[CFD2016]

Hi everyone,

I am meshing for the container ship which is located in the wind tunnel.
The problem is the meshes long the intersection between the stern and bottom wall have high skewness (>0.9).
I have tried to use sizing function and mesh sizing (body, face, edge),however, the problem could not be solved.
Attached is the mesh elements with high skewness.
Please help me!

Thank a lot!

Best!!

ANSYS Mesh 1.jpg

ANSYS Mesh 2.jpg

[Kapi]

did you use "HARD" behaviour in Sizing function?

[CFD2016]

Originally Posted by Kapi

did you use "HARD" behaviour in Sizing function?

Hi Kapi,

Thank you very much for your quick reply.
I have tried both "Soft" and "Hard" behaviour in sizing function. However, there are some remained high skewness elements.
What could I do?

Mesh 3.jpg

Mesh 4.jpg

[Kapi]

mate, use Edge sizing with hard behavior and reduce your element size and it will solve the problem!

[CFD2016]

Originally Posted by Kapi

mate, use Edge sizing with hard behavior and reduce your element size and it will solve the problem!

I reduced the element size half, even 1/10, but the problem could not be solve.

Mesh 5.jpg

[CFD2016]

Could you give me your email address? I want to discuss more detail about my problem.
Thanks!

[Kapi]

Please discuss whatever you want to discuss! its beneficial for others having similar problem. Reduce your element size in that area by 1000 and run it. See what the problem is!
If you dont have any edge thr then use sphere of influence!

[CFD2016]

Originally Posted by Kapi

Please discuss whatever you want to discuss! its beneficial for others having similar problem. Reduce your element size in that area by 1000 and run it. See what the problem is!
If you dont have any edge thr then use sphere of influence!

There is an edge between Hull and bottom wall as you can see in the image.
I had tried to use body and sphere of influence but there was no improvement.
Now I will try with smaller edge elements size as your suggestion.
Thanks!
Edge sizing.jpg

[CFD2016]

Originally Posted by Kapi

Please discuss whatever you want to discuss! its beneficial for others having similar problem. Reduce your element size in that area by 1000 and run it. See what the problem is!
If you dont have any edge thr then use sphere of influence!

I reduced edge elements size to 0.0001m (Lpp=1.5m). I supposed this is small enough, however, the number of high skweness cells has increased. May be the large cells were divided into many smaller cells.
Pls take a look at the photo.
Edge sizing mesh.jpg

[Kapi]

umm!!
try Bias type with small parts towards end and bigger parts away from that area and give bias factor of 10 and see how it looks. Reduce the bias factor if required!

[CFD2016]

Originally Posted by Kapi

umm!!
try Bias type with small parts towards end and bigger parts away from that area and give bias factor of 10 and see how it looks. Reduce the bias factor if required!

Hi,

I have tried to use bias factor (10-5-2-1.2) with the element size 0.0001~0.001.
However, it could not be solved.
Do you have any idea?
Do you think the problem may come from the geometry, the angle between the hull (stern) and the bottom surface? If so, how to fix this issue?
When I used patch independent, the maximum skewness < 0.9 (around 0.88). But the mesh is very coarse.
It does not success with other meshing methods like Hexa or Cut cell etc.
Attached is the details of mesh which I am using.
Mesh details.png
The length of ship model is 1.5m.
Total number of elements is approximate 5million for patch independent and 3.7million for patch conforming.
Even I tried to reduce the min and max size, It still could not be solved yet.

Thank you!

[Kapi]

you dont have to use full model of ship.
use symmetry and reduce your mode size.
I dont see this as geometry problem but i have never done ship modelling, so I will leave it expert who has done ship modelling.
Search the website for similar topics and you mite be able to get some answer!

[CFD2016]

I am going to calculate the wind resistance acting on the ship, the wind direction changes in case (0~180deg). That is why I use the full model.
Any way, thank you so much for taking your time and discussing.

Best regards,