← Back to team overview

yade-users team mailing list archive

Re: PSD triaxial

 

Bruno Chareyre said:     (by the date of Mon, 12 Jul 2010 11:47:47 +0200)

> 
> > The TriaxialTest generator is trying to guess the box size from the
> > number of particles. This is why Bruno suggests changing porosity.
> >    
> Nope, the generator is guessing size of particles form number and box 
> size, its a bit the same.
> The "guess box size" part has been implemented by Vaclav I think.
> 
> > But personally I have found that using fixedBoxDims='xyz' and
> > controlling the box size myself gives better results.
> >
> >    
> Mmmmh... How "better"?
> 
> If you find the TT algorithm too slow, keep in mind that it can be 
> speeded up a lot by changing growth coefficients, confining stress and 
> other parameters. CPU time for packing generation has never been a 
> problem for me yet.

by better I mean: there is no error "Exceeded 1000 tries to insert ...".

I prefer to set radius and radius deviation at a fixed value, so that
I have control on the sieve curve.

In fact I want a sample of target size, eg. 10x10x10cm (after the
initial compaction phase) and with a given sieve curve.

To achieve that I am fixing the sieve cure by controlling radius, and
I am trying different xyz and different number of particles, until I
get what I wanted. Can this be done in some better way?

-- 
Janek Kozicki                               http://janek.kozicki.pl/  |



Follow ups

References