgarlisch]

Please attach a zip of the pw file.
List the names of the connectors giving you the problem.

In general, if all normals are consistent, the remaining issue will be incompatible BC types on opposite sides of internal, inter-domain cons. You cannot have a permeable BC type on one side, and a non-permeable type on the other. Permeable types are fan, radiator, porous IIRC.

Another test to run is to remove all BCs and see if export succeeds.

[amcook3]

Originally Posted by jchawner

Wait. The messages you quoted are from Pointwise during export? Or from Fluent during Import?

Pointwise error message

[amcook3]

Originally Posted by dgarlisch

Please attach a zip of the pw file.
List the names of the connectors giving you the problem.

In general, if all normals are consistent, the remaining issue will be incompatible BC types on opposite sides of internal, inter-domain cons. You cannot have a permeable BC type on one side, and a non-permeable type on the other. Permeable types are fan, radiator, porous IIRC.

Another test to run is to remove all BCs and see if export succeeds.

I will upload Monday when I’m back at my desk.

If I remove the BC and switch to unspecified on one of my grids, it created the CAE. I’m only trying to use interior,wall,interface conditions. Nothing fancy. It’s a transient heat transfer problem with natural convection. During the debug stage I recreated the drawing multiple times by drawing it directly in pointwise making sure to have connections and grid points on the ends of connectors. Then moving from one section at a time assign the BCs. Out of the 150BCs the connectors only 6-10 BCs (3-5 connectors) cause the error.

I will upload first thing Monday, and I appreciate the knowledge.

[amcook3]

Originally Posted by dgarlisch

Please attach a zip of the pw file.
List the names of the connectors giving you the problem.

In general, if all normals are consistent, the remaining issue will be incompatible BC types on opposite sides of internal, inter-domain cons. You cannot have a permeable BC type on one side, and a non-permeable type on the other. Permeable types are fan, radiator, porous IIRC.

Another test to run is to remove all BCs and see if export succeeds.

Attached is the .pw file via dropbox link
https://www.dropbox.com/sh/uii6ewdlt...BzdgEjpna?dl=0
I have set the BCs which are interior already and cause no issues
The BCs that are set as unspecified can be changed inside of fluent, so those are not an issue either.
The 6 connector BCs are the issue at this point

Con-1 same/opposite (Vial tip/vial air gap)
Con-17 same/opposite (Vial body base/vial air gap)
Con-99 same/opposite (Vial gap right/base upper pdms)

I appreciate you taking a look at this for me. I feel as it is an easy fix, but I'm stuck.

[dgarlisch]

I downloaded your file.
I loaded it into the current release of Pointwise V18.1 (2018-03-20 20:27:00) Windows 64-bit

Without changing anything, the file as you downloaded exported without error.

Then, I set the BCs to the three connection cons you listed:

• Con-1 same/opposite (Vial tip/vial air gap)
• Con-17 same/opposite (Vial body base/vial air gap)
• Con-99 same/opposite (Vial gap right/base upper pdms)

This also exported without error.

I noticed the pw file header indicated you are using Pointwise V18.0R3 Mon May 21 14:07:17 (GMT) 2018

I was able to repeat the error!

Error: (May 21 12:03:07 pm) Face processing failed.
Info: (May 21 12:03:07 pm) Check your grid for one of the following invalid situations:
Info: (May 21 12:03:07 pm) * Inconsistent domain normals.
Info: (May 21 12:03:07 pm) * Incompatible BC types used on opposite sides of a baffle or connection.
Info: (May 21 12:03:07 pm) * Non-manifold connectivity.
Warning: (May 21 12:03:07 pm) Aborted Export CAE.

After further investigation it appears that this is a bug (PW-20935 Fluent Export Fails) that is already fixed in Pointwise v18.1R1.
You will need to update your version.

[amcook3]

Originally Posted by dgarlisch

I downloaded your file.
I loaded it into the current release of Pointwise V18.1 (2018-03-20 20:27:00) Windows 64-bit

Without changing anything, the file as you downloaded exported without error.

Then, I set the BCs to the three connection cons you listed:

• Con-1 same/opposite (Vial tip/vial air gap)
• Con-17 same/opposite (Vial body base/vial air gap)
• Con-99 same/opposite (Vial gap right/base upper pdms)

This also exported without error.

I noticed the pw file header indicated you are using Pointwise V18.0R3 Mon May 21 14:07:17 (GMT) 2018

I was able to repeat the error!



After further investigation it appears that this is a bug (PW-20935 Fluent Export Fails) that is already fixed in Pointwise v18.1R1.
You will need to update your version.

I appreciate you taking time away to assist in this matter. My university hasn't updated the maintenance package, so I'm hoping that I can still install the newest version. I have reached out to our technical support team to have this version installed (if possible).

The exact version installed currently is...Pointwise V18.0 Release 3 -- Optimized Build 0209161753 (2017-05-16 17:53:00) Windows 64-bit

Again, I appreciate the prompt replay and knowledge.

[shubz]

Originally Posted by amcook3

I am working on a 2-D axisymmetric simulation where I have meshed using interior partition lines to make every domain structured. During the boundary condition phase, I continue to have the same error message regarding only a few of the boundaries.

Error: Face processing failed
*Inconsistent domain normal
*Incompatible BC types used on opposite sides of a baffle or connection
*Non-manifold connectivity

I have checked the domain normal(All positive K) as well as the non-manifold connectivity(I think) which leaves the BC types used on opposite sides as the problem. It is only a few random boundaries that cause this issue after some painful debugging(less than 10 out of the 150 boundaries).

Some times when I leave them set as unspecified it imports, sometimes it doesn't. After struggling for over a month trying to search and figure out the issue, I need some help.

If anyone has and incite to try and help fix the issue, it would be greatly appreciated.

Again, this is a 2-d axisymmetric problem with boundary conditions causing the error when exporting to CAE.

I have attached the overview of the drawing as well as an example of two boundaries that are causing the issue.

Hi , I had a similar problem when I was trying to model a thin/ baffle surface in Pointwise

I had mixed grids on my baffle surface and when the volume mesher was done , it would locally modify the grids on the baffle surface. When I brought in the .vml file that pointwise suggests , it showed me which cells were problematic . From the comments in this and other posts I figured that the problem might be related to mixed grids near that part of the geo. So I locally meshed the baffle in the problematic area only with tri cells and it worked.

Hopefully this helps someone struggling with the same issue