Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
(Solved) A problem with deformed mesh
Posted 11 févr. 2015, 21:03 UTC−5 2 Replies
Please login with a confirmed email address before reporting spam
Hi all, I am trying to solving a very simple heat transfer problem with one moving boundary due to corrosion. However,
Description:
1. 2D transient heat transfer in a rectangular solid block;
2. Left boundary is constant 600 K, right boundary is constant 300 K, top and bottom are insulated.
3. Right boundary is moving because of corrosion. Corrosion rate (boundary moving speed) is linear to height.
Problem:
If the corrosion is faster on upper part of right boundary, the calculation is reasonable. If the corrosion is faster on lower side, the moving mesh deformation is weird. Please see the attached figure and model files.
I think the problem is in the definition of arc length function s( ). Is there anyway to bypass using s( ) here?
Description:
1. 2D transient heat transfer in a rectangular solid block;
2. Left boundary is constant 600 K, right boundary is constant 300 K, top and bottom are insulated.
3. Right boundary is moving because of corrosion. Corrosion rate (boundary moving speed) is linear to height.
Problem:
If the corrosion is faster on upper part of right boundary, the calculation is reasonable. If the corrosion is faster on lower side, the moving mesh deformation is weird. Please see the attached figure and model files.
I think the problem is in the definition of arc length function s( ). Is there anyway to bypass using s( ) here?
2 Replies Last Post 12 févr. 2015, 13:36 UTC−5
Please login with a confirmed email address before reporting spam
Posted:
10 years ago
Hi
For the bottom I used
-(2-y)*x*1e-5[1/(m*s)]
Otherwise also the bottom left corner starts to travel.
Note the units; also in the first deformation change the brackets to [1/s] making the unit correct. I also used extra fine mesh. Now it is almost correct, only a small nagging of an inverted element at the lower right corner. See attached image.
br
Lasse
For the bottom I used
-(2-y)*x*1e-5[1/(m*s)]
Otherwise also the bottom left corner starts to travel.
Note the units; also in the first deformation change the brackets to [1/s] making the unit correct. I also used extra fine mesh. Now it is almost correct, only a small nagging of an inverted element at the lower right corner. See attached image.
br
Lasse
Attachments:
Please login with a confirmed email address before reporting spam
Posted:
10 years ago
Hi Lasse,
Thank you for your reply!
I've solved the problem. It's because the direction of the bottom edge is from right to left (you can see it by turning on edge direction arrow in View node), therefore the prescribed x velocity should be -(2-y)*(1-s)*1e-5[1/(m*s)].
I've tried replacing s with x, it works, but is only an approximation. s always ranges from 0 to 1 along the edge, but with the corrosion the x value of the right end is no longer 1. In more complicated cases, this may not be acceptable.
Thank you for your reply!
I've solved the problem. It's because the direction of the bottom edge is from right to left (you can see it by turning on edge direction arrow in View node), therefore the prescribed x velocity should be -(2-y)*(1-s)*1e-5[1/(m*s)].
I've tried replacing s with x, it works, but is only an approximation. s always ranges from 0 to 1 along the edge, but with the corrosion the x value of the right end is no longer 1. In more complicated cases, this may not be acceptable.