Journal 3.0 Combine Functions and Analyze in Embodiment
Dizzy’s Journal Entry (A Broken Arm Wins!!!)
While Eve and I have been working on the embodiment for Function #2 (collect sticks), the “arm team” has been making CRAZY discoveries. I’m so glad they did analysis before committing to a design. Here’s the story…
Vex was working on Function #4 (Reorient sticks for AON insertion) and 4-hour was working on Function #6 (Reposition sticks for AON insertion)…We were all sure that they would just wind up building a simple arm and putting a servo on the end of it to rotate the stick before inserting it. But, coach insisted that 4-hour brainstorm the arm and do some analysis.
So, 4-hour finds this freeware to do physics modeling and he makes a basic kinetic model of the arm. Now, when we choose a final design in this scale model, we’ll know just how long to make all the parts of the arm and the best places to put the joints. Turns out, it is way better if you add a couple linkages to the arm (I’ve got to research “4-bar linkages” more…they are awesome!!!)
[download and install Physion to create quick 2D models and see how they work]
4-hour’s scale model of Gatekeeper stage 1 in Physion:
Right off the bat, we learned some important things like how a four-bar linkage system allows us to customize the path of the manipulator. It is weird how the path completely changes as you change the length of one of the linkages or move pivot points around just a little. We also learned that grabbing the stick at the base and allowing gravity to do its work, may save us from needing a motor to rotate the stick. (Way to Go, Vex…you managed to fill Function #4 without a motor!)
But, we also learned that inserting the stick without a motor to rotate it is going to rely on coordinated efforts between the spotter and the driver because the stick jambs in the pipe if the spotter doesn’t move the trolley (the green box). Once again, Coach insisted that we do some research. We found that real-world robotics engineers deal with this peg jamming issue all the time. They use a “Remote Center of Compliance” to fix the problem.
Basically, all we have to do is BREAK THE ARM…well, the wrist actually. When Vex broke the arm in the model and added stiff springs at the wrist, the stick could be inserted into the pipe without any coordination by the spotter (i.e. the green box can be anchored and it still works).
We might not go with this broken-arm design, but it is cool the way engineering analysis allowed us to combine two functions (at no cost) and provide us another option. My guess is that anytime we can reduce the required driver-spotter coordination, we are better off. Even if we mount a servo at the wrist, I’m guessing we should build some compliance into the wrist.
I can’t wait to start prototyping our top choices for each function.
This article demonstrates how simulations can help evolve embodiment ideas.