3Dblade
 

Hello,

I'm doing the 3D flow and heat transfer simulation over a turbine blade. The domain is the flow passage between pressure and suction sides. There is only one volume. I can mesh the volume using Tet/hybrid elements; but, if I attach a BL mesh on the blade surfaces the mesh fails because of high skewness elements. Any suggestion how to avoid this problem? Thanks in advance

Re: 3Dblade
 

hello , i faced same problems, i am workin also on this project , u can contact me if u want . but to avoid the high skewness , try to mesh separetely the faces failed to mesh because of high skeweness . So . mesh these faces first , then mesh all the volume . or , u can reduce the interval size , but it will need more time . regards, H.A.S

Re: 3Dblade
 

I have also had similar problems with attempting to use boundary layers in a 3-D simulation. Some types of mesh seem to be more forgiving than others. What mesh scheme are trying to use? Also, I kind of circumvented the whole issue by using size functions rather than using boundary layers. Good luck, Garth

Re: 3Dblade
 

I would suggest that you took the time to learn how to create a hexahedral mesh, and use that for you calculations.

Re: 3Dblade
 

Dan,

it sounds like you are using a passage centered topology? You may wish to try a blade centered topology: here you have the freedom to choose your periodic surfaces as being any shape you want...this gives you more freedom in creating a single or multi-block topology which will ease the issues of highly sekewed cells? Just something to think about.

Regards,

Bak_Flow

Re: 3Dblade
 

Thank you for all of your comments and suggestions. I tried to mesh the blade surfaces first and attach the BL mesh on them. I can preview the BL mesh and the EAS quality is very good. Then if I try to mesh the entire volume, none of the schemes is able to mesh it, because of the high skewness on some surface. Then I meshed those faces first, but then none of the scheme is able to mesh the volume at all. I'm also working with a blade centered topology, in this case I need to split the single volume into 3 volumes, and because of the shape of the domain boundaries (they are twisted according to the blade profile and the flow makes a 90deg turn) the resulting volumes will have more sharp corners which lead to higher skewness. I'll try to use size functions and hexahedral cells.

Thanks