We mostly worked on the indexer. Increasing angle of panel improved flow speed a fair bit. We also decided to move the elevation belts from parallel to sides of bot to in front and just behind v-exit. This reduced the chance of a ball jam halfway up due to it being too far forward and simplified ball exit flow from the v. Also saved some vertical depth. Balls move up and down two belts pretty well, but the current prototype needs to be improved before we can properly test that attached to the v. Some quick tests with 2-in mecanum and compliance wheels on Gary made it pretty clear that a single 4-in roller easily outperforms the 2-in. We’ll still probably go with a double roller, the current sketch makes that look really easy and it means direct pickup from high loading bay is possible, which will improve our intake speed. We’re now considering developing an elevator as our climb, after looking at how many subfunctions Richard needs to have. No work was done directly on the shooter today, but running a belt right into the bottom of it for feeding gave promising results. It was a really sketchy prototype, but it worked well enough to suggest that’s what we’ll do in the final design. Nearly all the prototypes are at the point where we can stop building and start doing semi-final cad.
January 9, 2020
We worked on the intake, we determined that the best conditions were 8 inches from bumper to wheel, 6.25 from ground to bottom of wheel. The intake reliably sucked up the ball by pushing over the bumper. Improvements would be to increase speed, which shouldn’t be an issue with the actual bot as there won’t be a piece of wood intercepting high speed intakes. We also worked on the indexing system. We had a V shape indexer to get from the intake to a single file line we found 18 degrees with a 12 inch gap very effective with belts moving in opposite directions. Lastly we worked on our indexing in a straight line, we found 6 inches from inner belt to inner belt very effective. The indexer successfully moved balls uphill with only one belt moving. Next step is to try to see if the belts can take balls up a piece of curved polycarbonate.
January 8, 2020
Today we mainly focused on the shooter. We increased the compression and moved the rails outward to provide better guiding. We then did a precision test from about 18 ft back from a wall shooting to a target on wall 10ft high. We were able to hit a box about the size of the inner goal (12") consistently. Over time the prototype would drift to the left slightly but stay consistent. We believe it's probably due to some issue with the prototype construction. So far, our tests have answered most of our questions. The remaining one is to see if we have the range to shoot from behind the dj station consistently. So far it definitely seems doable but we probably need more RPMs to do so. We also started an indexer (conveyor belt) prototype and are waiting on printing more hubs for some stealth wheels.
January 7, 2020
We started some prototypes! Check out the videos on our social media. Over-the-bumper intaking will work, about 13/16" compression on the initial pickup worked well. Richard (the climb idea) was not as stiff as we had hoped, but it did work. We'll need to look at alternatives from the igus cable chain we have now. Indexer discussion continued, though without multiple power cells we can't move forward. Shooter testing with a wood model went more horizontal than vertical, but was fast. (I didn't catch how accurate it was)
Three prototypes within the first three days of meetings is way more than we've ever managed done before, so good work team!
January 6, 2020
Today, we built on the priority list and strategy discussion we started at Kickoff and we finalized our priority list. We started some rough design work and mostly settled on a hooded shooter. We also came up with a climb idea, involving doing the up step with cable chain and the down with something else.