FRC 6328 Mechanical Advantage 2020 Build Thread

Good answer

22 Likes

This thread has mentioned compression twice, once in the beginning stating 2 in. of compression and now 3.5. Now I had doubts that it was correct, but now that it has been said twice it clearly is. My main question is: does the compression delta refer to the diameter of the ball, or is it noting something else. From both looking at the first picture of the ball in the hood doesn’t look quite like 2 inches and when putting a board above a power cell and crushing it until the ruler said it had compressed two inches, the ball looked like it was completely crunched.

Could someone elaborate more about this? I understand that compression is one of the key variables in a consistent shooter, and I’ve read that 2 inches worked well in stronghold. But when doing my board test, I’m lost as to how even 1.5 inches would work.

thanks in advance, good luck teams.

7 H’s. No more. No less.

5 Likes

So before explaining something big to remember is that the wheel is only so wide, between 1.5in and 2in in this case, so when the ball is compressed the sides of the ball roll over the wheel like you can clearly see in the picture of the first shooter. So taking a board and smooshing a wheel down two inches isn’t a good representation of what it’s actually like when the ball is compressed in the shooter.

With our newest revision of the shooter we’ll have the ability to change the type of the wheel, motor speed, amount of compression. In our given case we’re planning on testing each combination of different options to see what yields the best results consistently. This is a relatively straightforward approach to testing but is time consuming so we’re making some assumptions. We know we need a 4in wheel for packaging reasons, so that cuts out a lot of testing, and we’re starting with at least an inch of compression based on Stronghold results. We’re also looking at what some of the best teams are using for wheels and trying to cut down the number of wheels to test. We’re left with a 4in colson, this urethane wheel, and this neoprene wheel.

Let me know if you have any other questions!

2 Likes

Dave, are you guys planning on testing the various durometers of the urethane wheel and neoprene wheel as well or just one of each?

Zach

1 Like

Just have plans to test the 60A/blue green wheel for 2477K37 and the 35A/black 2476k37 as those were the ones used by similar teams from previous years.

3 Likes

Wanted to put together a document that covers pretty much everything for our robot and will turn into our engineering handbook/technical guide, its pretty much just high level sections right now but will get more filled out as we progress this weekend and get all our thoughts together but feel free to check it out and add comments and stuff https://docs.google.com/document/d/1jASczFnH2G1aesc16QLurNLFoVJ9MQUTvbwSlOkT0Uw/edit?usp=sharing

3 Likes

Out of curiosity where did you guys buy the 6in colson you prototyped your shooter with?

http://www.aiecasters.com/wheels-performa-rubber.htm is our source

1 Like

1/13 END OF WEEK ONE

Busy end to week one with the team, lots of focus on building a prototyping some major mechanisms, and working on CAD for the drive train. The team has started putting together a document that will eventually turn into our Engineering Notebook, you can find that here. That is a living document, so as we work to develop our major strategies throughout the season, we will be updating this daily.

Strategy

Here is a major strategy points that we’re looking to hit this year that will drive our prototyping and robot design:

  1. Auto
  2. During auto, each team can hold 3 balls and needs to start on the initiation line. There are 5 additional balls placed in the TRENCH, and 5 balls placed inside the BOUNDARIES inside each ALLIANCE’S RENDEZVOUS POINTS.
  3. Initial breakdown states that for week one we will have three different auto options.
    1. Stay on the initiation line and shoot with our intake down with the option to collect the balls from our alliance partner
    2. Shoot 3 balls, collect balls in TRENCH, shoot 5 balls at the end of the trench
    3. Shoot 3 balls, collect balls at RENDEZVOUS POINTS, Shoot 5 balls back near initiation line
  4. Week 5 and beyond autos will be added to this document as the season progresses.
  5. Shooting
  6. The team has decided that in order to give ourselves the best possible opportunity to make as many shots as possible, we will work to position the robot up against as many, “hard stops” as possible. This means that the robot will be pushed up against something to ensure the robot is properly aligned before shooting so we can get a consistent shot.
  7. Three major positions
    1. Up against driver station wall in protected zone
    2. End of Trench still in protected zone
    3. INITIATION LINE
  8. Be able to get all 5 shots off with 3 seconds
  9. Aim for the 3pt goal, be happy with 2pts
  10. Hanging
  11. The team has decided that hanging is a must
  12. Balancing will be critical but initial analysis leads the team to believe that the robot should be able to balance the rig without an additional mechanism on the climber hook
  13. Have the ability to hang anywhere on the bar
  14. In order to secure a ranking point, putting a lot of effort into developing a buddy hang will be important to have early in the season, will likely matter less towards the end of the season as more teams gain the ability to climb
  15. Wheel of doom
  16. This is the lowest priority for the team
    1. Reasoning is the team is under the impression that 49 balls will not be scored before the end of the match, so that ranking point will be hard to get and the time can be used to do more important tasks

Prototypes

In order of priority that the team decided on last Saturday, the first thing we wanted to prototype was developing the most consistent shooter possible. After our tests last week, we discovered that compression of the ball and rigidity of the shooter hood, as well as motor speed greatly impact how far the ball goes and how consistently it hits it’s target.

After last weeks sketchy wooden prototype we decided it was time to develop something that we could gather some accurate data for. Below is the picture of the CAD of the shooter.

Some things that we were looking for when developing this prototype was the ability to easily change the compression on the ball, easily change the type of wheel, and easily change the angle of the shooter hood. With this we’re planning to conduct some data utilizing four different wheels and three different rates of compression to see exactly how each impacts the shooter.

For compression we’re looking at 1.5in, 2in, and 2.5in of compression, we developed this hood so we can easily unscrew the hood and bring it closer to the wheel. For wheels we will be testing four different wheels, 2476K37 in 35A and 60A durometer, and 2477k37 in 35A and 60A durometer. We’re planning to take 5 shots with each combination of each wheel and mapping the trajectory. We will be presenting this data by the end of the week.

Here are some videos and pictures of the shooter prototype test.

Ball Launcher Prototype Rev 4

Scouting App Development

Something that we believe is incredibly important and severely underrated is a teams ability to develop a way to scout during the competition and get reliable data on each of the teams so they can make an informed decision come alliance selection time. Here’s a quick write-up and demo from our scouting team.

“Over the summer and fall, the scouting systems team has been working to develop a new electronic scouting system. This is an evolution from previous years, where we used a “scantron” system on paper. The electronic system is designed to run on a set of kindles which connect to a laptop via Bluetooth. One of our goals in creating this new system was to make scouting easier. Instead of using more traditional input fields, we show the scout a map of the field with buttons for recording data, which we call “visual” scouting. This makes many functions much more intuitive, like selecting a starting position.

Directly following kickoff, we worked to select which data to collect via match and pit scouting. This year, we are working to take advantage of the new electronic system by tracking the time between power cell shots. This provides a quantitative measure of the team’s speed in addition to accuracy. After setting up these fields, we began to create the layout and functions of the “visual” scouting system, which is defined via JavaScript code to be loaded over Bluetooth. The “visual” layout is now fully functional, and we are happy with the results thus far. The next step is to begin testing with potential scouts in order to find any confusing aspects of the design. We are also in the process of planning our analysis systems to be built in Alteryx and Tableau. See below for a video of the scouting app in action:

Scouting App Demo

Vision Development

New for this year the team want to purchase one of the new LimeLight 2+ to try out our hand at a COTS vision system to make our robot as accurate as possible. Will have more information on the development.

Ball Intake and Serializing Development

One of the things that the team believes is going to be a huge issue if not handled correctly is serializing the balls once they’re inside the frame perimeter. Intaking up and over the bumper seems like a clear cut choice after our testing and we’re currently working on a 148 2019 style slide out intake, pictured below is the first rev of the CAD we’re using to work out geometry.

@shirish

After looking at some 2012 robots, the team decided to try a system similar to 254 but with rollers on both sides might be a good system to prototype. Below is the initial geometry 2D drawing to ensure we will be able to properly package the ball handling system. With this layout, the balls would intake through the back and then travel towards the front of the robot, up to the second level, then back towards the back of the robot and up to some kicker wheels and into the shooter.

There are a lot of potential pinch points with this system, so it makes the most sense to try to build the entire system to see in real life what hangups it has.

We’ll be laser cutting these panels this week at ETM Manufacturing and will be assembling for testing this weekend.

Drivetrain Development

Over the weekend the CAD team met to start work on the drivetrain. We sat down and started working on the major requirements which are listed below:

  • 32in long by 26in wide
  • Single Speed, around 13fps adjusted
  • 4 NEOs(actually only motor you ever need(or can get))
  • 6in Colson Wheels
  • WestCoastProducts flipped gearboxes to give us some more space in bellypan

After doing all the math and utilizing JesseK’s calculator, we were able to reach a final ratio of 11.5:1 which gives us around 12.2 fps and a quick sprint distance. Below is a screenshot of the calculator.

After getting this, Our CAD team got to work modifying the WestCoastProducts Flipped Single Speed gearbox to match what we drew up.

There are some planning to see if we can shrink the distance between the plates since we wont be utilizing that space to have our sprockets.

While doing this, a few other students started work on the sheet metal for the drivetrain, hoping to finish this up in the next couple of days so we can get our practice drivetrain metal cut and assembled.

Random Other Updates

Some other cool updates, we did some work on an auto-closing hanger hook. Below is a screenshot of the CAD and a video of it in action.

Hanger Video

This week the team received an incredible donation of 9(!!!) more 3D printers for the company FlashForge, these printers will used to educate our students on additive manufacturing. Thanks again, Josh!

Expensive Shelf

4481 inspired basic tornado style hopper we’re playing with as a backup plan

Moving Forward

In the next week, the hope is to decide on a direction for our ball serializing system and nail down the specifics of the shooter. Once those are complete, we’ll start putting more effort into working on CAD for the entire subsystems. We’ll have an update Wednesday that will include more information about our climber ideas and prototypes. As always, feel free to ask any questions you may have!

18 Likes

A great resource. Thank you. Interested to see your data on compression and wheel types. How exactly are the neo motors directly connected to the shaft on your shooter?

1 Like

how tall is this entire design? does it fit under the control panel?

1 Like

Entire layout is 27in tall to allow for fitting under the control panel.

3 Likes

Thanks to all this wonderful information we can all get our own… mechanical advantage this year.

image

13 Likes

We 3DP some couplers on our MarkForged. It’s actually two couplers, one 8mm to 1/2 hex adapter, then a 1/2 hex sleeve that fits over both shafts. This isn’t a long term solution but has worked well for testing.

Hahaha literally after every time we say mechanical advantage while talking about robot design in the shop we’re obligated to add a, “heh” after

5 Likes

this is really interesting. how do you plan to move the balls throughout the system. Most teams have stated that rolling them one 1 side is prone to jams, so you have to have movement on both sides.

It’s hard to tell from the 2D drawing but the intent is to have rollers on the top and bottom and I totally forgot to post this video but to get around the curve at the front of the robot we’d use a system shown in this video.

To keep the balls from rolling we’d gear the rollers so the rollers on the outside were spinning faster than the center roller so there is little ball rotation.

There is an open section on the top for balls to fall in for human loading.

3 Likes

oh interesting , do you intend to continue to use the the hex shafts themselves as the rollers? that could be very space efficient

Yes likely with some rubber o-rings connecting the shafts to provide a little more grip.