I had a good discussion and some testing last night with some students, and wanted to share a bit of it with other teams. In 2010, I put a "white paper"* together that talked about CG Migration going over the bump.
I will try to put something tobether better, but in the mean-time, some points of discussion:
For a Robot that weighs 120 lbs, and has a Center of gravity at 6 inches above the ground, carrying a 50 lb stack of totes (roughly 6 totes and a recylcing container) can really move the CG up. For this stack, its CG is about 36" off of the bottom of the stack. This means that the net CG of a locked together 120 lb robot and 50 lb stack is:
(6"*120lbs=36"*50lbs)/170Lbs=14.8" or about 15 inches.
Now, assume this is centered over a wheel base of a traditional long robot (38") which really has a wheelbase of about 30" (38 - 1 inch of frame each end, - 6" for wheel diameters = 30"). In this scenario, your robot is likely to get tippy if you accelerate at 1 G (which most 4 CIM robots can approach).
A similar issue is illustrated by our friends at Ri3D team redacted:
http://www.chiefdelphi.com/forums/sh...hreadid=131925
Congrats on making a nice robot that plays the game so quickly.
Notice that even with 50 lbs of sand bags at the back of the robot, it is still pretty tippy, and the driver has to be very cautious. So, when designing, please keep center of gravity in mind in both the unloaded and loaded states.