DiscoBots 2587 - Build Blog 2024

[lynca]

Discobots are located in Houston, Texas. Since 2008, we have participated in various Competitive Robotics platforms including FRC. Our team is a melding of curious minds from all across the city of Houston, with significant participation from the following Houston ISD schools (Lamar, Westside, Carnegie Vanguard & DeBakey):

As part of the Open Alliance, our CAD, Github, and some additional resources will be public. We have online meetings every Sunday that are consistently uploaded to Youtube once the season starts. We will also be posting regularly in the Open Alliance discord in addition to this thread.

Team Resources:

• Github
• CAD in OnShape (will be posted soon)

Social Media:

• Instagram
• Facebook
• Tiktok
• Twitter
• Website
• Youtube

Competitions:

• Week 4 - Houston
• Week 5 - Space City

[Sentinel]

1898 Tombstone (Written: December 28th, 2023)

Students from Westside decided to build an off-season robot during the summer with the intent to compete in the Texas Robotics Invitational and eventually Robot Remix. We based our robot on Team 6672 Fusion Corp robot from the Charged Up season.

We used many of the same dimensions but adapted the robot to use the REV MAXswerve modules. We had to completely change many parts of the robot because we were unable to process steel parts. Additionally, we took this opportunity to test out many REV products such as the MAXTube and MAXSpline systems. With it being the summer, we had major manpower constraints that caused one person (me) to CAD the whole robot with little to no supervision. This proved to be our downfall. I was overconfident in designing this robot, which caused major mechanical issues to arise during the build phase. After I finished the design, I had enough free time to build a Bill of Materials from which we ordered a lot of our parts. I also had time to produce technical drawings for all of the custom parts for our design.

When the parts started to arrive, we only had two weeks to build our robot. I was very impressed by our newer members for their dedication and ability to work under pressure. As we built this robot, more and more issues started to arise. Some of these issues included galling between the screws and nuts we used (both stainless), inability to push a MAXSpline shaft through 8 MAX Pattern holes, unsupported tower, belly pan, and intake, and a lack of chain tensioners. Although we worked diligently, we were unable to get our robot fully operational during TRI. It was kind of a disaster.

Although the robot didn’t work, we still got some value from this project. For me, this project showed the limitations in planning and the downfalls in my skills in CAD. As a team, this project allowed us to train our new members and test out the REV system. I’ve collected many of the lessons we learned in a document and I’ve recorded a video elaborating on them.

Although we intended to continue to develop the robot for Robot Remix, little progress had been made only a couple of weeks before the event, causing us to pull out.

Project Folder

• Lessons Learned
• Video
• Drawings
• CAD

[Sentinel]

Student Development (Written: November 3rd, 2023)

In past years, not as many of our newer members have been as engaged as we would like. Although some students push past the steep learning curve, many more get lost on the way, never having a chance to get the value out of FRC. Several veteran students at Westside believe that we lose a lot of these students because of the flexible yet chaotic nature of our training program (or lack thereof). To combat this, we have been building a more formal and structured curriculum to teach students how to build robots. We created these lessons with a few core ideas in mind:

1. Students learn and retain information best when they apply their skills
2. Projects and presentations are both useful tools to teach students, but each fulfills a different role in the learning process
3. Although informal lessons can be useful at times, structure is critical in holding the attention of the students since many get lost when an instructor is unclear.
4. The majority of students need some kind of 1-on-1 instruction at some point.
5. The end goal is for students should be mostly self-sufficient by the start of the season

We first create lessons by writing out what information we have to cover in our lesson plans then we move on to creating slides to cover that information. For each lesson, we try to include physical examples (old parts, tools, etc.) to show the students what we are talking about. Additionally, we also try to include a practical application of the skills at the end of the lesson, which usually consists of making a part or assembly.

Our first lesson was an overview of FRC and our team, so the new members can have an idea of what is happening. After our initial test, we started developing a connected series of lessons that covered the process of building a robot. The first lesson in this series covers drawings, marking, and cutting and is accompanied by a quick practical of cutting a piece of tube stock (it’s at the end of the presentation).

We showcased this lesson on October 26th, 2023, and it was a great success. Overall, we were very pleased with the engagement the students had with the lesson. Even with the structure, the small group we had (~15 students) still allowed us to have fun and joke around a little bit. The physical examples, some scrap parts we had lying around, were passed around and acted as the spark for several questions some students had. To do the practical exercise, we split up into two groups and made two parts. Students in each group switched roles in the marking and cutting process. Although a little more chaotic than I would have hoped, I was very pleased when students were able to recall what to do from the presentation.

As of the time I’m writing this post, we have not yet created the next lesson in this series, but it will cover screws, nuts, and rivets (we do lessons every other week).

For CAD, we are going to rely mostly on the self-paced Onshape Learning Pathways to teach the basics since our CAD team is at varying levels of proficiency and CAD is mostly done independently. As we approach the season, we’ll come together to review some team processes and then practice our skills as a team in the F4 CADathon.

As for programming, most team members who are interested in programming are either new to FRC as a whole or have minimal knowledge of the Java programming language. Due to this, we have taken a step back and started by introducing various electrical components and software. Since most of the programming team is completely new to FRC programming, we will be doing thorough lessons on basic WPILib knowledge as well as other software like REV Hardware Client, Shuffleboard, and Driver Station. In between lessons, the programming team will work on programming a base tank drive to help solidify knowledge.

Some might wonder why we bothered to make our own curriculum when there are plenty of ready-made lessons online. For example, Team 3847 Spectrum has an amazing curriculum complete with projects and libraries. We decided to make our own curriculum to better suit the members of our team and our production processes and equipment. For example, I’ve noticed that our students get bored when we don’t have something physical for them to do, which is why we implemented the practical exercise. Additionally, I haven’t found any curriculums that combined both a presentation and a practical exercise in one lesson. I find that presentations are better suited for explaining concepts while practical exercises allow students to apply these ideas. Additionally, building these lessons helps our veteran members better articulate and document many of the lessons they have learned over the years, something that we have to work on as a team.

As part of the Open Alliance, we have kept all the parts of these lessons public so that other teams can use them or take inspiration from them.

Lesson Plans:

• Lesson 1 - Drawings, Marking, Cutting

[Sentinel]

Westside Pre-season Retrospective: (Written: December 28th, 2023)

After the success of the first lesson in the series, we continued this program and created 3 additional lessons. Although we had hoped to include a practical for every lesson, we did not include any practicals for any other lessons. Thankfully, we did at least include many physical examples for our members. Almost all of our lessons started to be created only a day or two before we used them with most of them being finished 15 minutes before our meeting started. This is purely because of a lack of time management from our end and a lack of consistent effort to build this program. We forgot that our senior members who were developing this program were going to be drowned in college apps. To properly execute this program, I think we would have to plan out this curriculum over the summer to reduce the load on our veteran members who are teaching these lessons. Needless to say, I have even more respect for teachers than I did before.

Initially, we wanted to alternate between lesson days and build days, where we would continue to build a simple robot. However, after a couple of build days of absolute chaos and some panic of realizing that we only had a few meeting days left, we turned to doing lessons for every meeting to make it more manageable for us and increase information density. Unfortunately, this made a significant portion of our members disinterested as these lessons were less engaging than we had hoped.

As for programming, we are still lacking programmers like in past years. We had hoped to practice programming during the build days, but the chaotic nature of these days took a greater toll on our success than we had thought. Some progress was made in this area with our last lesson that covered electronics, but we still have fewer programmers with a good understanding of WPIlib than we would have hoped. We intend to use the first couple of weeks of the upcoming seasons to get programmers at least somewhat ready.

As for CAD, we were quite disappointed in the pre-season in this area. We initially assembled a group of students who were interested in CAD and let them go with an Onshape learning pathway. However, I haven’t seen any success or evidence of progress from any of our prospects in this area. We should have had more of an emphasis on this area and put more time and effort into developing this particular skill. Unfortunately, we were swamped with creating lessons, so I’m not sure we had any time to push into this area. Additionally, with the F4 CADathon not continuing this year, many of us will be ill-prepared for the season and end up relying on veteran members. We might do a collaborative CAD project in the coming week ahead of the season to practice working as a group.

From an administrative side of things from Westside, we have fallen short of what we wanted to do concerning organization. We wanted to be on top of attendance so that we would be able to have some idea of what information each member has gained through our lessons. However, this fell through after our second meeting or so. Additionally, we wanted to have greater organization within our lab, and although the arrival of our TStaks helped somewhat, it is still an absolute mess. Hopefully, we can use the first week to clean up a little bit.

Although it may not sound like it, I am quite happy with this pre-season overall. At least from Westside’s Point-of-view, this has been the most work that we have done before a season. I am impressed by the resilience and effectiveness of our newer members and look forward to working with them in the upcoming season.

Lesson Plans:

• Lesson 1 - Drawings, Marking, Cutting
• Lesson 2 - Rivets, Screws, and Nuts
• Lesson 3 - Introduction to Motion
• Electrical - Lesson 1

[lynca]

The students put together a few CAD ideas on design types.

Recordings of DiscoBots’ Sunday virtual meetings will be posted on the DiscoBots youtube.

[Hobob123]

The Trap Isn’t as Hard as it Seems

Thanks to the polar vortex thing, I’ve had plenty of time to marinate about the trap, and here’s what I think.

After much discussion internally and on Chief Delphi,

Stop Trying to Do Trap!

I run an educational design community, and I’ve designed a trap robot for this game and posted the progress here. It’s unbelievable how many teams I’ve talked to that are going for the trap as a “must have,” not knowing how ridiculously hard this objective is… for not that many points, and the opportunity cost is much greater. Compared to previous endgames, where you can reserve space (2022) (2019), cheesecake it onto a different mechanism (2018), or literally drive onto a platform. (2023), the…

it seems like there’s a consensus that the trap is a lot of complexity for minimal point advantage.

However, I think the trap is almost necessary for an average team to have a chance at ranking high enough to be an alliance captain, especially since the ensemble ranking point requires three robots and a harmony/high note without any trap scores.

Additionally, I feel like designing for the trap can be easier than it seems if we approach it in the right manner.

Here’s a more detailed explanation of all of this:
[https://docs.google.com/document/d/1bQH09aFGUsv7RcPL7QeEbK8uWx__qE2WUDQJ4uzm2TQ/edit]

For this rendition of our robot cad, we decided to use a winch + measuring tape climber. The benefit of this was that it is decently simple, but can also pull the robot all the way to its bumpers without requiring extra design work.

The winch mechanism also helps create a lower center of gravity and balance out the battery, allowing the robot to be more stable when climbing.

Since running trap cycles is just not an efficient use of time, it seems reasonable to assume a team will only use their trap mechanism once in a match. In that case, it would also be easy for the trap mechanism to load directly from the human station.

In the document I talk about using surgical tubing or springs to unfold the mechanism, however, since it is just a single pivot, We’ve decided to just use a brushless motor to rotate the arm.

Here are more CAD pictures:

Onshape

Hope ya’ll enjoy - Sid

[Sentinel]

Week 1 Recap - Westside

First we had kickoff at Westside with the whole team. We got to know each other and brainstormed approaches to this year’s game, but we did not settle on any particular approach.

More Pictures

During our Sunday meeting, we showed some approaches we had to the game:

Westside finished building our MAX swerve drive so programming can get familiarized

Saturday Meeting at Lamar

We made a priority list on what we want on our robot/what we want to do:

1 - Drive

• REV Swerve
• Bumper Height varies based on other systems

2 - Shooter

• Flywheel shooter
• A shooter is the primary scoring aspect of this game and we suspect that we can gain a significant amount of points by simply participating in it.
• We were questioning if a vertical (axles parallel to the face of the note) or a horizontal (axles perpendicular to the face of the note) will be better for our shooter.
• Horizontal shooters can easily induce spin onto the note, but require a wider package in order to work. They also require a significant level of compression onto the note.
• Vertical shooters have a better package (less required width, which is a premium, in exchange for more vertical space). These shooters do not easily induce spin, but this can be mitigated by using opposing spiral grip tape patterns on the top and bottom.
• We can also add the number of wheels to increase the exit velocity of the note.
• We are leaning towards a vertical intake with the opposing grip tape pattern for our shooter because it provides better packaging. We will have to test a prototype to confirm however.

3 - Climb

• Minimum height of climb is ~2’-5"

A few options available:

• Climber in a box with telescoping tubes
  • Easy to use (little design needed, only build)
  • Susceptible to side hits and painful to repair in these cases
• Constant Force Linear Slides (Penn State)
  • Looks fragile
  • Easier to repair than telescoping tubes
  • Still requires elevator parts and constant force springs
• Pivot (Soumil)
  • Simpler than everything else
  • May be easier to integrate to another mechanism
  • Requires a lot of space on robot and more testing
• Reaction Bar + Cascade Lift
  • Requires parts for the lift
  • Dedicates a large part of the robot to just climbing
  • Opens the door for a trap possibility
  • Super stable climb
• Cascade Lift
  • Simpler than reaction bar
  • Dedicates a large part of the robot to just climbing
  • Super stable climb
  • Requires purchasing lift components
• Hook with Winch (Disco Stronghold Robot)
  • Prototypes already available at Lamar
  • Small package
  • May be unstable if done incorrectly
• We’re leaning towards the telescoping tubes and the winch because of their relatively compact horizontal area. We would like to see a prototype of the Soumil’s pivot idea and consider it depending on other systems of the robot.
• A significant part of our decision is based on part availability for elevator and the constant force springs

4 - Ground Intake

• We think the ground intake is very important because it allows us to capitalize on the extremely heavy autonomous points as well as the missed opposition source loads. Additionally, it will probably provide a faster alternative to a source load.

• Over Bumper intakes are much more common in FRC and have been proven by many RI3D teams. They also fit very well this year considering the lack of a bumper gap. As usual, the main concern of an over bumper intake is the susceptibility to damage; a hit from another robot or just by hitting the wall could make us useless for a match.

• Under Bumper intake is less common among RI3D teams. They are more protected because they are in the frame perimeter throughout a match. This makes them harder to package. Additionally, it is harder on the driver to know if they have intakes the note due to the limited visibility.

• A design that could alleviate the packaging problems with the under bumper intake is team 95 grasshoppers who put their intake in front of the swerve drive but still a part of the frame perimeter.

• We are leaning towards the over bumper intake but are exploring designs that involve team 95 grasshoppers’ idea.

5 - Strong Autonomous

• With our back-heavy schedule, we have put more importance into the programming team. This means that we plan on making design decisions that speeds up the manufacturing and assembly process to get a robot to programming quicker and subsystems that help them excel.
• Some of these include a vision system, aluminum extrusion–based construction as well as more COTS parts.

6 - Amp

• We think the amp will be overlooked because of the ability of the kitbot to shoot. However, the coopertition bonus and amplification mean that this is an important niche for us to fill.
• The main idea for scoring in the amp is making the note move downward

Options:

• Cranberry Alarm Note Reaction Bar
  • Solid concept and execution
  • No example robot that uses this idea
  • Allows for the shooter to be a fixed angle
  • Questions on reliability
• Minnesota Blooper/ramp
  • Solid concept, good reliability
  • Takes up lots of space
• Unqualified Quokkas arm
  • Solid concept, most reliable
  • Requires a flywheel that points downward and a way to get it there
• Soumil Push-in idea
  • Can be a reliable if done well
  • Unsure on heights
• We are leaning towards the arm idea, but we recognize that it would be a large allocation of resources within our robot.

7 - Computer Vision

• We think that this is an important part of having a strong autonomous rather than relying on dead-reckoning
• Our main goal is the ability to do a 3-note Auton

8.5 - Drive Under Chain

• Maximum Robot Height: 2 ft - 3 in
• The bottlenecks around the stages will make it extremely tough for a robot to complete cycles. By being able to go under the stage, we can be more flexible and effective.
• We don’t think this is too harsh of a constraint

8.5 - Harmony

• Requires ~ 8 in of extra climb height ⇒3 ft minimum climb height
• This is basically required for the ranking point

10 - Multiple Shooting Locations

• Being able to shoot from different distances allows us to be less reliant on the subwoofer area. This also makes us less susceptible to defense since we can also shoot from the protected zone.
• Additionally, this makes our auton cycles much faster.
• We can change the RPM of a flywheel to increase the exit velocity of a note
• We can also change the angle of the flywheel to change the launch angle of a note
• Another way of changing the launch angle is to have a blooper that diverts the note after it leaves the flywheel
• One concern of moving a whole flywheel is that it is less responsive from a programming side of things and requires more support in order to accomplish it. The blooper would fix this by lowering the moment of inertia of the moving section of the shooter.

11 - Source Load

• In order to make source loading useful, it has to be faster than ground loading. Our main concern is how long it will take to align with the source, so we think that we would need to be able to slam into the source in order to align.
• However, source loading could be used for certain mechanisms if it shows to be quicker or more convenient for it.

12 - Trap

• We largely think that the trap is a trap
• It is tough to do and designing a whole robot to do it is not worth it. However, its utility for the ranking point can make it a powerful tool.
• We are considering a single use mechanism to do it currently.
• Refer to the picture on the whiteboard

13 - Buddy Climb

• We don’t consider a buddy climb important at all because of the ranking weight that a trap would come with.
• We don’t think the time it would take to accomplish it would be very useful.
• Unqualified Quokkas have a viable design to do it.

We watched some videos of the various RI3D teams:

Videos

Bison Robotics - https://www.youtube.com/watch?v=vnc_ZvgAcJY

Penn State - https://www.youtube.com/watch?v=tpeo8EcOMhU

REV ION Starter Bot - https://www.youtube.com/watch?v=fm0M-4Gmf0c

Big Sky Robotics - https://www.youtube.com/watch?v=zllVfVgOKLA

Unqualified Quokkas - https://www.youtube.com/watch?v=vlUoh5eGic8

Minnesota Robotics - https://www.youtube.com/watch?v=CxoqA1WbW2M

Zoukeepers - https://www.youtube.com/watch?v=W3ITseRQoaU

After evaluating our options, we split into smaller design teams to come up with cad concepts of various robot types. Teams 1 and 2 were decided during the meeting and Team 3 is left for everyone else. The intent of this split was to explore the different approaches to this game in a more granular sense by going forward with CADing robots.

Teams and Robot Specifications:

Team 1: (Sid, Cedric, Peter, Vatsal, Dhriti)

• Under Bumper Intake
• Trap
• Winch
• Vertical Shooter
• Blooper
• Amp with blooper

Team 2: (Greg, Sebastian, Cem, Anthony, Michael)

• Over Bumper Intake
• Telescoping Climber
• Pivot Shooter
• Amp with Pivot Shooter
• Vertical Shooter

Team 3: (Lamar and everyone else)

• Over Bumper Intake
• Pivot Climber
• Horizontal Shooter
• Amp

That’s how we ended Week 1. More will come in the Week 2 Recap (Hopefully not 3 days late this time).

[Bean]

Week 2 Recap - Lamar, Debakey and more…

Based on the nature of this game, a well-functioning ground intake seems more and more important to a successful robot. As a result, we have spent a lot of time iterating ideas and evaluating strategies of a successful ground-intake. The main debates were:

• Over-bumper or under-bumper?
• Moving or stationary?
• Integrated with the indexer/shooter or separate?

To begin, we made a combined intake + indexer + shooter subsystem inspired by the Unqualified Quokkas Ri3D concept:

This design uses the first row of rollers to pick the note off the ground, the second rollers powered by a separate motor as an indexer, and the big, blue wheels as flywheels that launch the note.

Next, we refined this design to include motors, belt-spacing, and cut down on size. In the process of this development, we realized that the concept of mounting this integrated intake & shooter to an arm could be achieved both under, and over the bumper.

After presenting this design, building the prototypes, and evaluating it amongst the whole team, we were able to decide on the criterion of our ground intake debates:

• The intake should be under the bumper: the extra security of the intake will outweigh the extra reach or driver-visibility we’ll get otherwise.
• The intake will not move like the Quokkas’ design, instead, it will remain stationary in our robot
• The ground intake will be separated from the other subsystems and notes will be passed from one mechanism to another through handoffs.

Here was the initial design concept after those discussions:

This design is meant to attach to the side of our Maxswerve and behind our bumper. All 4 rollers would be powered by one motor to deliver the note from the floor into our belly-pan. The concept looked promising, but needed some changes to integrate into the drive base.

Resolving the clearance issues, we finally arrived at this refined design:

[Sentinel]

Week 2 Summary - Westside

We used the polar storm to get ahead on our CAD work, and we had a CAD meeting on Tuesday (1/16).

Tuesday CAD Meeting

Team 3 was not able to attend

Team 2 did not progress their design as fast as expected. They had several design concepts and geometries but no tangible robot. Their concept consisted of an over bumper intake mixed with an arm to shoot both at the speaker and the amp. This is similar to Unqualified Quokka’s design, but with a separated intake and shooter. However, this created several problems.

• The intake must handoff the note to the shooter
• A shooter + indexing assembly must fit within the robot and be able to both receive from the intake and be lifted for scoring into the amp. This was tough because it needs to be ~16-18” long in order to hold the note.
• A large arm assembly might interfere with a climbing assembly.

They were also utilizing a climber-in-a-box (AndyMark Telescoping Climber) to complete the climb but with no intent of going for the Trap.

Pictures

Onshape Link

Team 1 had a more complete robot with several major systems nearly completely designed. They had an under bumper intake that fed to a static shooter and a diverter to change the launch angle of the note. To score in the amp, they utilized a plate that prevents the note from sliding back out of the amp after it has been shot into it. A winch with tape measures to push up a hook is used for the climber, and the trap will be done using another source-loaded arm that shoots downwards into the trap.

We were quite happy with Team 1’s design despite some concerns:

• The tape measure mechanism was unnecessary since we have enough height to place a hook onto the chain. Team 1 was designing the tape measure system as if it was a cascading elevator system, which it is not.
• The movement of the winch through the middle of the indexer is quite concerning. It is likely to get jammed and may have problems with deployment. With the concerns regarding the winch and the tape measure, we will likely redesign this particular system.
• The accuracy of the diverter mechanism was put into question. These notes are already being compressed into shapes we are unable to fully predict, how consistent will diverting their movement work?
• How consistent is using the “hood” to score in the amp?
• The under bumper intake is a mechanism Disco has little experience with.
• The angle that the note has to move as its being intakes is extremely steep. Will this damage the note?

Overall, we are quite happy with our CAD progress and think that splitting into teams was a very valuable exercise. With Team 1’s trap design, we have more hope of doing the trap than we had before. The next step for the CAD team is to settle on a design. It will be most likely a mix of both concept robots.

Pictures

Onshape Link

Westside Magnet Showcase

Westside High School had a showcase and open house on Thursday (1/18). We took this opportunity to market our team to Freshmen and 8th graders from some of the local middle schools in HISD.

We made a new poster board to use at our booth. It includes a lot of pictures that we can reference while we are talking to people who are interested.

Poster

Additionally, we used a portable monitor to show people the current state of our CAD for this season.

(Pretend Mike Miles is our cad)

Finally, the programming team got our swerve drive moving, so we used that as our main attraction.

Overall, we think it was a great success. Slowed down progress on our robot though.

Saturday Meeting at Lamar + Sunday Online Meeting (1/20-1/21)

During our weekly Saturday meeting, we split up into different groups in order to accomplish several tasks.

CAD

The CAD team focused on completing the base of the robot. We decided on an under the bumper intake and took inspiration from Spectrum and various other teams in order to create our design.

Our design will have a normal square swerve drive that is sized to 24 x 24. An upper skirt made out of 1x1s is a rectangle that will define our frame perimeter, which will be different from the 24 x 24 swerve perimeter in order to accommodate our under-bumper. This rectangle is placed some distance above the drive rails (as of this meeting it is 2” above, but this was changed to 3.5” at a later date). This rectangle is defined as the “Upper Skirt” and will act as both structure and mounting positions for various systems.

We were considering stacking another 2x1 under the drive rails to lower our belly pan and push notes away. Instead of doing that, we are keeping the belly pan attached to the drive rails, and we are adding a skirt made out of polycarb that covers the outside perimeter to stop notes from getting under our robot. These polycarb plates constitute our “Lower Skirt.”

We decided upon an under-bumper intake for our robot after looking at some of spectrum’s designs. However, we had to design the structure of the robot differently because we are using MAXSwerve instead of MK4is. By placing our intake within the bumper, we do not have to worry about deployment and more safety would be afforded to the intake. However, we are still weary of this design because this is the first time Disco has ever done this type of intake.

Another aspect we were considering was reducing the front drive rail to a 1” height instead of 2” by using a 1x1 and stacking a smaller piece of it for the swerve module mounts. However, we found this unnecessary after assessing the height that the note can climb with our intake.

Build/Test

We also tested our intake and shooter. Build Week 2 1/21/24 - YouTube

Refer to Lamar’s Post above for more details

We also started building our competition swerve modules. (I have no pictures of these)

Week 3 recap is OTW

[Sentinel]

Week 3 Recap - Westside

Westside focused on completing the CAD for the drivetrain and intake. We knew that this point is where the structure for these systems will be finalized.

CAD

We had a CAD meeting on Wednesday to discuss the viability of our drivetrain and establish what progress needs to be made. Here are the notes from that meeting:

Wednesday CAD Meeting Notes:

Drivebase

• We currently have no issues with the 24x24 drive base. Subsystems are fitting in fine. We are looking for confirmation from Andy and other mentors on this sizing so we can proceed with cutting drive rails.
• We are unable to find reasons as to why we want this sizing in particular or why we don’t want it.
• The top skirt made out of 1x1 will likely be raised to provide more room for swerve maintenance while still being able to mount the bumper on it. This will also allow us to not have to cut gussets to fit into this spot.
• Electronic plates have been made, we just need to put electronics on them and mount them. We are considering cadding some of the more important wires.

Arm

• We are considering different methods to move the arm particularly in the axle.
• Vastal has a prototype based on Sid’s Sketches.

Intake Problems

• When the upper skirt is extended out to fit it, we interfere with the whole gearbox + block any line of sight between the pulleys, causing belts to not work. We think the gearbox has to be moved.
• The 2x1 being jut out one inch is unnecessary. We are already working with space as a premium and we don’t see a reason it can either be changed to a 1x1 and pushed back or kept a 2x1 and pushed back and up.
• The bottom roller should be pushed up 0.2" to fit the dimensions we had on Saturday.
• We think the whole intake assembly can be pushed in 1" and have the indexing rollers push up over the front drive rail. Currently the pulley interferes, but we can change that.
• The indexing rollers do not have to run the full width of the intake, consider using centering 3d-printed pieces that spectrum uses to make it center into smaller rollers.
• Please try to use standoffs instead of churros as spacing please. It’s less parts we have to manufacture.
• (minor) Also, the current design uses thunderhex bearings, but we have already bought a ton of normal hex bearings and will probably using those.

More Pictures

On Thursday (1/25), we finalized the dimensions of our drive base:

• Side to Side - 24”
• Front to Back - 31”
• Swerve modules are placed in a square formation of 24” x 24” at the back of the robot.
• The Upper Skirt (1x1s) stick out 7” forward past the swerve modules.

Here are some progress posts from our CAD channel:

Electronics (1/25):

Electronics are starting to come together. This plate would extend from bracing to bracing, with cut outs on the corners for gussets. In terms of the winch, would be tucked underneath the plate itself. The construction lines on the drawing show the “no-go” zones which represent an estimation as to where the wires would be and how much space they take up. Thinking of putting our breaker(on/off switch) directly on the plate slightly above the Rio.

Some other concerns:

• Would spark Max’s on a plate mounted vertically work?
• What other electronics could there be that are missing on the plate? (Network switch, LEDs, Rass pi, etc.)

Intake (1/26):

Intake based on Soumil’s design is ready for the test robot.

• The idea is that we can attach it to the westside swerve and use that to test driving and intaking for drives and autonomous routines.
• I can add another set of rollers above if we want an indexer, but one set of rollers makes it simple to manufacture and assemble.

Intake rollers

• These use a dead axle and 3d printed pulley hubs to attach to the tube.
• We can get away with 2 pulleys on bottom since we aren’t attaching other rollers.
• The Custom pulleys we are using mate to versaroller and a hex bearing. They spin on a thunderhex shaft that acts as a Dead Axle.

Intake

• Attaches to the 1x1 skirt and the drive rails.
• Uses 2 30t HTD pulleys with 75t belts that we just got from swyft.
• Two sets of unique plates One for motor mounting and one for the actual sideplate

More Pictures

Drivetrain (1/26)

Here’s some updates to the drivetrain:

• The 1x1 rail has been moved up to 3.5 inches above the drivetrain. The gussets have also been changed to L brackets.
• The side skirts are now multiple smaller pieces that can rivet or through bolt to the drive rails.
• Some minor updates to electronics layout.
• Dead axle pulleys seem to be good to go.
• Bellypan was removed to make space for other stuff bellow the robot
  
  

More Pictures

Cut List for Drive Base (Extrusion only)

2x1:

• 22" - 2x
• 17" - 4x
  1x1:
• 31" - 2x
• 22" - 2x
• 3.5" - 6x

Programming has been steadily progressing. They have been focused on getting telemetry working as well as simulating our drive base. The next step is to mock up and simulate some autonomous routines.

Pictures

During our Saturday meeting (1/27), we cut the extrusions of our drivebase and did a mock assembly using screws instead of rivets. Currently, we are using MAXTube extrusions but we may change them out for other material in the future.

CAD work continued between our Saturday and Sunday meetings

Rollers (1/28):

I have updated the deadaxle pulley part in onshape. This new version should be a simpler geometry, give us more clearance for through shafts, and be a little stronger.

It is also an 18t pulley instead of 16t which means we can probably run 60t belts on the intake.

More Pictures

Theoretically if you make the pulley even bigger, you can have it slide over the tube, getting the tube closer to the bearing.

Currently the outside of the bearing holder is flush with the versatube. Since our roller is really low to the ground I prioritized having that be flush rather than having the whole width of the intake. Perhaps for the indexer dead axles we can have larger pulleys and more tube length

Bill of Materials for the Rollers

Note: All belts/pulleys are 5mm HTD

• 4x 18t timing pulley - 1/2" hex bore
• 2x 28t 20dp gears - 1/2" hex bore
• 80t belt
• 2x 60t belt
• 45t belt
• 35t belt

More Pictures

First Time Everything Together (1/28):

Oh yea there is an arm too by the way…

We discussed some issues and concerns during our Sunday meeting (1/28), so the design has drastically changed since then. The only system that is finalized is the drive base dimensions. Our CAD Lead will be making a post of the changes soon.