[gmx-users] Strange dgdl-value together with lincs

[Maik Goette]

Hi

While trying to find, whats wrong with some of my free energy 
calculation, I stumbled upon a strange thing, I don't understand.

I'm simulating the interconversion of ethane to methanol in vacuum.
I'm splitting the process into 3 steps:
1. charge of perturbed atoms -> 0
2. morph of LJ-parameters and bonds
3. charge of perturbed atoms -> B-state values

Each system is morphed with 1 fs timestep and within 50 steps.
The total energies at the "borders" of the simulations (qqoff/vdw, 
vdw/qqon, qqon/vdw and vdw/qqoff) perfectly match and show no jump; the 
following simulation always got energies and velocities from the one before.

Now, when running the LJ morph A->B with lincs, the dgdl looks like this:
http://www.mpibpc.mpg.de/groups/grubmueller/start/people/mgoette/images/dgdl_lincs_01.png
For B->A:
http://www.mpibpc.mpg.de/groups/grubmueller/start/people/mgoette/images/dgdl_lincs_10.png
Doing both without lincs yields:
http://www.mpibpc.mpg.de/groups/grubmueller/start/people/mgoette/images/dgdl_nolincs_01.png
For B->A
http://www.mpibpc.mpg.de/groups/grubmueller/start/people/mgoette/images/dgdl_nolincs_10.png

I find this spike in the first step extremely strange. Additionally it 
only occurs, when using shake or lincs as constraint solver. I guess, in 
a longer process it will contribute to the integral in a neglectable 
manner, but maybe theres some bug. The latest CVS shows the same behaviour.

Any comments on that?

Regards

-- 
Maik Goette, Dipl. Biol.
Max Planck Institute for Biophysical Chemistry
Theoretical & computational biophysics department
Am Fassberg 11
37077 Goettingen
Germany
Tel.  : ++49 551 201 2310
Fax   : ++49 551 201 2302
Email : mgoette[at]mpi-bpc.mpg.de
         mgoette2[at]gwdg.de
WWW   : http://www.mpibpc.gwdg.de/groups/grubmueller/