Why is it that a demo that works perfectly well when there are no students in the room will spontaneously derail when performed for students? I had a reasonably simple demo set up for demonstrating the Impulse-Momentum Theorem. I had a force sensor, a cart on a level surface, and a motion sensor. I hit the cart with the force sensor, and the motion detector tracks its motion. My computer calculates velocities and displays everything, real-time on the SMARTboard. It worked beautifully when I practiced it yesterday afternoon and this morning before the students showed up.
Once the students showed up, however, it stopped working properly. I thought perhaps that the motion detector was getting cranky about keeping too-close quarters with the cart, so between my first and second classes, I rearranged my classroom and set up the demo on the lab table I've neglected all year. This involved untangling all of the cables for my computer, the SMARTboard, the printer, and the network and USB hubs and then putting the system together again. As you might imagine, when the second class arrived, it wasn't anywhere near perfect. I could show them that the bump in the force graph corresponded to the (relatively abrupt) change in the velocity graph, but the impulse just would not agree with the change in momentum.
I glossed over that, spent some time reworking the system when I could have been at lunch, and thought I had it settled by the time my third class walked in. No luck. The impulse still wouldn't agree with the change in momentum. We talked about the various features of the velocity graph (abrupt acceleration from the push, slow deceleration due to friction, and abrupt deceleration when the cart was stopped to prevent it from crashing into the motion detector). We thought perhaps the force sensor was off, but it accurately weighed both a 100g and 1kg mass. My students left without me having successfully done the demo.
I continued working on it during the period between my third and fourth classes, but now not only was the impulse not numerically agreeing with the change in momentum, they weren't even lining up properly in time, according to the graphs.
Rather than continuing to frustrate myself trying to figure it out, I worked problems with my fourth class. I hope I'll be able to figure it out before I do the collision demo for them next week...