We started disassembling our 2024 half-field, and constructing the coral station and reef. Neither will be completed this week (potential snow this week cancelled school), so we’ll have more field photos ( and prototypes) next week.
We redesigned to prioritize consistent, fast L4 scoring using knowledge from Ri3D teams. This design uses a 1 stage elevator with a wrist to score coral at all four levels.
L1, L2, and source intake are achieved without using the elevator, and L3 and L4 use the elevator for quick scoring (Reef algae hasn’t been sketched yet, but should be removable with just wrist).
The current idea is to use a 1/4” plate with WCP inline bearing blocks for the stage.
We’d also have room for a ground intake (maybe post about that tomorrow).
The design’s seeming simplicity does have a few downsides:
We also lost a lot of knowledge and experience from last season, so there are feasibility concerns.
However, we did run a plate-based carriage last season and have designed the robot to be easier to program than the past. The REV Starter Bot also proved a lot of the design concepts for us.
We also have our previous 2dof arm design, and imagined a rack and pinion end effector capable of reaching L4 without a wrist.
We’re still designing it, thanks to the help from alumni who have prevented us from making the following mistakes (and provided these solutions):
We’ll have more updates tomorrow once the elevator file is started. This design looks awesome so far.
We’ve put a lot of work into developing our design. Here’s where we’re at:
Our elevator is mostly complete, with the base and first stage designed. We are still developing the torque transfer, which has only been sketched out.
Our elevator is extended by a 16t sprocket and 25 chain rack, similar to 3005 in 2023.
We added a second set of rollers to the end effector, allowing our robot to score on both sides (and reach L4).
We are planning to use 1/5” Maxcomposite and dual-sided belt to lessen the weight for faster elevator and wrist extension.
A dedicated reef algae intake would not be difficult to add to this design, and a slapdown algae intake would also expand our abilities. While algae doesn’t seem worthwhile to us yet, faster algae removal and the ability to score algae into the processor could assist when the algae wars begin.
We’re trying to have the preliminary design done over the next two days, then prototype across the next week and finalize geometry. We want this robot done as soon as possible to let the programmers cook.
There has been plenty of emphasis on coral passthrough by teams so far, and I’m not sure if it’s worth the added complexity.
Here are the cycle ranges for two coral-oriented robots splitting the reef by the midline between them. Both of these robots have passthrough coral intakes, and their cycles would be linear (find faster route between station and reef).
While this is efficient, if you give the red robot a uni-sided intake, the cycle routes become closer to this:
In this cycle map, the red robot prioritizes the reefs farther away from the alliance, while the blue robot prioritizes the reefs closer to the alliance. This route has several benefits for both playoffs and qualifications:
This path map allows blue to focus on scoring as quickly as possible, thus taking more defense, while red goes for higher value, longer cycles.
While at champs and on Einstein the first map is more effective, for lower alliances and at less-stacked events the second map can also put up significant points and avoid heavy, 2019-style defense.
What’s the reasoning for doing the bearing block stack on a sheet metal stage vs doing a more standard 2*1 inner rail that allows for bearings at the top of the stationary rails?
It seems like this design loses a lot of stability when the elevator is at the bottom of its range of motion, since the bearing spacing doesn’t extend like ti does in the traditional setup.
Also only 4 inches of bearing space is pretty small for such a tall elevator, we’d probably want to keep a larger 6"+ overlap even on a traditional elevator when fully extended.
Have you done or tested something like this before?
The 4 bearing blocks were mainly a CAD error; I did a linear pattern with 4 to add separation between the top and bottom but forgot to make the middle holes construction circles.
The original idea behind the plate-based elevator was to have it lighter so our center of gravity would still be moveable to the center. However, it ended up being significantly heavier than box tube, so we we’re in the process of switching to 2x1s for the elevator.
If we were using plate for our elevator, we’d add tapped holes with bearings on the side of the elevator while it’s stowed.
Yes, this is a concern I have as well. A potential solution we considered was using wcp plate-based bearing blocks on the sides at the top, which would add the extra stability. But again, just switching the 2x1 would solve this issue much more easily.
No, but if we were using plate we’d prototype it this upcoming week.
Our 2024 robot’s carriage did use 1/4" alu plate, but that was mainly to save horizontal space.
Last night, we went through the same process and realized there is no advantage to plates over boxtube. We will hopefully have a design update later tonight with some modifications.
