yade-dev team mailing list archive

[Bruno Chareyre <bruno.chareyre@xxxxxxxxxxx>]

Hi Vaclav,
> I am trying to put energy trackers throughout the code. A point that 
> is not clear to me is how to compute energy dissipation due to the 
> numerical damping (especially with the aspherical integrator --). Can 
> elucidate that subject a bit?
I never thought about that before.
Thinking in 1D :
- numerical damping is like an added force in "acceleration=(f+f_d)/m"
- the work of this force on dt is (first order approximation) 
dW=velocity*f_d, and this product is always negative by definition of 
Cundall's damping.

dW is defined at the body level, not interactions. It is removed from 
the system in Newton's integration.
This definition should apply for defining 6 dW_i corresponding to all 6 
DOF's present in Newton integrator, since each DOF is damped independently.
Aspherical integration is apparently not damping rotational motion => no 
dissipation.
If torque was damped like in spherical integration, there would probably 
not be anything specific.

I hope it helps. Cheers.

Bruno