# Problems of Computing two rigid bodies with dihedral_coeffs~~

Dear Axel,

Recently, we try to calculate two rigid bodies connecting with N14.But when the dihedral angles imported(active),it seems the timestep can not be successfully increasing to the ideal value we want(1.0). Also the missing atoms message will appear. I offer the following as IN.SCRIPT and includes read data file(GB1GB2ok.end.data), Can you help me to solve this problem? Thanks a lot.

units real

boundary p p p
atom_style hybrid ellipsoid molecular

echo both

pair_style hybrid gayberne 0.73 1.87 1.43 25.0 lj/cut 0.3

pair_coeff 1 1 gayberne 3.8 3.6 2 2 10 0 0 0 # 3.8 Ben GB1-Ben2 GB1
pair_coeff 1 2 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 1 3 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 1 4 gayberne 3.8 3.6 2 2 10 0 0 0 #Ben GB1-Ben2 GB2
pair_coeff 1 5 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 1 6 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 1 7 lj/cut 0.000001 0.3

pair_coeff 2 2 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 2 3 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 2 4 gayberne 0.00001 1.75 2 2 2 0 0 01 #
pair_coeff 2 5 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 2 6 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 2 7 lj/cut 0.000001 0.3

pair_coeff 3 3 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 3 4 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 3 5 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 3 6 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 3 7 lj/cut 0.000001 0.3

pair_coeff 4 4 gayberne 3.8 3.6 2 2 10 0 0 0 #Ben GB1-Ben GB2
pair_coeff 4 5 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 4 6 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 4 7 lj/cut 0.000001 0.3

pair_coeff 5 5 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 5 6 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 5 7 lj/cut 0.000001 0.3

pair_coeff 6 6 gayberne 0.00001 1.75 2 2 2 0 0 0 #
pair_coeff 6 7 lj/cut 0.000001 0.3

pair_coeff 7 7 lj/cut 0.000001 0.3

GB1GB2ok.end.data (1.39 KB)

Dear Axel,

Recently, we try to calculate two rigid bodies connecting with N14.But
when the dihedral angles imported(active),it seems the timestep can not be
successfully increasing to the ideal value we want(1.0). Also the missing

how is this ideal value determined? was this under exactly identical conditions?

atoms message will appear. I offer the following as IN.SCRIPT and includes
read data file(GB1GB2ok.end.data), Can you help me to solve this problem?

there is no way to increase the timestep beyond a value where the time
integration becomes unstable, for example when forces become large.
you cannot simply make LAMMPS ignore that fact and expect a meaningful
simulation. your simulation will just generate garbage.

you are using a custom dihedral style. have you verified it is working
correctly?

axel.