Re: [ESPResSo-users] problem in calculation of local_stress_tensor

[Axel Arnold]

On Friday 29 June 2012 09:50:18 Olaf Lenz wrote:
> Hello!
> 
> On 06/20/2012 05:38 PM, Tarun Khanna wrote:
> > I am using espresso 3.1.0 for calculation of local stress tensor
> > for our system but following error is coming "background_errors 0
> > {analyze stess_profile: skin cannot be negative}" can you please
> > explain me the error so that i can rectify it.
> 
> Sorry for the late reply.
> I have never seen the error before. Could you please provide the script
> that produces the error? Otherwise it is hard to debug.\


Hi,

I think what happened is simply that you try to calculate the pressure before 
you set the skin. That however is a necessary tuning parameter, without which 
you cannot calculate any short ranged interactions, including stresses.

The skin does not change the physical results, but depending on the value, 
your simulations might run faster or slower by a factor of 2-3. For long runs, 
you should try out what is the best value, but if your simulations are short, 
then about 30% of the typical length scale, e.g. LJ diameter, is a good value. 
That is, for standard Lennard-Jones with sigma=1, setmd skin 0.3 is probably a 
good guess.

Axel

-- 
JP Dr. Axel Arnold
ICP, Universität Stuttgart
Pfaffenwaldring 27
70569 Stuttgart, Germany
Email: address@hidden
Tel: +49 711 685 67609