Elastic constant script and minimize command

Hi,
I am using the elastic constant script provided in the LAMMPS distribution.

I have a question about the “minimize” command. In the “in.elastic” file it is given as :

minimize {etol} {ftol} {maxiter} {maxeval}

The values of these variable as given in the “init.mod” file are as follows:

variable etol equal 0.0
variable ftol equal 1.0e-10
variable maxiter equal 100
variable maxeval equal 1.0e-2

Now, in LAMMPS manual ( http://lammps.sandia.gov/doc/minimize.html ), we can see the following :

“Either or both of the etol and ftol values can be set to 0.0, in which case some other criterion will terminate the minimization.”

What does it mean ?
Doe it mean that if either ‘etol’ or ‘ftol’ equals to zero (as in the above case mentioned), the stopping criterion will be decided by either ‘maxiter’ or ‘maxeval’ ?

In that case, how can we guarantee that the structure is relaxed properly ?

I have another question:

I want to calculate elastic constants for a damaged system (having lot of lattice defects like frenkel pairs, dislocations, stacking faults etc) using this script.

In that case, attending a stopping criterion like ftol= 1.0e-10 will be very difficult.

So, how can we judiciously choose these parameters in such cases ?

Thanks,

Comments below.

Steve

Hi,
I am using the elastic constant script provided in the LAMMPS distribution.

I have a question about the "minimize" command. In the "in.elastic" file it
is given as :

minimize \{etol\} {ftol} \{maxiter\} {maxeval}

The values of these variable as given in the "init.mod" file are as follows:

variable etol equal 0.0
variable ftol equal 1.0e-10
variable maxiter equal 100
variable maxeval equal 1.0e-2

Now, in LAMMPS manual ( http://lammps.sandia.gov/doc/minimize.html ), we can
see the following :

"Either or both of the etol and ftol values can be set to 0.0, in which case
some other criterion will terminate the minimization."

What does it mean ?

Just what it says. A setting of zero for ftol (or etol) means
that setting is ignored when deciding to end the minimization.
The other criteria are still considered.

Doe it mean that if either 'etol' or 'ftol' equals to zero (as in the above
case mentioned), the stopping criterion will be decided by either 'maxiter'
or 'maxeval' ?

In that case, how can we guarantee that the structure is relaxed properly ?

That's up to you. There is no simple definition of "relaxed properly".
LAMMPS uses the criteria you define.

I have another question:

I want to calculate elastic constants for a damaged system (having lot of
lattice defects like frenkel pairs, dislocations, stacking faults etc) using
this script.

In that case, attending a stopping criterion like ftol= 1.0e-10 will be very
difficult.

So, how can we judiciously choose these parameters in such cases ?

Again, that's not a LAMMPS Q. That's a conceptual Q. I suggest
you try your system and see how it works.