[math-fun] Gravity Simulators (centrifuges)
I'm sure that NASA must have looked into all variations of this problem, but I haven't been able to find anything. Suppose that I wanted to build a centrifuge that incorporated a very active load (heavy mass that moves around a lot at the end of the centrifuge arm. The problem is that I would also have to keep some sort of counterweight also moving around in order to keep the arm statically and dynamically balanced, in order to not put huge loads on the axle & bearings. The most obvious solution (to me) would be to incorporate some sort of microprocessor and motor system to actively move the counterweight(s) to reduce the vibration and loads on the axles and bearings. But I was also wondering if there were a completely passive solution. I thought of putting the dynamic load inside of a rigid sealed box, and then suspending the box in a fluid having the same (average) density as the box. Presumably, the box wouldn't be able to move very far in any direction within the fluid. Can anyone think of any other "passive" type solutions? Can anyone point me to any literature references? Thanks.
reminds me of a zero-gravity experiment in sedimentation, where a slow rotation of a bottle proved to have equivalent effect as suspension in zero-gravity. Just make the driving force sinusoidal with average zero. nice. W. ----- Original Message ----- From: "Henry Baker" <hbaker1@pipeline.com> To: <math-fun@mailman.xmission.com> Sent: Monday, October 25, 2004 8:07 PM Subject: [math-fun] Gravity Simulators (centrifuges) I'm sure that NASA must have looked into all variations of this problem, but I haven't been able to find anything. Suppose that I wanted to build a centrifuge that incorporated a very active load (heavy mass that moves around a lot at the end of the centrifuge arm. The problem is that I would also have to keep some sort of counterweight also moving around in order to keep the arm statically and dynamically balanced, in order to not put huge loads on the axle & bearings. The most obvious solution (to me) would be to incorporate some sort of microprocessor and motor system to actively move the counterweight(s) to reduce the vibration and loads on the axles and bearings. But I was also wondering if there were a completely passive solution. I thought of putting the dynamic load inside of a rigid sealed box, and then suspending the box in a fluid having the same (average) density as the box. Presumably, the box wouldn't be able to move very far in any direction within the fluid. Can anyone think of any other "passive" type solutions? Can anyone point me to any literature references? Thanks. _______________________________________________ math-fun mailing list math-fun@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun
participants (2)
-
Henry Baker -
wouter meeussen