[gmx-users] BD and SD equations

Mark Abraham Mark.Abraham at anu.edu.au
Sat Jul 9 04:16:24 CEST 2011

On 9/07/2011 3:20 AM, Marcelo A. Carignano wrote:
> Thank you Mark.
> On Jul 8, 2011, at 11:17 AM, Mark Abraham wrote:
>> On 9/07/2011 1:36 AM, Marcelo A. Carignano wrote:
>>> Hi all,
>>> Stochastic dynamics is described in section 3.8 of the Manual (v. 4.5)
>>> Equation 3.108 is dimensionally incorrect, or at least I cannot 
>>> understand it.
>>> Left hand side is a force,
>> Specifically, mass * length * time^-2. Note that (d^2 r)/(d t^2) is 
>> (d/dt) (d/dt) r, which makes the dimensionality more clear.
>>> and the first term on the right hand side has a missing length in 
>>> the numerator to be a force.
>> It has a missing length dimension to be an *energy* :-)
> you are of course right, I should not work in the mornings!
>>> This is according to the definition of xi, right below the equation, 
>>> as having units of 1/ps.
>>> Brownian dynamics, described in section 3.9, also has a problem 
>>> (again, in my opinion)
>>> gamma is defined as mass/xi, but it is using gamma=mass*xi that we 
>>> can recover the two terms of the SD case
>>> (still with the wrong dimensions)
>> Seems OK to me, similarly.
> But what is gamma?  mass*xi (still is before noon over here)
> If so, the manual has a missprint.

Seems like it does. I posted a Redmine issue.


More information about the gromacs.org_gmx-users mailing list