curve the mesh failed in PointWise for high order mesh

Hello, 

If I understand correctly, you have a grid in CGNS format and the CAD data on a separate file in some format. You import the CAD file first and then the CGNS file with the grid.

In this case, note that the domains (i.e., surface grids) are likely not constrained to their corresponding underlying database quilts. This relationship CAD-domains is needed for the High Order process to curve the grid as it elevates it. 

The domains can be constrained to the underlying database surfaces using the controls in the Grid, Solve, Attributes tab. I would recommend to check the Grid, Elevate page in the Fidelity Pointwise User Manual (Help, User Manual) for additional details. You will find text, images, and videos explaining what is needed for the curved elevation to proceed. 

I hope this helps. 

Thank you for your quick answer.

Your understanding is completely correct, that's it: I have a grid in CGNS format and the CAD data on a separate file in some format. I import the CAD file first and then the CGNS file with the grid.

I now realize that the domain is not constrained to their responding underlying database results. So I get the high-order straight grid.

I tried your method: constrain the domains to the underlying database surfaces using the controls in the grid, solve, attributes tab. But get the error "Domain is not fully constrained on its interior." I do read the the Grid, Elevate page in the Fidelity Pointwise User Manual, but I can not understand what should I do next.

I found someone else suggesting use menu edit, project ... to associate the domain interior points to the infrared apportionable database entity. Is there any scientific basis for this method?

Regrads.

You are very welcome. 

After a domain is assigned to an underlying database quilt, you will need to refine the domain with the button located in the Grid, Solve panel. This will ensure that the points on your domain (that was unconstrained) are indeed constrained to the underlying database entity.

As you pointed out, another approach to this would be to simply do a projection of the domains onto the underlying database quilts. I am not sure I understand the scientific part of your question. Both methods will end up projecting domains onto underlying database surfaces. 

I hope this helps. 

I do that like this:

First, constrain the domains to the underlying database surfaces using the controls in the grid, solve, attributes tab.

Second, refine/poject the domain with the button located in the Grid, Solve panel.

Now everything looks like good.

However, When elevating mesh, there is an error: Computational Jacobian is negative (-0.000122) for cell with centroid at xxx. And then Error: Could not elevate mesh , Aborted Export CAE.

emm... It means the mesh do not match the geometry? Is there any other possible solution?

Thanks.

I do that like this:

First, constrain the domains to the underlying database surfaces using the controls in the grid, solve, attributes tab.

Second, refine/poject the domain with the button located in the Grid, Solve panel.

Now everything looks like good.

However, When elevating mesh, there is an error: Computational Jacobian is negative (-0.000122) for cell with centroid at xxx. And then Error: Could not elevate mesh , Aborted Export CAE.

emm... It means the mesh do not match the geometry? Is there any other possible solution?

Thanks.

Hello, 

Thank you very much for the additional information. 

Please note that this particular error is explained in the cited page of the User Manual. We also recommend a possible solution that may not apply to this case since you completed grid in CGNS format. 

That being said, it is not clear to me if the grid was generated using the CAD data being imported. If this isn't the case, there could be differences causing cell skewness after the domains projection. 

I would recommend inspecting the region indicated by the error message. Note that you can  use the available Examine functions in addition to a visual inspection. As always, I would recommend consulting the User Manual for additional information. 

One possible solution would be to identify the skewed region (see above) and perhaps perform a manual projection of the appropriate domain(s) using a more relaxed projection tolerance.

Last but not least, a more robust approach if nothing else works would be to recrate the mesh using Fidelity Pointwise starting from the CAD data. 

I hope this helps. 

Hello again,

On another thread, you asked the question below. As mentioned in that other thread, in order to keep things organized, I will reply to this question here. 

"By the way, after the high order curved mesh is generated, the Jacobian message is printed.But when I open the generated high order curved mesh with Gmsh and compute it using Gmsh’s mesh quality module, I find that the difference between Gmsh's minJ/maxJ and the Pointwise's Scaled Jacobian are not equal, please forgive my ignorance, I always thought Scaled Jacobian= minJ/maxJ, where did I misunderstand? How can I find out how the Jacobian, Scaled Jacobian, Normalized Jacobian is calculated after the high order curved mesh is generated and printed? I want to compare it with Gmsh's."

Unfortunately, I am not familiar with Gmsh and how it computes Jacobians. As for Fidelity Pointwise, please refer to this paper for additional details on this topic: https://arc.aiaa.org/doi/10.2514/6.2016-3178.

I hope this helps!