Scan-and-Solve for Rhino

Simulate Early, Simulate Often... In Rhino

Rating:
  • Currently 0/5 stars.

Views: 153

Comment

You need to be a member of Scan-and-Solve for Rhino to add comments!

Join Scan-and-Solve for Rhino

Comment by Michael Freytag on November 27, 2012 at 11:36am

10 MB by email should not be a problem.  Send it to support@intact-solutions.com.

I agree that computation time seems excessive, but in looking at the screenshot, the beams appear to be made out of some REALLY thin material.

Technical Aside:  During the "Building integration tree" phase, Scan&Solve is building an tree-like data structure (the integration tree) with leaves that adapt to detail in the model.  A structure that is thin EVERYWHERE is algorithmically-interpreted as having detail EVERYWHERE, so the data structure becomes very large.  In the "Building linear system" phase, the integration tree is processed to build the system of equations to approximate the structural behavior of the model.  The larger the integration tree, the longer the time required traverse it to build the system of equations.

As the size of the elements in the simulation approaches the size of the detail in the model, the computational effort required for adaptation can go down.

So... for models that have thin features everywhere, running at high resolutions may actually reduce the overall computation time.

~Michael

Comment by Mike Calvino on November 26, 2012 at 6:41pm

Current state - running since 7:28am, now 6:41pm.

Comment by Mike Calvino on November 26, 2012 at 6:37pm

I have x,y,z restraints at the ends of the long truss (welded to W14's) + I have X,Z restraints on the vertical flat side of the perpendicular, short trusses.  Load is along top surface of long truss as a pressure of 4psi (effective pressure from roof deck live/dead loads).

Comment by Mike Calvino on November 26, 2012 at 6:33pm

I can send the model, but it may be a little large.  Maybe 10Mb, not sure if it will upload.  Also, I'd like to try getting the lower chord attached to bottom flange still, have not been able to do that one yet.  I just started another solution this morning at 7:28am after I realized that it was still working when I cancelled it last time.  It's now 6:24pm & it's about 85% now.  It is still progressing.  Does this sound peculiar?  Should it take this long?  I know there are many thin members in it & I figured it would take awhile, but it seems excessive.

Should I upload the model here, or can I send via email?

Thanks for your input Michael.

Cheers!

Comment by Michael Freytag on November 26, 2012 at 9:53am

Mike,

Hmmm... that's interesting.  What kind of restraints are you applying?

Is this a proprietary model or can you send it to me for debugging?

~Michael

Comment by Mike Calvino on November 25, 2012 at 8:03pm

I could not get it to resolve, ran for like 23 hours (running a quad core laptop w/ 4@2.3Ghz & 20GbRam)

It did not resolve.  Maybe I am doing something wrong.  At first I had the resolution at 10,000 elements (all the way to min.)  So I just started it again with resolution about 75% (154400 elements), the progress bar is about 66% and the task label is "Scanning geometry".  Thats where it hangs every time & does not get past this point (the approx. 66% mark with progress bar).  It's been going about . . . maybe 2 hours.  What do you think I'm doing wrong?  maybe something wrong in the geometry/solid? 

I just thought of something I'm going to try.  I think maybe the lower chord of the trusses is not connected to the lower flange of the can & this may be posing some problem.  I'll tr to fix it & see what happens.  Thanks for the tip about raising the resolution.

Cheers!

Comment by Michael Freytag on November 25, 2012 at 1:36pm

Mike,

At what resolution were you running this?  You will definitely want to run at a high resolution given the many thin features you have in this structure.

~Michael

Comment by Mike Calvino on November 23, 2012 at 9:18am

You can see the trusses running perpendicular to the long one, these are attached to another truss and would not be completely restrained.  I tried running the simulation with this whole structure as one solid, but it never resolved even after about 14 hours.  So I stopped it & just ran the sim. on the skylight using the flat surfaces where the short trusses attach as restraints & placing the loads (tension / compression) for the long truss where it attaches.  At least it gives me an approximate idea of the stresses in the angled cylinder, but would like to be able to test the entire assembly.

FOLLOW SCAN&SOLVE

© 2024   Created by Michael Freytag.   Powered by

Badges  |  Report an Issue  |  Terms of Service