Hello Everyone!
Sorry for the late week 3 post, but due to exams and a snow day, This week was a bit less productive than the other but still fruitful nonetheless. Detailed below is our third week of build season, as well as a detailed breakdown of our competition design direction.
Overall Architecture
Inspired by 95’s Intake, and 4481’s “shamper” mechanism, our competition design direction differs from our alpha bot direction in the way that it uses a cannon-style shooting methodology instead of a backhanded Ri3D Quokkas-style shooting like what our alpha design was supposed to do.
This new method allows for theoretically faster auto routines due to its minimized handoff time and less actions on the arm to shoot. One of our main priorities was to optimize the bot for shooting as much as possible as well.
It also allows for our note to be stowed more securely, making feed more consistent and makes it less likely for the note to be dislodged off the bot during field cycles, collisions and general arm movements.
As you can see below, our arm intake and near speaker shoot positions are the same, this is so we eliminate a large aim movement that can speed up the first 4 autonomous notes shot.
This new method also shoots lower at the same angles for speaker and wing-line shots, allowing us to have a slightly larger margin of error when it comes to wing shots.
Robot in Amp Scoring mode
This has yet to be tested but it is supposed to be a shamper based mechanism meaning that we should have little difficulty in scoring.
Robot in shooter stow, and climb mode
With the shooter stored and the climbing mechanism extended, we should be able to climb on to the chain at endgame
Robot in intake and near speaker shot angle mode
The shooter angles towards the speaker and prepares to launch a note.
Bot in multiple modes in a field CAD
Visual overview of the versatility of the robot design and all of the actions that can be completed.
Physical Trap Mechanism (and Abandoning it)
Another big architecture attribute that we decided was to abandon a physical trap mechanism. We decided that taking on this endeavor is going to take way too much time that could be dedicated to other subsystems. Also, the fact that we do not have the space to build a partial stage to test it reliably with.
Instead, seeing the viability of shooting into the trap by teams such as 5406, we are betting on being able to shoot into the trap instead. This was a big gamble that the team decided to be worth it considering time, and resources.
Our direction for the trap:
- If shooting trap works well, YAY!
- If it does not, well too bad but its fine, we are still able to be competitive without it
Comparison to Week 1 Priority List
Overall the current architecture lines up with our original week 1 priority list. We seem to have followed our picks pretty well and are happy with the capabilities that our final design has.
Alpha Bot
Due to the architecture change, we felt that it was not worth it to pursue the alpha bot to a fully operational state, given the incompatibility and differences from the comp design.
It will be continued to be used in its existing state for advanced intake testing and preliminary driver testing though, as well as testing auto pathing. It was a good experience to be able to have something solid for testing and is something that our team will discuss about continuing with for future build seasons.
Deciding the Architecture
The overall architecture decision was unfortunately decided by external data and insights from other open alliance teams’ blogs versus our own testing data. We were supposed to test the feasibility of the alpha architecture, as well as this new one last week, but a snow day plus exams put a big damper to that decision which delayed that to the end of this week. This is very not ideal and will put our build season behind schedule so, in order move ahead with more certainty, we decided to go with this direction instead.
Fortunately testing that commenced right after exams has confirmed that the direction we chose works
Competition Robot Design Overview
We have decided from the start of the season that we are going to leverage existing knowledge and techniques from the past year into this robot, seeing the potential for a lot of commonality and easily transferrable mechanism and subsystem design methods that worked for us then, unlike in previous years.
Modularity is also a big factor that we wanted to continue, all of our subsystems are designed to be easily detachable mechanically using bolts from each other for easy iteration, as well as maintenance. If something needs to be fixed or replaced during an event, we want to be able to either quickly diagnose the issue or replace any broken parts.
Chassis
Our chassis design is pretty straightforward and similar to last years. It’s a 27 x 27-inch SDS MK4i based swerve chassis.
The decision to do 27-inch square chassis over something smaller was simply easier packaging. Compared to something larger like our 2023 28 x 28 drive base size, it would have given us suboptimal space to integrate an under-the-bumper intake.
Material wise we chose a solid 0.120 aluminum belly pan to hopefully help reduce our CoG.
One big change for this year is regarding the implementation of a brain pan. This allows us to put half of our chassis electronics into the bottom facing the ground, allowing a larger and more accessible space for wire management and maintenance.
Covering the bottom is a metal pan lined with a thin sheet of polycarbonate. This was chosen due to its cost effectiveness as it is being made by our sponsor, and in result, us not needing to get an appropriately thick (and conversely expensive) polycarb sheet of that size. It also allows for easier climber mechanism installation/removal for servicing.
Superstructure
The base is a 1x1 bar sitting on top of the chassis, only secured by 4 bolts. It hosts the same construction and mounting methodology as our 2022 and 2023 superstructures. This allows for faster dismount and installation of the upper structure., and has been proven to work for us for very well.
Our arm gearbox resides on the superstructure as well. The gearing setup is exactly the same as our 2023 robot, 60:15 final chain stage, then a 24:84 stage to the initial 4:1 maxplanetary gearing. Outsourcing the what would be 2nd maxplanetary stage to a spur gear reduction was ideal due to packaging, as well as protecting the planetaries further from sudden loads.
A change we have implemented was to add a cam-based tensioning mechanism. Our 2023 chain mechanism did not have it and we got very lucky that one of the holes happened to allow for a tensioner shaft and bearing, allowing it to tension just perfectly.
This is not exactly replicable for future years and something we have to keep in mind.
Arm
The arm is very simple, not much to talk about. It houses 2 60t sprockets that interface to our arm gearbox and holds mounting provisions for our shooter mechanism.
The arm crossbar is held by WCP tube blocks to the arm side rails which is the first time that we are trying such a mounting method for one of our bots. The pulley on the arm is for an absolute encoder that resides in the chassis. it also acts as a dead axle shaft stopper for the arm.
Intake
The intake is a very slow-to-intake subsystem for us to get up to speed and into the condition we want it to be for comp. We expect to have a lot of intake geometry and wheel type modifications over the coming weeks. The current design employs 3 rollers, one of which is a dead axle powered guide roller lined with grip tape, while the other two are custom urethane rollers we call “petunia wheels”.
IT’s a 95-style under the bumper extension intake. It is not really under the chassis but just simply a frame perimeter intake.
It’s the simplest theoretically and most robust for us to implement due to its protection given by the bumpers, as well as its modularity potential. It mounts to our MK4i’s mounting provisions and the frame is decoupled from the actual intake plates for easy iteration.
The Internal Gusset construction for the frame (inspired by 1114’s 2019 elevator gussets) give us a strong platform to mount the intake to the frame given the geometry limitations posed by the MK4i Modules.
We also had a minor design requirement that necessitated this mounting solution as we wanted to keep our 2910-style bumper mounting system, and that meant no protrusions or plates outside the physical frame itself. Since this is partly uncertain and new for us, we will subject it to more durability testing when it gets built.
Indexer/Shooter
The shooter uses a horizontal shooter architecture with 0.5 inches vertical and 2 inches horizontal compression. It has been demonstrated to work really well both from our own internal prototyping as well as the testing done by OA teams. Each roller is powered by NEOs on a 1:1 ratio
Our wheels for the shooter are 4 3x3 inch lightweight custom-cast urethane wheels for the flywheels. The Indexer is composed of 30 2 inch urethane compliant wheels and compresses the wheels by about 0.25 inches. This is powered by a single NEO, currently shown as 1:1 but may change to a 3:1 or 4:1 with maxplanetaries when we do come to assemble it.
The Indexer floor is composed of 3/8 inch shafts on bearings. We felt that making the floor as frictionless as much as possible was important for feed, and this worked for us.
The shooter sides are lined with a thin sheet of hdpe to reduce friction and shield the note from the metal pockets. However this may not be the case if we decide to make it out of polycarbonate instead.
While the top and bottom are currently separately driven, we might and have already designed provisions for a mechanically linked top and bottom flywheel rollers in a bid to mechanically match their surface speeds.
The spacing between the flywheels and last indexer shaft allows for a shamper redirect of the note for amp scoring.
also it kinda looks like a shark
Climber
We initially wanted the climber to be integrated to the arm, like WCP’s CC robot for 2024, and RI3D Quokkas. However that proved to be incompatible without some major sacrifices to our shooter arm mechanism, due to our existing geometry being barely over 28 inches at max arm angle to boot. So instead we opted to use TTB 1-stage climbing blocks.
The gearing is not yet final but a fixed 72:20 3rd stage and 2 Stage MAXplanetaries make it versatile enough for us to several ratio combinations to test. probably looking at around 70:1 to 90:1 per gearbox.
The climber is placed at the approximate center of the robot, and the geometry allows for us to climb either facing forwards or backwards giving us versatility when it comes time to start game strategy.
Thank you so much for reading our bot breakdown summary. We hope to have been able to provide insight into our design choices and our current design path. If we make any changes or come across any other information when creating the design we will be sure to share the progress. Please be sure to check out any updates we release as we get our parts from our sponsors and begin the assembly of the final design.