hmm. Do you know how much your model will weigh? I'm thinking that using a pneumatic cylinder might be a good way to go with simulating the recoil and thus the vibration. The cylinder stroke could be around 1 inch and cycle at a rate similar to the firing rate of the real mechanism. You'd have to find a relay that would cycle the pneumatic valve at a rate that is acceptable. Of course, you could build a simple 555 timer circuit which cycles the relay, thus cycling the pnematic valve. The 555 timer could be triggered by the actual trigger mechanism, using some sort of electric contact, or physical switch mounted at the rear of the trigger stroke. You may get lucky and find that the pneumatic discharge could effectively simulate the actual firing. That is, if you have compressed air available?