+- Obi Official Forum (https://obi.virtualmethodstudio.com/forum)
+-- Forum: Obi Users Category (https://obi.virtualmethodstudio.com/forum/forum-1.html)
+--- Forum: Obi Cloth (https://obi.virtualmethodstudio.com/forum/forum-2.html)
+--- Thread: Help Long Metal Sheet Simulation (/thread-2605.html)
Long Metal Sheet Simulation - IT_Tao - 02-12-2020
Hello there,
I've been working for a while with the ObiCloth plugin, trying to get a bit deeper in it. My goal is to simulate a very long (about a hundred meters) sheet of metal (copper actually). The sheet is about 20cm-wide, and I want to implement some kind of rigidity to it.
My first goal as been to remove the flexibility in the middle of the band. The solution I've found was to remove all the particles in the middle : I'm generating a long mesh with particles on the sides, and none on the inside. Here's an image of my generated bands :
[attachment=820]
This was made for two reasons : removing the possibility to bend in the length of the band, and considerably reduce the number of particles involved in the simulation (because the bands I use are much much longer).
I've then tweaked a bit the solver settings (increasing the distance constraints iterations to reduce the stretching, and disabling most of the other constraints to lighten the calculation charge). I got to the point where I can use the band almost like I want, attaching and detaching it dynamically with an unwinding - winding setup. I want to optimize it as much as i can : the problem i face today is the bending of the band.
I want it to stay as straight as it can, as if i was holding a sheet of paper by the side :
[attachment=821]
On this example, the band of the left bends a lot. It's the current way the simulation works. I'd like it to act more like the example on the right.
So this si mainly a bending problem. I also have question about the way the forces are calculated, because I'd like to simulate a traction sensor, and i need to get the force applied by the band on a cylinder for example.
Thank you for reading this if you have any info I'm really interested.
RE: Long Metal Sheet Simulation - josemendez - 02-12-2020
Hi there,
Bending is controlled by bending constraints: http://obi.virtualmethodstudio.com/tutorials/bendingconstraints.html
Assuming you haven't deactivated them (they should be enabled in your case), expending more iterations on them will allow the solver to reach higher bending stiffnesses. Also, using more substeps will noticeable improve simulation quality (which means higher stiffness will be easily achieved).
If you haven't already, take a look at this page: it describes in detail the way simulation works and how iterations/substeps affect its quality:
http://obi.virtualmethodstudio.com/tutorials/convergence.html
RE: Long Metal Sheet Simulation - IT_Tao - 10-12-2020
Hi josemendez ! Thank you for your reply.
I've been playing around with the bending constraints parameters, and I've come to satisfying results at some point, but with a very specific type of sheet.
I tried working on a smaller scale, with a band (2mx0.4m, 20edges per meter). The problem I now face concerns the mass of the particles. It looks like the particles are very very heavy, and struggle to keep their distance (they seem to be stretching a lot). Therefore, I've once again played with the number of iterations in the distance constraints and bending constraints calculations. It sometimes resulted in a very very laggy simulation, but I've come to a point where the fps is ok and the band seems to work ok too.
Is there some way for me to interpret the mass parameter ? Because I still need to simulate effort calculation, and the mass seems to have a lot of impact on my measurements.
Currently, I'm rendering forces via the sum of the forces in the particles touching my "measuring cylinder" but I'm not quite sure wether my calculations are logical or not.
Also, I'd need to cut the band at some point, and I haven't a way to do so. I can get the particle indices that I need to cut, but I still don't know how to separate the cloth in two parts.
If you have further detailed info, I'm listening. I don't mind DMs at all if it get too specific.
Thanks again,
Tao.