Circular Motion Simulation Follow-up

I last posted about a new circular motion applet that I was planning on using with my classes as the quantitative part of their UCM paradigm lab. Some reflections:

• When students came up with a list of variables that might affect the size of the centripetal acceleration, the list was: speed, mass, radius (always in that order). The visual accelerometer on a rotating table showed the qualitative effect of speed nicely, and the thought experiment about driving a car around a corner (tight or wide) addressed radius. We couldn't do mass with the given stuff, so I told them to check that out in the applet. A few seconds' work with the slider showed that it's irrelevant.
• The applet is framed in terms of string length (radius) and rotational frequency - instead of speed. This means that students had to confront (and figure out) the relationship between rotational frequency (or period) and speed just to get their data for the acceleration's dependence on speed. I like that.
• The other way that they have to confront it is to control speed while investigating acceleration's dependence on radius - changing the radius but not the frequency would change the speed. The students have the figure out the proper frequency for each new value of the radius in order to keep the speed constant during the second experiment. I like this a lot as well.
• Students still have trouble reconciling their two models ( and ) to determine the complete function of v and r. Even when they have figured out the units of the two constants, the connection is hard for them to make. I'm very open to suggestions of ways to make this go more easily - I don't have a great handle on what the conceptual difficulty is for them here. In the second section, I framed those two models as "OK, so a is proportional to v-squared, and a is proportional to 1/r," and that may have helped.
• Overall, the quantitative modeling went much more quickly, had some good conceptual things to think about, and was good practice with function modeling, so I'm pretty happy about it, at this point. We'll see how things go over the next couple of weeks; did this begin to build lasting understanding?