[lammps-users] bug? and question

LAMMPS LAMMPS Mailing List Mirror
1

Hello all,
I’m a new user of Lammps and I have a few questions:

  1. Morse potential
    I’ve been trying to read the coefficients for the morse potential from the data file however I get this “ERROR: All pair coeffs are not set”. For example, if I change the pair style to “morse 10” for the peptide example (and change the kspace style to none), I get this error. Is there something special about secifying the morse coefficients in the data file?

  2. In the Charmm dihedral type, the parameter d is meant to represent the sign, as such it only takes value of 180 or 0. What would be your suggestion if I wanted this value to be any value between -180 and 180 (a sort of shifted dihedral). Any suggestions on where I should change the code to accomplish this? Finally, does the variable “c” (calculated in line 170 of dihedral_charmm.cpp) represent cos(phi)?

Regards,

Tamara
It seems that if you try to read the coefficients from the f

2

Tamara,

  1. Morse potential
  I've been trying to read the coefficients for the
morse potential from the data file however I get
this "ERROR: All pair coeffs are not set". For
example, if I change the pair style to "morse 10"
for the peptide example (and change the kspace style
to none), I get this error. Is there something
special about secifying the morse coefficients in
the data file?

Make sure your Morse coeffs are in the correct format
as detailed here:
http://www.cs.sandia.gov/~sjplimp/lammps/doc/pair_coeff.html

If you still can't figure it out, send us your input
files and we'll take a look.
    

  2. In the Charmm dihedral type, the parameter d is
meant to represent the sign, as such it only takes
value of 180 or 0. What would be your suggestion if
I wanted this value to be any value between -180 and
180 (a sort of shifted dihedral). Any suggestions on
where I should change the code to accomplish this?
Finally, does the variable "c" (calculated in line
170 of dihedral_charmm.cpp) represent cos(phi)?

If you have the latest version of LAMMPS with all of
the patches and updates (LAMMPS upgrade):

http://www.cs.sandia.gov/~sjplimp/download.html

You can input any integer value for the parameter d.
See the Dec 7 2005 patch:
http://www.cs.sandia.gov/~sjplimp/lammps/bug.html

The variable "c" in DihedralCharmm::compute is indeed
cos(phi).

I hope this helps!

Paul

3

Hello,
Is there a way to output the value of a particular potential interaction
during the simulation. I noticed that the bond classes have a single
function that can potentially be used to track a particular bond
interaction.

Naveen

4

The bond::single() routines are used by hybrid bond potentials
to invoke a bond interaction for a pair of atoms. But there
is no user-mechanism to print the results of just one interaction.

You could easily write a fix to monitor a bond and print it.
Or stick some printf()
lines in the bond routines to print out debug info ...

Steve