Re: [ESPResSo-users] conservative force in DPD

[Stefan Kesselheim]

Dear Zhicheng Guo, dear Espresso-Users,

On Nov 27, 2012, at 3:37 PM, guozhicheng222 <address@hidden> wrote:

> Dear all
>        We constructed a system containing TATB molecule (typed 0) and F2313 
> polymer (typed 1). We used the DPD methods to describe the interaction 
> between beads 0 and 1. In DPD simulation, the contribution of conservative 
> force defined as follow:
> FijC=aij(1-rij)
> It should be a soft repulsion.
> The dissipative force and stochastic force can be depicted by the given 
> parameters of gamma and r_cut in dpd command (inter 0 1 inter_dpd $gamma 
> $r_cut $wf $tgamma $tr_cut $twf). Our question is that how to obtain and link 
> the conservative term (aij) to the interaction parameters (such as ε, σ, 
> rcut, cshift and roff) if we choosed the lennard-jones potential to represent 
> the conservative contribution. Thanks for your response.

You probably want to use either the HAT potential (which you mentioned) or some 
other short range interaction potential (LJ, .. ) or even no conservative 
potential between the DPD particles. The thermodynamic properties 
(compressibility, …) will depend only on the conservative interactions and 
transport coefficients (viscosity, speed of sound, diffusion constant) will 
depend on both, conservative and the dissipative forces.

There is no general theory of how to obtain the transport coefficients in 
general, and you will have to calibrate them: E.G. create a test system and 
measure the viscosity. This is one of the drawbacks of DPD, but you have to go 
through it. The right choice depends very much on your system and the related 
scientific question.

Did this answer you question?

Cheers and good luck
Stefan