[tkw954]

I have a theoretical problem where I'd like to solve for the flow through a long curved pipe, where the pipe curves through more than 360 degrees in the same plane. Think of a long multi-turn helical pipe and I'm interested in the limit as the helix angle goes to zero. Obviously this is not a physically possible geometry.



This results in a toroidal shape, but where the multiple turns of the pipe overlap in space, but do not interact. This makes more sense in cylindrical coordinates where the space at an angle of theta is *not* the same as the space at theta+2pi. For example, I'd like to use cylindrical coordinates and apply a boundary conditions at theta=0 and 1080 degrees.

Does anyone have any advice on this?

Not sure if this is really a "meshing" problem, but it didn't really fit in any other categories.

[hjasak]

Interesting.

Unless you are tracking particles or similar, the solver will never find out that the domain is overlapping with itself - and it will work.

You may kill post-processing though

Hrv