The effect of one gamma photon on a point particle can be evaluated by considering that the momentum of the gamma photon is p=E/c, where E is the energy of the photon, E=h*f, h being the Planck constant anf f the frequency of the photon.
Once you now the momentum of the photon, you can evaluate the momentum of one point particle after the collision satisfying the conservation of the momentum and energy and add it to the particle. Then, you can follow the evolution of the system by MD.
For a nanoparticle, you should repeat the process picking one particle of the cluster at time, and average over the particles and orientations of the cluster with respect to the photon. If the cluster is symmetric you can spare time avoiding equivalent geometries.
Problems can arise if the photon is highly energetic. In this case, the (repulsive) potentials should not be too steep to avoid too large forces, and accordingly the time step must be tuned to avoid errors in the evaluation of the trajectories.
The LAMMPS package should permit to change the velocity of an atom through the “velocity” button. Before changing the velocity obviously it necessary to equilibrate the system at the desired temperature. It is also better to ensure that after the velocity change the total momentum of the system remains zero, unless the effect on the velocity on the momentum is negligible.
From a short Google search “gamma ray effects on materials MD simulation” I found some interesting information about your problem.
Finally there is a special issue of the journal “Applied Science”, which is still open to contributions, but at least an interesting open access paper is available. Please see: