Regarding what gets produced…
In the case of electron/positron colliders, all of the energy of the colliding particles goes into the particles produced in the collision. Since you know how much energy you gave the incoming particles, you know how much is available to make stuff, and theory tells you what you can and can’t make. Of course, theory could be wrong, especially as you move into uncharted waters, so folks also look for things that don’t fit expectation.
With a proton/proton collider, however, it’s a little messier. You give the protons a fixed amount of energy, but that energy is shared among the protons’ constituent parts (namely, quarks and gluons – collectively, “partons”). It’s actually individual partons that collide, and at any given moment there’s no telling how much energy a given parton will have. So, you end up with collisions where some of the energy is available to produce stuff, and the rest of the energy/quarks/gluons are just debris making a mess of things.
Despite the messier outcome with proton/proton collisions, the basic idea holds though: theory tells you (via simulations of the collisions) what to expect. You try to rely on the simulation only when necessary, and in many cases you can base expectations firmly on decades of real life measurements.
Regarding detector damage…
Absolutely, the detectors get pwn3d. A lot of R&D goes into developing “radiation hard” components, but it’s still not a very nice place to be. The worst case scenario is if the beam shoots through a piece of the detector. There are plenty of safeguards in place to keep that from happening, but if it did, it would ruin whatever it hit.
Under normal operation, the more delicate components will eventually need to be replaced. Particularly sensitive are the silicon tracking chips. These are basically electronic chips on circuit boards that can tell you precisely where a particle went through. After a while, the silicon wafer gets too beaten up to work well.
Neutron activation is a problem, although that’s more of an issue for so-called “fixed-target” experiments. A fixed-target experiment is one where a particle beam is directed at, well, a fixed target of something (say, carbon) with a detector downstream to do physics with the produced particles. Such targets can become brittle from radiation damage, and in many cases cannot be approached by people because they are too radioactively hot.