Help with minimization

LAMMPS LAMMPS Mailing List Mirror
1

Dear Users

I’m trying to minimize a system which has two large parts: a nanotube and a helix of nucleotides which is non-covalently bound to the nanotube. I will call the system A & B, respectively. I’m using a Conjugate Gradient minimization with a force tolerance of 1.0-8. First I minimized the whole system until the convergency criteria is met. This means that the gradient with respect to the whole set of coordinates (A,B) is zero. Then I take the minimized geometry and I minimize it again but now freezing the helical structure (freezing B). I would have expected no further minimization. However, system A changes and the total energy now reaches a new minimum., considerable lower than in the first minimization. Why is this? I’m using lj/cut/coul/cut with a 32.0 Angstrom cut-off. I don’t understand why the system reaches a lower minima after partially freezing part of the system. If it was at a local minimum without constraints it should still be in the same local minimum after freezing part of the system.

Thanks

Milinda Samaraweera
Computational materials chemistry and Computer programming
University of Connecticut
Department of Chemistry
55 N Eagleville road
unit 3060
Storrs CT
USA

2

A tolerance of “1.0-8” is not equivalent to “1.0e-8”. If indeed you literally typed the first one in the input script, it will be read as “1.0” and the “-8” following discarded. The minimization will proceed after freezing because the force criteria you think you have is not the one the code is following. A simple look at the log file should show the “Force two-norm final” reporting a value above 1.0.

Carlos