[MBDyn-users] Time calculation with 1.3.15

Romuald NORET romuald.noret at cdr.hutchinson.fr
Tue Jul 20 09:04:42 CEST 2010


I spoke with the person who installed MBDyn : he thinks this can be 
related with Umfpack. It seems that this library was missing, he will 
install it as soon as he can.
I will compare time calculation and convergence when it will be done.

Best regards,
Romuald Noret

alessandro.fumagalli at polimi.it 
Envoyé par : mbdyn-users-bounces at mbdyn.org
10/07/2010 15:47

mbdyn-users at mbdyn.org

Re: [MBDyn-users] Time calculation with 1.3.15

Hi Romuald,

strange behaviour... Could you be a little more precise? 

1) Could you check the message given by mbdyn at the end of each run 
"The simulation required XX seconds of computational time" and compare 
this between 1.3.9 and 1.3.15?

2) Which version of gcc did you use?

3) Did you change something in the OS as well?

4) what do you mean with "sometimes" simulations don't run? When? When 
you change some parameters or something else?

By the way, no difference regarding overconstrained systems. And, in 
any case, I do not think your model is overconstrained because you 
have beams in your model and, so, many dofs related to flexibility.

Please let us know.


Def. Quota Romuald NORET <romuald.noret at cdr.hutchinson.fr>:

> Hello,
> I recently upgraded fom MBDyn 1.3.9 to 1.3.15, while working on a beam
> study and I noticed a strange behavior :
> My model can  be summarize as two spindles of beams in parallel, with 
>  joints between them. So it looks to an overconstrained model but its
> required to simulate an torsional effect.
> The load has two phae : a preload in the direction of the beams and an
> oscillation perpendicular to them.
> I haven't recorded time calculation, but it seems it much more longer 
> the new version, at least 10 times, may be more . And in some case,
> simulations don't runs.
> That is why I wonder if new version is more strict with overconstrained
> system and if there is an other thing I must take in account? Or may be
> its a compialtion problem, I don't know.
> Thanks in advance.
> Best regards,
> Romuald Noret

This message was sent using IMP, the Internet Messaging Program.

MBDyn-users mailing list
MBDyn-users at mbdyn.org

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.mbdyn.org/pipermail/mbdyn-users/attachments/20100720/399326f2/attachment.htm>

More information about the MBDyn-users mailing list