← Back to team overview

yade-users team mailing list archive

Re: Exception std::bad_alloc when a simulation is started

 

Hi,

I met other problems concerning the version YADE-0.5:

	1) In fact, I can load my xml file, but for example if I execute
O.forces.f(any id) in the yade console the result is vector(0,0,0) which
is not true. I need at least to run one iteration to answer that.(an
actualisation problem ???)

	2) Concerning the nan values met also by Luc,  I edited my xml file in
"Law2_ScGeom_FrictPhys_Basic" and activated "useShear" and then I loaded
it, Yade run some iterations and when it comes to record, it crashes and
quit telling me "nan value for somewhere rotation" (Recording
problem ???)


Nejib  


Le vendredi 26 novembre 2010 à 18:37 +0100, Luc Sibille a écrit :
> Yes it is by using always the same version (0.5), so it should not be 
> the problem pointed by Chiara.
> 
> The problem appears after a triaxial compression, but I often reloaded 
> xml files after a compression without problems. However, as you said I 
> will double check the registered variables.
> 
> Of course I could use the last bzr revision, but it would mean remake 
> all the sample preparation which is quite long in my case. In addition I 
> make computations on computation servers and it is always quite long to 
> install a new Yade version and not always possible (you need to ask to 
> the administrator for the installation of the new necessary libraries, 
> and these libraries are not always compatible with the configuration of 
> the servers...)
> 
> Thank you for your help,
> 
> Luc
> 
> 
> 
> 
> Bruno Chareyre a écrit :
> > If I understand correctly, the problem is with scenes saved and loaded using the same
> > version (0.5).
> > Usually, when I get this type of behaviour, it is because I forgot to register a variable
> > that was needed, or (the opposite) because I registered a variable that should always
> > start with the default value (typically, a flag isFirst).
> > Maybe this sort of mistake results in memory leak in an infinite loop in your case?
> > Did you find correct behaviour with earlier versions?
> > 
> > Bruno
> > 
> > _______________________________________________
> > Mailing list: https://launchpad.net/~yade-users
> > Post to     : yade-users@xxxxxxxxxxxxxxxxxxx
> > Unsubscribe : https://launchpad.net/~yade-users
> > More help   : https://help.launchpad.net/ListHelp
> > 
> 




Follow ups

References