yade-users team mailing list archive

[Anton Gladky <question253045@xxxxxxxxxxxxxxxxxxxxx>]

Question #253045 on Yade changed:
https://answers.launchpad.net/yade/+question/253045

    Status: Open => Answered

Anton Gladky proposed the following answer:
Hi Dominik,

thanks for double-checking the code!

2014-08-22 17:07 GMT+02:00 Dominik Boemer
<question253045@xxxxxxxxxxxxxxxxxxxxx>:
> 1.  The forth equation of (22)[Pournin2001] should be the following:
>
>  ct = - 2 / (tc * (1/meff + R1^2/I1 + R2^2/I2)) * ln(et)
>
> Why?  First, the differential equations in (21)[Pournin2001] is similar
> (i.e. it has the same form) to the second one and the initial conditions
> of both equations are also similar.  Thus, the results should be
> similar.  Why is the second equation of (22)[Pournin2001] correct?
> Because this equation is the same than the one in [2], which seems to be
> correct.
>
> [2] Nagurka & Huang_A, Mass-Spring-Damper Model of a Bouncing
> Ball_19_Ijee1697
>
>
> 2.  Taking into account the previous correction, the equations of [Pournin2001] are correct.  This is the reason why
>
> cs1 = cs2 = -2.0/7.0 /Tc * log(Et)*massR; [3]
>
> should be replaced by
>
> cs1 = cs2 = -4.0/7.0 /Tc * log(Et)*massR;

Hmm, it seems, you are right, there is really an error in the paper and in the
code. Let me check it one more time, and I think I will fix it on Monday.

> 3.  As the equations of [Pournin2001] are correct (with the correction),
>
> const Real massR = 2*mass1*mass2/(mass1+mass2); [4]
>
> should be replaced by
>
> const Real massR = mass1*mass2/(mass1+mass2);
>
> according to [Pournin2001, just below equation (19)].
>
> [4]
> https://github.com/yade/trunk/blob/master/pkg/dem/ViscoelasticPM.cpp#L173

As far as I remember we discussed it with Raphael some months ago.
After the code refactoring it was decided to keep the previous behavior
of the initial implementation. It seems, it is time now to fix it....

Thanks again

Anton

-- 
You received this question notification because you are a member of
yade-users, which is an answer contact for Yade.