Back when I was doing sciece fair projects it was more of an experiment vs. explanation of how things work.
So, when I read the first part of your post I had in mind taking out the AR with various types of grain weights, differnet bullets (hollowpoint/jacketed/lead/ect,) and seeing which shoot the best (tightest groups, best/flatest trajectory) depending on a given a barrel twist or grains of powder. Do this using the scientific method (background research/observation, hypothesis, test, conclusion).
Any number of other options, such as semi-auto vs. single bolt action. Test which is more accurate (this plays to "
were thinking something along the lines of detailing how the AR15 works internally, comparing the action of the AR to more traditional firearms (bolt, lever, etc) or something in this area". Another option, does a 55 grain bullet .223 vs. 67 grain bullet with the same powder load give "better" baslistics, likely you find the 67 trajectory drops faster but maybe you get better groups at 100 yards.
Key to this being minimiuze the number of variables per trial to one. For example, use the "exact "same shooting position every time so your at the same level and inclination every time. That sort of thing. For example, if you do the 55 vs 67 as described be sure both rounds are the same (i.e. hollowpoint/nosler/fmj/ect/), wind is consistent/ clean barrel and same way of shooting (slow squeeze no jumping). For example shoot the first three 55 grains with a completely clean barrel with say 20 seconds between shots. Then thouroughly clean the barrel and repeat, three rounds 20 seconds apart.
Anyway, I'm sure this is more than you wanted, but I felt like I had to give some examples of what I was talking about.
Hope this is useful
GL.
have