Stone arch FEM

About the development of the FEM module/workbench.

Moderator: bernd

User avatar
-alex-
Posts: 298
Joined: Wed Feb 13, 2019 9:42 pm
Location: France

Re: Stone arch FEM

Postby -alex- » Tue Sep 10, 2019 5:34 pm

freman wrote:
Tue Sep 10, 2019 4:04 pm
do you have a FCStd for that model?

Sure, the file is below attached, as a study case.
In order to get the reaction forces on pilars you could get inspired with this thread:
https://forum.freecadweb.org/viewtopic. ... on#p167047

I guess you speak french, if I'm right you could have a look at : https://forum.freecadweb.org/viewtopic.php?f=12&t=37162

Firstly, I'm designer engineer, not structure engineer, so....be carful with this methode and results.
My strategy here is to use contact constraints, in order to get a realistic model.
Saying that, about your need, I'm not so sure such a model will give you true results.
It's hard and long to explain that in english, and maybe I'm wrong, anyway I'm going to try:
I mean, this arch stone is a singular case in regards of an FEM modelisation, because stones can slide along slopes during loading, so that the contact faces are not face to face. You know what I mean? I hope I'm clear enough.
This kind of model involves instability, so usualy the solver fails with an error "too many cutbacks".
With a 3 stones arch you can "cheat" the solver by locking the master stone which is located in the plane of symetry (locked along X and Y direction). Then the 2 other stones are constrained (XnY) to the ground. Then you get results.
But with a model which includes more stones like the one I attached here below, the intermediate stones bring unstability. And you can't lock them because these stones have to be free about displacement in the XZ plan.
In order to solve this issue I had to replace the contact card "surface to surface" by "TIED" in the inp file. Hence the beaviour of the stone arch is maybe not realy realistic. For e.g.I get not realistic reaction forces on the fixed face of this model (some reaction forces along X and Y axis, but it should not :? )
So, be carful with results. Maybe Thomas Schrader is right, graphic static could give better and more reliable results.
FEM-stone-arch-self-weight.FCStd
(540.35 KiB) Downloaded 8 times
FEMMeshGmsh.inp.txt
(76.44 KiB) Downloaded 8 times
FEMMeshGmsh.dat.txt
(126 Bytes) Downloaded 8 times
(remove the .txt extention to both.inp and .dat files)
User avatar
freman
Posts: 904
Joined: Tue Nov 27, 2018 10:30 pm

Re: Stone arch FEM

Postby freman » Tue Sep 10, 2019 6:13 pm

many thanks alex. That other thread looks informative, I go through it later.

Thanks for the word of caution. Slipping won't happen until the stones are already dislodged.

The most likely movement is the lower edges spread apart with a vertical load. I may try just tying the top edges rather than the faces. With cut stone segments that edge is where the force will be, I think. Tying the faces, as you say, will not give a correct result for lateral forces, since it effectively becomes a solid stone arch which will only push down at its points of contact.

If I wanted to simulate a collapse under vertical load maybe tying alternate top and bottom edges and double the number of cut segements. If an arch buckles, one side goes up and the other collapses inwards. That is why you need material above the arch of stones to stabilise it. That is the next problem.

Anyway, that's a great help, I'll have a look if master has been fixed enough to load the FEM WB. It's had a rough few days ;)