Hi all,
I have started to do an MD simulation of glycol(in water) but having and issue on the interatomic interactions of the alcoholic hydrogens. I am using OPLS ff for the simulation. From the trajectory, it is seen that those hydrogens overlap with nearby oxygen atoms (which are non-bonded) and flew out of the box very far with a huge energy and temperature. It looks like that those hydrogens are not facing any repulsion with a nearby nonbonded oxygen atom, overlap with the oxygen atom as soon as the NVT is started and flying very far. My query is that why this bad dynamics is happening. Is it because the L-J parameters are zero for the hydrogens of di-ol and therefore the interatomic potential is zero for any cross-interaction? If this is the case then is there a flag in Lammps to handle it? I am fairly positive that all parameter values are correct and in the right order. If anybody else has had this issue, do you have any idea what my mistake might be? Thank you very much and waiting for a reply.
Hi all,
I have started to do an MD simulation of glycol(in water) but having and issue on the interatomic interactions of the alcoholic hydrogens. I am using OPLS ff for the simulation. From the trajectory, it is seen that those hydrogens overlap with nearby oxygen atoms (which are non-bonded) and flew out of the box very far with a huge energy and temperature. It looks like that those hydrogens are not facing any repulsion with a nearby nonbonded oxygen atom, overlap with the oxygen atom as soon as the NVT is started and flying very far. My query is that why this bad dynamics is happening. Is it because the L-J parameters are zero for the hydrogens of di-ol and therefore the interatomic potential is zero for any cross-interaction?
the lack of cross-interaction should not be a problem, since the hydrogen atoms are bound to oxygen atoms which will have a repulsion (LJ and coulomb) with each other.
If this is the case then is there a flag in Lammps to handle it?
this is a property of the force field, so if you are using the force field correctly and the force field is suitably parameterized for the compounds you are simulating, then this situation should not arise in the first place. logically, there cannot be a flag to tell LAMMPS “don’t do that”, since this behavior is a direct consequence of assigning atoms types, partial charges, bonds and the corresponding parameters.
I am fairly positive that all parameter values are correct and in the right order.
i admire your confidence. but the statistics speak against you. in my observation, people that observe behavior like you do, the cause is almost exclusively either a mistake in force field parameter assignments or the use of an unsuitable force field.
axel.