[Evogrid-dev] Bonding improvements

[Tom Barbalet]

It's interesting to think of GROMACS as a visual language. I suspect people who use it heavily (or other chemistry simulation packages) would be able to visually tell what physics simulation method was being used and when it was hand rolled.

Is anyone on this list in that "visual language" group with regards to GROMACS or other packages?

Best regards,

Tom Barbalet.

On Mar 10, 2010, at 1:08 AM, Peter Newman wrote:

> Not really, I shouldn't have left them unimplemented. But they are done 
> now. I'm not sure they were the cause of the problem but they seem 
> related, and it's always best to be as correct as possible.
> 
> I also fixed another error where all the bonded atoms were using the 
> same interaction force, usually a weak one specified for non-bonded 
> interactions. So now atoms bond together and are pulled to each other. 
> This is currently introducing other problems, but I think they are 
> mostly due to bonds being too strong, or being formed between 
> inappropriate atoms.
> 
> So, before atom bonds weren't acting like bonds. Now they do, too much 
> so. On with the developing!
> 
> Peter N
> 
> On 10/03/2010 3:21 PM, Bruce Damer wrote:
>> Excellent catch Peter!
>> 
>> Bruce
>> 
>> At 10:42 PM 3/8/2010, you wrote:
>>> I'm working on improving why bonds act so strangely, and I think I found
>>> why. My code had a few //TODO's marked in about generating some of the
>>> data structures used for bonds! It looks like the bonds were being made,
>>> but the array that told GROMACS to simulate the forces involved in those
>>> bonds wasn't. D'oh!
>>> 
>>> So that's what I'm working on now. Hopefully that's what's missing to
>>> make bonds actually behave like tight couplings between atoms, rather
>>> then the big pin's through the simulation pin-cushion.
>>> 
>>> There's still improvements that could be done to the bonding process
>>> (such as per pair type thresholds, rather then a singular one for all
>>> combinations), but this may not be needed in the short term, and in the
>>> long term we're looking at moving to pure QM.
>>> 
>>> Peter N
>>> _______________________________________________
>>> Evogrid-dev mailing list
>>> Evogrid-dev at lists.digitalspace.com
>>> http://lists.digitalspace.com/cgi-bin/mailman/listinfo/evogrid-dev
>> 
>> 
>> _______________________________________________
>> Evogrid-dev mailing list
>> Evogrid-dev at lists.digitalspace.com
>> http://lists.digitalspace.com/cgi-bin/mailman/listinfo/evogrid-dev
> 
> _______________________________________________
> Evogrid-dev mailing list
> Evogrid-dev at lists.digitalspace.com
> http://lists.digitalspace.com/cgi-bin/mailman/listinfo/evogrid-dev

 --- --- --- --- --- --- --- --- --- --- 
Tom Barbalet http://nobleape.com/tom/
Noble Ape    http://nobleape.com/sim/
Community    http://biota.org/podcast/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.digitalspace.com/pipermail/evogrid-dev/attachments/20100310/0d0a6e09/attachment.htm