[ANSYS Meshing] can floating point exception correspond to bad mesh quality?

Red Ember · June 30, 2017, 03:45

I ran cfx-solver and get such error:

+--------------------------------------------------------------------+
| Mesh Statistics |
+--------------------------------------------------------------------+
| Domain Name | Orthog. Angle | Exp. Factor | Aspect Ratio |
+----------------------+---------------+--------------+--------------+
| | Minimum [deg] | Maximum | Maximum |
+----------------------+---------------+--------------+--------------+

+--------------------------------------------------------------------+
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| Floating point exception: Invalid number |

So is it something like negative volume anywhere?
It's first time I faced such error.

vasava · July 6, 2017, 03:02

Yes and no. Yes, bad mesh quality can lead to floating point exception error.

No (it is not ONLY the mesh that can cause floating point error), sometimes it can also be because of improper boundary conditions, large time-steps, improper interface setup and so on.

It is advisable not just to check mesh before starting simulation but also fix the mesh. I am not sure about CFX but Fluent has commands like repair-face-handedness and repair-face-node-order that can help fix the small errors in the mesh without having to make the mesh again.

Red Ember · July 10, 2017, 11:07

"improper boundary conditions, large time-steps, improper interface setup" - no. that was not the rub definitely.As you can see, the problem popped out even before mesh statistics estimation not to mention the first iteration. and as I know, there are no useful commands such as repair, alas...

divergence · July 11, 2017, 04:41

It sounds to me that you might have something big wrong in the mesh. Perhaps you should try to go back to your meshing program and try to find the error there. Is it possible to do some kind of visualization of the mesh parameters before the actual solving of the case? You might me able to locate the assumably low quality cells in the mesh that way.

If you are using DesignModeler and Meshing, I would suggest that you should try the "Share Topology" tool in DM, which would usually give the user a heads up if there's something amiss in the actual computational volume ( for instance overlapping geometries). Meshing (the program) itself provides many tools to evaluate the qualitative parameters of the mesh.

June 30, 2017, 03:45	can floating point exception correspond to bad mesh quality?	#1
Red Ember Member Alex Join Date: Feb 2016 Posts: 81 Rep Power: 10	I ran cfx-solver and get such error: +--------------------------------------------------------------------+ \| Mesh Statistics \| +--------------------------------------------------------------------+ \| Domain Name \| Orthog. Angle \| Exp. Factor \| Aspect Ratio \| +----------------------+---------------+--------------+--------------+ \| \| Minimum [deg] \| Maximum \| Maximum \| +----------------------+---------------+--------------+--------------+ +--------------------------------------------------------------------+ \| ERROR #001100279 has occurred in subroutine ErrAction. \| \| Message: \| \| Floating point exception: Invalid number \| So is it something like negative volume anywhere? It's first time I faced such error.

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 06:38
[mesh manipulation] Importing Multiple Meshes	thomasnwalshiii	OpenFOAM Meshing & Mesh Conversion	18	December 19, 2015 19:57
[snappyHexMesh] snappyHexMesh and cyclic boundaries	Ruli	OpenFOAM Meshing & Mesh Conversion	2	December 9, 2013 07:51
simpleFoam Floating point exception error -help	sudhasran	OpenFOAM Running, Solving & CFD	3	March 12, 2012 17:23
[snappyHexMesh] snappyHexMesh won't work - zeros everywhere!	sc298	OpenFOAM Meshing & Mesh Conversion	2	March 27, 2011 22:11

July 6, 2017, 03:02		#2
vasava Senior Member Paritosh Vasava Join Date: Oct 2012 Location: Lappeenranta, Finland Posts: 732 Rep Power: 23	Yes and no. Yes, bad mesh quality can lead to floating point exception error. No (it is not ONLY the mesh that can cause floating point error), sometimes it can also be because of improper boundary conditions, large time-steps, improper interface setup and so on. It is advisable not just to check mesh before starting simulation but also fix the mesh. I am not sure about CFX but Fluent has commands like repair-face-handedness and repair-face-node-order that can help fix the small errors in the mesh without having to make the mesh again.

July 10, 2017, 11:07		#3
Red Ember Member Alex Join Date: Feb 2016 Posts: 81 Rep Power: 10	"improper boundary conditions, large time-steps, improper interface setup" - no. that was not the rub definitely.As you can see, the problem popped out even before mesh statistics estimation not to mention the first iteration. and as I know, there are no useful commands such as repair, alas...

July 11, 2017, 04:41		#4
divergence Member Join Date: Mar 2014 Posts: 56 Rep Power: 12	It sounds to me that you might have something big wrong in the mesh. Perhaps you should try to go back to your meshing program and try to find the error there. Is it possible to do some kind of visualization of the mesh parameters before the actual solving of the case? You might me able to locate the assumably low quality cells in the mesh that way. If you are using DesignModeler and Meshing, I would suggest that you should try the "Share Topology" tool in DM, which would usually give the user a heads up if there's something amiss in the actual computational volume ( for instance overlapping geometries). Meshing (the program) itself provides many tools to evaluate the qualitative parameters of the mesh.