[Rohith Giridhar]

Hi All,

I am new to ANSYS-ICEM. I am currently creating an unstructured mesh of UAE PHASE VI wind turbine blade placed in a domain.
The mesh that I am creating is not taking the trailing edge of the blade into account and I do not know how to correct this.
I have attached a picture of the same along with this post.
Could anybody please help me regarding this?
Thank you!

[shereez234]

unstructured patch dependent or independent?

[Rohith Giridhar]

I used the volume mesh option and the mesh method was Robust (Octree). I see that the patch dependent/ independent is a mesh method used while creating a surface mesh. The mesh that I am creating is for a 3D-blade. Am I required to perform surface mesh also before I create the volume mesh?

[shereez234]

you could approach both ways. octree uses patch independent meshing by default I guess. Did you build topology? and what is your smallest curve length?

[Rohith Giridhar]

Yes, I used the Build Diagnostic Topology option with all default values. Also, the smallest curve length is 0.285m.

[shereez234]

Originally Posted by Rohith Giridhar

Yes, I used the Build Diagnostic Topology option with all default values. Also, the smallest curve length is 0.285m.

Make sure you build topology with 0.285/10 =0.0285 or less so it could capture this detail. And let me try and explain to you from what I know;

Two ways to mesh - Patch dependent or Patch independent surface meshing. Two types of volume meshing Octree and Delaunay. Delaunay is better than Octree in giving smooth transition.

Patch dependent works best if your geometry is air tight ( no or less holes with good build topology ( red curves mostly). on surface meshing choose ( respect line elements and set minimumm size below your smallest element length) so you can set curve parameters and surface mesh will form with respect to the node spacing you give.

If you are using patch independent surface meshing then you don't need to set the curve parameters and it is best practice to delete the unneccessary curves. you can still select curvature/proximity based refinement to give a better mesh.

More specifically from judging by the picture you choose I will say reduce the size of your element ( more nodes on that curve?) give it a try and come back!

[Rohith Giridhar]

Thank you very much Shereez! Ill try the same and get back to you on this

[Rohith Giridhar]

Hi Shereez,

I followed the steps you had mentioned earlier and used patch dependence and mesh type of "All tri" for my surface mesh and Quick(Delauney) for my volume mesh.
The mesh took the trailing edge into consideration this time. However, I was unable to provide a finer mesh at the trailing edge or the tip of the blade. I tried increasing the number of nodes along these curves and also decreased the size of the element in the surfaces of these regions, but I didn't find any change in the mesh. I have attached the pictures for the same along with this post.
I also tried providing a small number for the minimum size limit option during curvature/proximity based refinement but did not see any changes.
Could you please let me know how I may go about providing a finer mesh near the trailing edge and tip of the blade?

Thank you!

[shereez234]

curvature proximity/based refinement works well for patch independent surface meshing. Is this what you used? I am free for few days so if you need help share your geometry or PM me.
and I can't tell you what you are exactly doing wrong unless I give it a quick try on your project. Because there is so many things you can check on. Geometry quality,Blunt or sharp trailing edge? sizes set individually and global size limits you set, Triangulation tolerance, geometry tolerance, topology.. So errors could come from any where.

[shereez234]

Also another method you can use and some times I use when I have enough headache from this problem. Put a curve close to trailing edge curve through points. Project the curve on the surface. Split the surface by that curve. And set a smaller size limit to that surface.

[Rohith Giridhar]

Hi All,

I was able to obtain the mesh for the blade and the domain. With the guidance from Shereez, these were the following steps that I took:

a.) To make sure the trailing edge of the blade is taken into account and to obtain a fine mesh on the tip of the airfoil, I set the minimum size limit ( Global Mesh Parameters=>Global Mesh Size=>Curvature/Proximity Based Refinement=> Minimum Size Limit) as 0.001 ( This value was chosen because this was the smallest surface element size that I had used during surface mesh set up) and Shell Meshing Parameters=>Respect line elements and Ignore size=0.0009 ( a value smaller than 0.001 was chosen for this.).

b.) When the mesh was made finer, the transition region in the blade was not captured properly. To capture this, I split the surfaces in the transition region of the blade using curves that were projected onto this surface. I used the following steps for this:

i.) Create points=> Based on 2 locations=> Method: N-point
(This created points on the curves that bounded the two surfaces.)

ii.) Join points using lines.
iii.)Using the icon "Project Curve on Surfaces" ( found in Create/MOdify lines) I was able to project these lines on the surfaces.
iv.) Geometry=> Repair Geometry=> Build Diagnostic Topology=> Apply.
v.) Now, the curves had split the surface into smaller surfaces and using suitable values for number of nodes for curves and Maximum size for the surfaces, I got the mesh in this region also.

I have attached pictures for all of these below