I had one of those ideas you wake up with and I had to share.
Consider a “hook” for this year’s game which looks something like a segment of a large radius pulley, or perhaps this conduit bender, but large enough to engage the rung and without a stop on the end.
Mount it at the top of a pole (or two) so that it doesn’t rotate relative to your robot’s body. Then set up a nominally balancing climb with your alliance partners. Oops, our alliance partner got too close to the middle, so they’re going up and we’re going down. Then, as the rung tilts, our point of engagement with the rung also moves towards the center, offsetting part of the error! This type of hook should make any tilted hanging configuration closer to level. The larger the radius of curvature, the better the effect.
Added: The robot will swing to keep its CoG under the engagement point, of course.
In the top image, the torques from the two gravity forces cancel each other out, so the switch is balanced. In the bottom image, robot 1 moves to the right, which unbalances the switch. The switch tips clockwise, making the curved “hook” of robot 2 also roll to the right. So now robot 2’s weight is acting with an even smaller lever arm, meaning it balances even less of robot 1’s weight. The same thing happens if robot 1 moves to the left; robot 2 would also roll to the left, making the switch even more unbalanced than it normally would be.
As the switch pivoted clockwise and you had the rigidly mounted curved hook on the top of your robot (assuming it does not slide on the bar), your robot would “roll” uphill rather than downhill as the angle of the robot relative to the bar changed.
The main problem is that you’re assuming d2*<d2, but what @GeeTwo is saying is that the opposite happens. It rolls uphill. With a big enough radius you could roll very far on the bar to counteract the action of your alliance member.
I’m confused then on what would cause the curve to roll uphill. If the curve is rolling without slipping on the bar then shouldn’t it roll down the bar?
Because the COG is below the arm, in your layout the arm will rotate CCW relative to the generator switch. In the reference frame of the world it doesn’t rotate at all, but when the generator switch rotates CW, the arm rotates CCW relative to it as long as there is a grippy enough material on it.
The concept is correct. Physics is weird! Love it.
The trick is whether you have clearance to roll very far before something on your robot hits the square rung above the round rung, and you effectively become rigid to the bar, but with your CG not as rotated as you otherwise would have had if the original mounting was rigid
I think there is a larger selfbalance effect associated with simply hanging rigid to the bar with a relatively small amount of climb/long lever arm to the robot cg than trying to implement this, but I haven’t done the physics in long form to prove it.
Thinking on this: Yes, hanging rigid to the bar is basically equivalent to the radius of curvature being infinite (the pulley shape becomes a half pipe) or just having a simple hook at each end. The curvature allows this transfer of weight to be more fluid and doesn’t require an active mechanism at rung level.
Hmm…now that I think about it, I think there’s an even easier mechanism that does the same thing.
Make an elevator that’s the width of the robot with two hooks, one on each side. If the bar is level, both hooks touch at the same time and the robot lifts straight. If the bar is tilted (let’s say your on the left side and the bar is higher on your end) then the left hook will touch first and carry all the weight of the robot. Once the robot gets off the ground, it will swing to the left so that its CoM is under the left hook. That moves the force further away from the pivot, which increases the torque on the switch. The same things happens in reverse if you’re climbing on the lower side: the right hook holds the weight of the robot and the CoM swings under it to the right, moving the force closer to the center and decreasing the torque on the switch.
The problem of getting two hooks to to land on the bar is that each hook will be at different points on the X, Y and Z axis in relation to your robot or a deployed mechanism to the relationship of the bar if the bar is at an angle (another bot has pulled the bar down on the opposite side) You might solve one but the other two will not line up without some creative work.
On a fixed bar this would happen. On a bar that pivots the original point would move to a balanced point between the two robots. Which may or may not get you to the red line in your diagram based on weights of the robots.
I think when possible, this is the way, and you’ll see it a whole bunch in elims. Coordinating climbs this year will definitely take a lot more time/finesse than in previous years.
Agreed. Smart teams may pause at this level (unless at the buzzer) even on solo climbs to help their scouting reports to team captains who want to balance (or at least hang) three robots in elims. I estimate that this will be roughly equal to all of them.
