Thanks for your response. I think I understand why the determinant is 0
now. But then, first I do MD in constant energy. I don't need to remove
rotations. Second, why it always crashes in step 900? Does that mean
amber corrects rotation every 900 steps?
Best,
Lishan
On Wed, 2003-08-13 at 09:50, Andreas Svrcek-Seiler wrote:
> Hi,
> > I use amber7 to do MD constant energy simulation for a diatomic
> > molecule. But I get something wrong here:
> >
> > ------------------------------------------------------------------------------
> >
> >
> > %CENMAS-F-INERTIA_TENSOR, determinant is zero ... stop
> >
> > Can some one tell me what this tensor is and why the program is stopped?
> Well, If I remember correctly, it's like the following:
> 1) Assume, for simplicity, that the center of mass coincides with the
> origin.
>
> 2) Then, let L be the angular momentum of your system, which can be
> calculated from the individual velocities (linear momenta) and positions.
> Classical Mechanics teaches L = I.omega, where I is the tensor of inertia
> and omega points in the direction of the 'momentary total rotation'.
> Alas, I don't know the correct english name for omega.
>
> 3) To remove overall rotation, you have to invert I to get
> omega = I^(-1).L . If your system is diatomic (or perfectly linear), I is
> singular and that inversion fails.
>
> 4) If you swith off the heat bath early, there's no need to remove the
> rotation, lengthy 'equilibration' of a diatomic molecule in vaccum
> does not seem to be too useful.
>
> Hope it helps,
> good luck
>
> Andreas
>
>
>
>
>
--
Lishan Yao <yaolisha.pilot.msu.edu>
MSU
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Wed Aug 13 2003 - 16:53:01 PDT
