From: Hadi (dinpajooh_at_gmail.com)
Date: Sun Aug 10 2014 - 22:29:55 CDT
Here is one citation regarding the flaws in the algorithms:
THE DL POLY 2 USER MANUAL, i.e., section
2.5.7 Rigid Bodies and Rotational Integration Algorithms
On Sun, Aug 10, 2014 at 8:10 PM, Kenno Vanommeslaeghe <
kvanomme_at_rx.umaryland.edu> wrote:
> n 08/10/2014 09:59 PM, Hadi wrote:
>
>> the algorithms that use bond constraints have
>> their own limitations. I am not sure how NAMD addressed those issues about
>> bond constraints.
>>
>
> The NAMD documentation quite clearly explains its use of well-established
> constraint algorithms that have been in use since the early '80s.
> Ubiquitously so for the last few decades (especially for water!) You're
> essentially telling us that you, the great future Dr. Hadi Dinpajooh, have
> found that these algorithms don't really work as advertised and that the
> bulk of the MD work in the last few decades is flawed.
> Here's more explanation from which you can rip quotes out of context,
> though I'm seriously considering to stop replying if you do so because I'm
> starting to doubt your good faith in this discussion.
> http://en.wikipedia.org/wiki/Constraint_algorithm
>
>
> Molecule.C
>> is just one subroutine that may or may not be related to the problems.
>>
>
> Maybe I should have made it clear this is the functionality responsible
> for reading the bond section of the .psf file and applying parameters to
> the list of bonds psf. It is critical to take note of what it does with the
> counters; when encountering a bond with 0 force constant, it effectively
> acts as if it never read it. When finished, it issues a message saying:
>
> "Warning: Ignored xxx bonds with zero force constants."
> ^^^^^^^
> Since you're probably going to claim I'm "misunderstanding" the meaning of
> the word "ignored", don't take my word for it, just generate a small water
> box and simulate it for a bit with "rigidBonds all" with various outrageous
> force constants. I can guaranteee you it will always gives the same results
> except when setting it to 0.
>
>
This archive was generated by hypermail 2.1.6 : Thu Dec 31 2015 - 23:21:06 CST