Dear all,
In my simulations using VERLET propagator, the definition of the pair_style is given by
pair_style lj_sf_dipole_sf 12 lj_charmm_lj_long 10 12
where the pair_coeff for AA-AA interactions is given by lj/charmm/coul/long and the hybrid interaction between CG-AA and the CG-CG sites is defined as
…
pair_coeff 22 23 lj/charmm/coul/long 0.135099963 3.0907894236 0.135099963 3.0907894236 -----> AA interaction
pair_coeff 22 24 lj/sf/dipole/sf 0.256904651 3.0395099605 -----> AA-CG interaction
pair_coeff 23 24 lj/sf/dipole/sf 0.289231741 3.1014564401 -----> AA-CG interaction
pair_coeff 24 24 lj/sf/dipole/sf 0.55 3.05 -----> CG-CG interaction
the alchemical free energy of the AA-CG intermolecular interactions (between explicit atoms and CG waters beads) is obtained via compute ti command:
…
compute TI all ti lj/sf/dipole/sf 1*23 v_minusl v_minusdl
and working well.
However, in my rRESPA simulations, using the follow definition for the pair_style:
pair_style hybrid lj/charmm/coul/long 11.0 12.0 lj/sf/dipole/sf 12.0 lj/sf/dipole/sf 12.0
and defining the pair_coeff as
…
pair_coeff 22 23 lj/charmm/coul/long 0.135099963 3.0907894236 0.135099963 3.0907894236 -----> AA-AA interaction
pair_coeff 22 24 lj/sf/dipole/sf 1 0.256904651 3.0395099605 -----> AA-CG interaction
pair_coeff 23 24 lj/sf/dipole/sf 1 0.289231741 3.1014564401 -----> AA-CG interaction
pair_coeff 24 24 lj/sf/dipole/sf 2 0.55 3.05 -----> CG-CG interaction
^
when the ti compute command is invoked (now, defining the number of the specific pair_coeff to be estimated - in this case ‘lj/sf/dipole/sf 1’)
…
compute TI all ti lj/sf/dipole/sf 1 1*23 v_minusl v_minusdl
^
I’ve the follow message ERROR: Illegal compute ti command (…/compute_ti.cpp:54). I haven’t found a similar scenario on the compute ti doc page or in the LAMMPS-list for set this specific type of calculation.
Thanks.