From: Thomas Brian (thomasbrianxlii_at_gmail.com)
Date: Fri Jun 22 2012 - 15:37:14 CDT
Anyone here know anything about force-grids?
For some reason I can get regular force grids to give me reasonable
results, but force-grid light has issues at my system boundary.
Is it possible to use force-grid lite with PBCs? I thought the restriction
was only that the particle coordinates are not wrapped into the grid space
like in regular force grids, but rather, that the force-grid lite must
extend over your whole domain. I did not see any restriction on PBCs, but
that is where the particles in my system experience large forces that cause
instability.
On Wed, Jun 20, 2012 at 11:08 PM, Thomas Brian <thomasbrianxlii_at_gmail.com>wrote:
> To update my question in case anyone else has similar questions:
>
> -the order of grid points is as I expected, with zindex being the inner
> loop.
>
> -its going to be difficult to replicate NAMDs lennard jones force
> approximation unless you know how they implement their lookup tables
>
> Also, in the .dx file format, you do no include the brackets that are in
> the example. My fprintf line is just:
> object 3 class array type double rank 0 items %i data follows
>
> Also, I don't know if it matters, but I just left the last line of my grid
> data with however many grid pts were left over, so my last line of grid pts
> might have 1 or 2 or 3 data pts.
>
>
>
> On Wed, Jun 13, 2012 at 6:36 PM, Thomas Brian <thomasbrianxlii_at_gmail.com>wrote:
>
>> Hello,
>> I would like to replace a large group of fixed atoms by the field they
>> impose on the mobile atoms using Force-grids. Is their an easy way to do
>> this? Do I need to rewrite the NAMD lennard jones implementation in my own
>> code, or perhaps someone has already done this say in VMD scripting, or
>> matlab? Thus i could just move an imaginary test particle around my grid
>> and have it compute the potential at each point.
>>
>> I am confused by some wording in the manual regarding the layout of the
>> force-grid file. I need to express a 3D grid as a 1D string of numbers. I
>> do not know what you mean be "fast dimension". If my potential function is
>> U(xi,yi,zi), what is the order of the indexes that I need to loop over? I
>> would guess
>> for xi=1..xn
>> for yi=1..yn
>> for zi=1..zn
>> print U(xi,yi,zi)
>>
>> Thanks,
>> Tom
>>
>
>
This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:21:42 CST