Hello,
I am trying to simulate the dynamics of an adsorbate (A) in a flexible framework (F) and the following things are required.
- Specific dispersion interactions - via pair lj/cut
- Real space and Ewald interactions for adsorbate with the framework. For this interaction the adsorbate charges are q_A and the framework charges are q_F_A.
- Real space and Ewald for framework/ framework interactions. The class2 force field that is required is quite complicated and is defined with charges q_F_F, which are different from q_F_A.
- Thus,
the framework/framework coul needs to be defined with q_F_F
the framework/ads coul needs to be defined with q_F_A
Now, I have modified pair_coul_long.cpp to take care of the real part of the coul interaction. This is done by having if statements that check itype and jtype and assign charges based on them.
Do you have suggestions about how the Ewald part of the interaction would work. Before I go ahead and try hacking the code, is it even possible to have ewald summations for two different charges?
Thank you,
Regards,
Ambarish
Hello,
I am trying to simulate the dynamics of an adsorbate (A) in a flexible
framework (F) and the following things are required.
1. Specific dispersion interactions - via pair lj/cut
2. Real space and Ewald interactions for adsorbate with the framework. For
this interaction the adsorbate charges are q_A and the framework charges are
q_F_A.
3. Real space and Ewald for framework/ framework interactions. The class2
force field that is required is quite complicated and is defined with
charges q_F_F, which are different from q_F_A.
4. Thus,
the framework/framework coul needs to be defined with q_F_F
the framework/ads coul needs to be defined with q_F_A
Now, I have modified pair_coul_long.cpp to take care of the real part of the
coul interaction. This is done by having if statements that check itype and
jtype and assign charges based on them.
Do you have suggestions about how the Ewald part of the interaction would
work. Before I go ahead and try hacking the code, is it even possible to
have ewald summations for two different charges?
No. Your whole approach is flawed. you are using an inconsistent set
of interactions. the mechanical coupling of force fields with pair
hybrid only make sense, if you have direct mechanical interactions
(i.e. mostly repulsive and short range). the reasonable approach for
your case would be to parameterize all compounds for the same force
field and have consistent interactions throughout. please keep in mind
that different force fields have different balances between coulomb,
dispersion and bonded interactions as well as solvent-solute. just
picking some parameters from one force filed and some parameters from
some other force field and trying to make them all work is asking for
trouble.
axel.