Simulate Early, Simulate Often... In Rhino
I´m comparing SnS results with analythical solutions on simple cases as part of my learning process with SnS. For me the most difficult thing (and the only one) with SnS is creating the restrains correctly as there is no possibility to let turn the restrained surfaces.
I´m applying perpendicular load on the long edge of a flat bar 1000 x 150 x 10 mm.
I want my bar to be clamped as in the picture. So I restrain part of upper and lower sides.
The resulting deflection is quite strange and I dont know why. Does any body know what am I doing wrong?
Apologies for your difficulties, in this case it is not your fault. You've stumbled across a bug in the current version (V1.3) of Scan&Solve. This bug appears on thinner structures (thickness < approximately 5% of total size) when restraining faces with normals pointing in the "thin" direction. In some cases, these restraints will spill beyond their boundaries, unduly extending their influence to portions of the geometry that are not restrained. I am working on a fix to be incorporated into the next version.
In your particular case, you can obtain a reasonable work-around by cutting off the portion of the geometry you want to restrain and restraining the cut face as depicted here:
Does this help?
~Michael - working diligently on a solution.
I have done what syou suggested me and it works fine.
I wish you could solve the bug as well as you would include the possibility of free turning in restrains!
Every time I use SnS if like it more so I encorage you to improve it and keep it simple!
I encourage you to try Version 1.4. It has improved handling of the restraint case you posted here, as well as the ability to restrain edges - so there is the possibility of free turning about these edges, within the limits of linearity.