I would like to know if it´s possibe to measure in FEMAP the inner volume of a FE mesh (plastic bottle) before load application and after loading application (internal pressure) in order to obtain the volume increment due to internal pressure.
How I should proceed?
Thanks in advance.
I may have a solution for this, but you're going to have to elaborate:
- is your mesh 2D?
- is your mesh convex (plastic bottle so i'm guessing yes)?
- does your mesh have 0 or 1 hole in it, no more (again, I'm guessing yes)?
I had a similar problem a few years ago and I wrote a small bit of code I can give you, if all your answers to the question above are "yes".
Here's a few tips on the matter:
- without API, FEMAP can create a 3D mesh from 2D elements enclosing a domain. If you've got a hole that' a problem because FEMAP won't be able to deal with (at least I don't think so), but that's a minor problem: you can easily create a 2D mesh to fill the hole. So once FEMAP has created that mesh (command: "Solid Mesh fom Elements") you can measure its volume.
Now as for the deformed model, you'll have to use an API. There is one, given with FEMAP, which moves nodes to their deformed location ("Move nodes by Deform" I think), then you can repeat the procedure and voila.
- my API asks for a "center point", then computes tetra/pyramid volumes from this point and deduces the total volume. This point needs to be chosen well if you have holes. Illustrating this is easy in 2D:
I had half a more-or-less spherical tank (thick black line) and wanted to know it's volume. I specify a point along the red line (which doesn't exist in the FEM, the elements are along the thick black line - nodes in green), then for each element I compute the volume of the tetrahedron which is formed. For this 2D illustration this means computing the area of each blue triangle, add them up and voila.
If the "center point" is above the red line, the computed volume is bigger and vice versa.
It's pretty easy to see why the 3 conditions I mentionned need to be met, otherwise the result is worthless. And the choice of the "center point" needs to be thought out.
Thank you very much for your answer.
I had not heared about this API (Move nodes by Deform) before and it´s perfect to solve this problem.