Re: [ESPResSo-users] Fwd: Columns for tabulated non-bonded potentials

[Axel Arnold]

Hi!

Yes, the formula below is correct. To verify your tabulated potential, it is 
in any case probably a good idea to use test script like this:

setmd box_l 10 10 10
thermostat off
setmd time_step 1
setmd skin 0

inter 0 0 tabulated "lj1.tab"

part 0 pos 0 0 0

# small step potential and force calculation
set dx 0.01

for {set x 0} {$x < 2} {set x [expr $x + $dx]} {
    part 1 pos $x 0 0

    # calculate forces and energies
    integrate 0
    set F [lindex [part 0 pr force] 0]
    set E [analyze energy nonbonded 0 0]

    # finite difference of energy
    if {[info exists lE]} {
        set dEdx [expr ($E - $lE)/$dx]

        # print out F and dE/dx, which then can be compared e.g. with gnuplot
        # difference should be acceptably small
        puts "$x $F [expr $dEdx]"
    }

    set lE $E
}

which verifies numerically that that the force and potential match.

Axel

On Saturday 20 August 2011 10:20:47 Olaf Lenz wrote:
> Hi everybody!
> 
> Can anybody help Steven? I am on holidays, and I can't check from here...
> 
> Olaf
> 
> -------- Original Message --------
> Subject: Columns for tabulated non-bonded potentials
> Date: Fri, 19 Aug 2011 05:02:38 +0000
> From: Steven Kirk <address@hidden>
> To: Olaf Lenz <address@hidden>
> 
> Hello Olaf,
> 
> I just wanted to confirm with you the necessary order of columns in a
> tabulated potential file for non-bonded interactions in the current
> Espresso release 3.0.1.
> 
> The manual currently says that the order is:
> r, F(r), V(r)
> 
> where F(r) = (-1/r)(dV/dr)
> 
> Is this correct?
> 
> Steven Kirk
> address@hidden
> 
> _______________________________________________
>   Message sent via/by Savannah
>   http://savannah.nongnu.org/

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