QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: Roc on February 25, 2009, 04:26

Title: some questions about the geometric optimization
Post by: Roc on February 25, 2009, 04:26

Dear everyone,

In order to relax the two-probe system like Au-DTB-Au, could we optimize the struture by ATK more precisely and quickly and how to set it in VNL? It seems that thare are two geometric optimization mothods Steepest decent and Quasi Newton, which method is better for optmizing the two-probe system and what's the  relative merits for these two methods?  how to set the parameter "Geometric Constraints" in VNL

Thanks!

Title: Re: some questions about the geometric optimization
Post by: Anders Blom on March 5, 2009, 11:13

QuasiNewton is often the faster method (requires fewer steps), but SteepestDescent can converge in situations where QuasiNewton has problems, e.g. if it gets stuck.

The geometric constraints are specified in VNL in the NanoLanguage Scripter, on the "Configuration" tab, and are given just as in NanoLanguage, i.e. you provide the indices of the atoms in the central region that should not move in the relaxation.

Title: Re: some questions about the geometric optimization
Post by: hellboy on November 17, 2010, 08:00

Dear Sir,

Can you give a reference of "Steepest descent algorithm based on a damped velocity-Verlet MD algorithm", which is used by ATK.

Title: Re: some questions about the geometric optimization
Post by: zh on November 17, 2010, 09:19

The information on the following links may be helpful to you:
Improved algorithm for geometry optimisation using damped molecular dynamics
M. I. J. Probert
http://dx.doi.org/10.1016/S0021-9991(03)00308-5 (http://dx.doi.org/10.1016/S0021-9991(03)00308-5)

 the method of steepest descent
http://en.wikipedia.org/wiki/Gradient_descent (http://en.wikipedia.org/wiki/Gradient_descent)