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/)
-   -   [ICEM] Negative volume error in hybrid mesh (http://www.cfd-online.com/Forums/ansys-meshing/79798-icem-negative-volume-error-hybrid-mesh.html)

siw September 4, 2010 07:43

[ICEM] Negative volume error in hybrid mesh
 
2 Attachment(s)
Hi, this is a long description but I've tried to explain everything to help get a fix. I've searched for help with this but all I could find was on Hexa meshes. I've also watched the video at http://www.ansys.com/demoroom/swf/de...ion/index.html but they did not talk about the tetras resulting from prism generation.

I've made a tetra/prism mesh in ICEM v12.1 for an aircraft flowfield simulation for solving in CFX v12.1. The mesh was made as follows:

a) Compute Octree mesh, delete volume elements, check mesh (delete unconnected vertices), smooth TRI_3 elements until quality > 0.3.

b) Compute Smooth (Advancing Front) mesh using exisitng surface mesh, check mesh, smooth TRI_3 and TETRA_4 elements until quality > 0.3.

c) Compute Prism mesh (1st layer height = total layer height = 0 so prisms will float, number of layers = 6), check mesh (fix orientations), smooth mesh but freeze PENTA elements. Could not get smoothing quality to be all > 0.3 and about 0.2 - 0.3% (~20,000) of the elements were < 0.3.

d) Split prism layers for the required y+ value and re-distribute, check mesh, output as CFX *.cfx5 file.

I imported the file into CFX and set that up ok but the CFX-Solver returned before the first iteration:

| ERROR #002100011 has occurred in subroutine cVolSec.
| Message:
| A negative SECTOR volume has been detected. Execution will proceed
| but this is a possible cause of robustness problems.
| The location of the first negative volume is reported below.
| Volume : -0.5435E-12
| Location : ( 0.66706E+02, 0.10685E+02, 0.76817E+01)
| This warning may be made fatal by setting the expert parameter
| 'negative volume option = 1'

| ERROR #002100012 has occurred in subroutine cVolSec.
| Message:
| A negative ELEMENT volume has been detected. This is a fatal
| error and execution will be terminated. The location of the first
| negative volume is reported below.
| Volume : 0.0000E+00
| Location : ( 0.66708E+02, 0.10689E+02, 0.76821E+01)

My questions are:

a) Why didn't ICEM pick up this negative volume element and fix it during the Check Mesh (which I used a lot and at the very end of the mesh generation process)?

b) Given the element co-ordinates by CFX how can I find it in ICEM to fix it?

Even though I had a high quality TRI_2 and TETRA_4 mesh before computing the prisms. I had about 20,000 poor quality (< 0.3) elements after computing the prisms even after multiple smoothing runs.

Visual inspection showed that they all occured where the floating prism layer total height reduces where two perpendicular walls meet (i.e. where the wing upper surface joins the fuselage). This is because the element size decreases here to improve capturing the flowfield. Because of this I increased the Fillet Ratio to 1.0 based on the Help Guide description but it made no difference.

Image 1 shows a cut-plane with the prism layers and Image 2 shows a close up with the poorest quality (< 0.05%) elements where they get trapped.

Other things I've tried but which have not fixed the problem:

a) Compute less prism layers, even tried computing a single prism layer.

b) Reduce the fuselage/wing/tailplane element sizes so the floating prism layers height is reduced.

c) Fillet ratio = 0.1 (default), 0.5 and 1.0.

d) Computing prism layers with specified initial or total height to stop the floating and height reduction on smaller surface elements. This stopped the poor TETRA_4 elements detailed above but instead made many (tens of thousands more) poorest quality TETRA_4 elements along the trailing edges.

The only option I can see is to not refine the element size at the fuselage/wing/tailplane joins and this will stop the floating prisms to reduce in height but the mesh resolution and therefore the flowfield will be lost.

encoced October 26, 2013 22:36

Hey Stuart,
You solved the problem?

siw October 27, 2013 05:06

Hi. Yes, I solved the problem but it was so long ago that I do not remember what I did.

Far October 27, 2013 05:34

Hi Stuart;

Nevertheless, you can post the best practices for such geometries with tetra + prism meshing you have learned so far.

oewnce932c September 3, 2014 05:25

[Meshing] negative ELEMENT volume has been detected
 
Error in Solver: The ANSYS CFX solver exited with return code 1. No results file has been created.

Hi! got the same issue with ANSYS Meshing. CFX Solver write
"ERROR #002100012 has occurred in subroutine Out_NegVol. |
| Message: |
| A negative ELEMENT volume has been detected. This is a fatal |
| error and execution will be terminated. The location of the first |
| negative volume is reported below. |
| Volume : -0.5079E-10 |
| Location : ( 0.14817E+00, 0.21324E-01, 0.80235E-01) "

I have fluid tetra-mesh around propeller. Solution for me was changing mesh parameters fine/row the cell size. In fact just little change in mesh parameters made new mesh without this tiny negative volume(-0.5079E-10) :cool:


All times are GMT -4. The time now is 07:29.