yade-users team mailing list archive

[Bruno Chareyre <bruno.chareyre@xxxxxxxxxxx>]

> I will
> gladly incorporate this to svn :) But before I start doing that I prefer
> to wait till you will have the "final" version. As it will be much
> easier for both of us, if there are less changes done at later time.
>  
I agree! In fact I didn't post the files in order to commit them. They 
are too fresh... 

Thank you for all expanations.
I is really nice to know how to debug with "core*" files.
I heard things about it many times but I've never been able to know how 
to do it.

_I recommend all Yade users who are intersted in debugging but don't 
know how to do it to read the previous post in this thread.
_
Thank you also for optimisation tips. In fact, I already knew most of 
them, but it is never bad to hear this one more time. Currently the type 
of things I think about (in order to optimize) are often related to 
algorithms themselves rather than the way they are implemented. For 
instance, I wonder if subdividing a whole metabody into separate groups 
of body and applying SAP collider to each group would not allow a faster 
detection of contacts. It simply requires that the groups are updated 
each 100 time steps (less/more depending on speeds of bodies). So 
perhaps, from your point of view, this is not optimizing Yade, it is 
writing a faster SAP collider..
You know, I don't plan to work on this type of things very soon, I 
prefer to concentrate on the triaxial test now. But when I read Yade 
code, I can't resist the temptation of asking me if something could be 
done faster. Perhaps I have few ideas, but I have to think twice (or 
more) to them, because I often discover that what I see initialy as a 
good idea is in fact just waste of programming time...

I will clean/rename my code for triaxial, make a file generator that 
trully simulate a triaxial test, and write the timestepper based on 
stiffness matrices. These are next steps.



--
_______________
Chareyre Bruno
Maître de conférence

Institut National Polytechnique de Grenoble
Laboratoire 3S (Soils Solids Structures) - bureau I08
BP 53 - 38041, Grenoble cedex 9 - France     
Tél : 04.56.52.86.21
________________

_______________________________________________
Yade-users mailing list
Yade-users@xxxxxxxxxxxxxxxx
http://lists.berlios.de/mailman/listinfo/yade-users