Design Help for offseason bot

[Sperkowsky]

Hello everyone.

So our team had some issues with our robot this year with a design error that was so big it was instant fatality. (The elevator was just bad no need for pics).

We ended the sbpli regional in last place and it really crushed our team.

So we decided to build an entirely new robot from scratch.

We are using a spare chassi from last year thats stripped down to 6 inch mecanums.

Our first issue was that it was in the long configuration. It was possible to work with but we all decided to go for a team 20 apporach and have the mecanums facing sideways.

our second decision was material choice. We made the bad decision of using bar stock for an elevator track so this time we chose something much more robust and went for 2x1 aluminum rectangular tubing.

For the elevator assembly we are going to take peices of bar stock with 2 bearings on the sides gripping the tubing. With 4 sets of 2 bearings we believe it will slide nicely with any weight load and again the assembly is held together with 2x1 rectangular tubing.

we also decided to do powered intakes with 2 andymark motors with plannetary boxes. We dont know if how we them mounted is the best way though.

one big decision for us right how is whether its a big deal to have a built in ramp or just tether one.

The next big decision is what material to use to power the lift. Some people vote webbing some vote chain. Chain seems more robust and doesnt seem to fall victim to back drive as much where webbing seems more forgiving. chain will allow us to do a built in ramp as there is a few inches where we can attatch the chain to the ends of the elevator assembly. Another question there is do you think we need a chain on each side.

for the actual manipulator we think we are going to reattach our plastic tabs similar to ri3d indiana and have a raised hook section for recycling bins similar to cybertooth robotics.

btw most of our structure will be rivited together with L brackets

Heres a pic
http://gyazo.com/dca65feb9971c9cbbb49cc609963b393
thanks

[Kevin Leonard]

Originally Posted by Sperkowsky

Our first issue was that it was in the long configuration. It was possible to work with but we all decided to go for a team 20 apporach and have the mecanums facing sideways.

Would not recommend.
We faced issues all year with strafing into our objectives, and only once we got partway through Championships did we finally solve them and our robot functioned mostly as intended.

Otherwise the design looks like a smart way to play the game, and backfeeding with "Indiana-tabs" seems to be the smart, simple design for 2015 human loaders. Even simpler is forgoing backfeeding and using a ramp, as many other successful teams did.

Good luck!

[ToddF]

Sam,
Your team is embarking on a great exercise. You are starting a robot design with a completely blank slate, but with the ability to look back at how the game was most effectively played. If you do it right, this is a rare opportunity.

I'm going to suggest that you are putting the cart way before the horse. You are jumping into the design phase before you have done the most basic step: game analysis and establishing priorities. Before ever beginning to design our robot, our team always begins with the game analysis process we learned from Den Black. I encourage you to go through this process yourselves. You can learn about the details by watching the following two videos (one, two)

When you have done the analysis, post the results, and I'll be happy to give you some guidance on your design. But until you have decided how you will play the game, and prioritized your scoring actions, you aren't engineering; you're just winging it. You are jumping right into detailed design before you have defined the problem in a way that will guide you to a viable solution. That's probably the single biggest mistake in FRC, and it's nearly always fatal.

Check out my youtube channel to see how we approached the problem. We were fairly successful, and won two regionals this year. This does NOT mean that you should copy our robot. It's the process you need to learn, not how to copy the solutions of others.

[Sperkowsky]

Originally Posted by Kevin Leonard

Would not recommend.
We faced issues all year with strafing into our objectives, and only once we got partway through Championships did we finally solve them and our robot functioned mostly as intended.

Otherwise the design looks like a smart way to play the game, and backfeeding with "Indiana-tabs" seems to be the smart, simple design for 2015 human loaders. Even simpler is forgoing backfeeding and using a ramp, as many other successful teams did.

Good luck!

The reason why we want to do the wide chassi with front facing wheels is because we kind of have to.

Changing the wheel config is not an option with the chassi and going facing the other way makes it very hard to build as there's no good way to mount the 2x1s.

Our only option would be building a chassi from the beginning something we have never done before.

What helped you sort the issues out was it Driving issues, code issues, any guidance would be awesome but the long chassi is something we really can't change.

[EricH]

Originally Posted by Sperkowsky

Our only option would be building a chassi from the beginning something we have never done before.

Ya know, a lot of teams will take the offseason to prototype new drive ideas, including building a custom frame from the beginning. Just a thought.

[Kevin Leonard]

Originally Posted by Sperkowsky

The reason why we want to do the wide chassi with front facing wheels is because we kind of have to.

Changing the wheel config is not an option with the chassi and going facing the other way makes it very hard to build as there's no good way to mount the 2x1s.

Our only option would be building a chassi from the beginning something we have never done before.

What helped you sort the issues out was it Driving issues, code issues, any guidance would be awesome but the long chassi is something we really can't change.

I have no idea how they fixed the Mecanum drive- its one of those darned programming things I've never been good at. However, even when the robot was working at full capacity, the strafing was never that great, and placing stacks on a scoring platform in the strafing direction is not easy to do.

Originally Posted by Sperkowsky

It's funny I followed your robot design process pretty closely throughout the season and one thing we will probably pull from is the disc brake.

But that aside

Our single objective is a simple reliable robot that can stack from the hp station or the landfill.

A ramp would speed up a load from the human player station and our powered intakes will allow us to be fast Through the landfill.

We want to be able to cap a six stack which is why we have the raised hook to lift the recycling cans off.

Although in higher level play cans are a determining factor a can grabber is more of an add on then a key design feature at the moment.

Our goal is 2 capped six stacks that is not too Crazy unreasonable but is 84 points.

Our biggest design factor has to be reliability

I would focus on either human player loading or landfill loading, not both. And then also consider how you're going to place the stacks, not just how you're going to build them.

ToddF is right, though. You need to come to these conclusions with your team via an in-depth strategy analysis, and pick the right design for you.

I would recommend building in the longwise direction, and for human loading using a ramp, re-designing the backfeed ramp, or considering a design like Team 1325, our fantastic alliance captain at championships.

[Sperkowsky]

Originally Posted by ToddF

Sam,
Your team is embarking on a great exercise. You are starting a robot design with a completely blank slate, but with the ability to look back at how the game was most effectively played. If you do it right, this is a rare opportunity.

I'm going to suggest that you are putting the cart way before the horse. You are jumping into the design phase before you have done the most basic step: game analysis and establishing priorities. Before ever beginning to design our robot, our team always begins with the game analysis process we learned from Den Black. I encourage you to go through this process yourselves. You can learn about the details by watching the following two videos (one, two)

When you have done the analysis, post the results, and I'll be happy to give you some guidance on your design. But until you have decided how you will play the game, and prioritized your scoring actions, you aren't engineering; you're just winging it. You are jumping right into detailed design before you have defined the problem in a way that will guide you to a viable solution. That's probably the single biggest mistake in FRC, and it's nearly always fatal.

Check out my youtube channel to see how we approached the problem. We were fairly successful, and won two regionals this year. This does NOT mean that you should copy our robot. It's the process you need to learn, not how to copy the solutions of others.

It's funny I followed your robot design process pretty closely throughout the season and one thing we will probably pull from is the disc brake.

But that aside

Our single objective is a simple reliable robot that can stack from the hp station or the landfill.

A ramp would speed up a load from the human player station and our powered intakes will allow us to be fast Through the landfill.

We want to be able to cap a six stack which is why we have the raised hook to lift the recycling cans off.

Although in higher level play cans are a determining factor a can grabber is more of an add on then a key design feature at the moment.

Our goal is 2 capped six stacks that is not too Crazy unreasonable but is 84 points.

Our biggest design factor has to be reliability

[EricH]

Originally Posted by Sperkowsky

Our single objective is a simple reliable robot that can stack from the hp station or the landfill.
[...]
We want to be able to cap a six stack which is why we have the raised hook to lift the recycling cans off.

[...]
Our biggest design factor has to be reliability

Now comes the fun part. I'm only going to pick on you for a few of the items you listed--see above--but as an exercise, you might want to go through this with your team. It'll give you a great insight into designing a competitive robot (or other product).

Why? Please answer for all three items. (Does not have to be made public.)
--Why do you want to be able to stack from either location? Why is that important, or IS it important? Why is stacking important? (OK, seriously, I'm getting a little on the off-base side here. But, this is something that if we'd never seen the game before, I'd most definitely be asking.)

--Why a 6-stack? Can you do shorter if need be (ability vs need, definitely something to look at)? Why is the recycling can important?

--Why is reliability the biggest design factor? What about competitiveness? Or coolness? (TBH, I know what you're getting at. But, again, if this were at the beginning of the season, I'd be looking at "is an unreliable high-point item better than a reliable medium-point item?"--I often come back with "no" myself.)


Now, back to your other question of what design will work the best: That's for you to figure out. Start by answering these questions... and that'll give you some requirements... and any design you come up with will need to be evaluated against those requirements.
1) Is it important to be able to go over the scoring platforms?
2) What orientation do we expect to handle totes in?
3) Is it important to be able to keep pool noodles out of the drivetrain?
4) Is it important to protect the drivetrain from robot-robot collisions?

Those questions would answer ground clearance, overall width, and a couple of key addons (or not-so-key addons). A good answer for #1 might come back as "It is not important, but doing so places these limits on play. We can't handle those, therefore it is important (or vice-versa)." Which would then translate into a requirement "This robot SHALL (or does not have to) be able to go over the scoring platforms with ease". This will then drive some ground-clearance decisions. Some searching of CD-Media or FRC Designs will probably turn up some passable drivebases that could be adapted to fit your requirements.

[Sperkowsky]

Originally Posted by EricH

Ya know, a lot of teams will take the offseason to prototype new drive ideas, including building a custom frame from the beginning. Just a thought.

Not a bad idea.

We put in a pretty big grainger material order and will have extra 1x1 tubing 2x1 tubing and a good amount of rev extrusion.

What design do you think would work best with our tuff box minis and mecanums.

[Tweedy399]

One thing that my team and many others teams did was add an RC "stabilizer". This would grab the RC initially and follow it all the way to the top of the stack to secure the RC and the whole stack. This can even not be powered but just ride with the tote manipulator and the stack. We have found great success in this and I would highly recommend it.

[waialua359]

Originally Posted by Tweedy399

One thing that my team and many others teams did was add an RC "stabilizer". This would grab the RC initially and follow it all the way to the top of the stack to secure the RC and the whole stack. This can even not be powered but just ride with the tote manipulator and the stack. We have found great success in this and I would highly recommend it.

Great recommendation.

I would also recommend an external ramp if you choose not to have one internally built as part of your robot.
We spent $10 buying 1/16" angle alumimun and a thin 1/32" acrylic sheet in making a simple ramp that worked wonders for us this year.
The same spectra cord used for our elevator, we used as our tether and it never got tangled in our wheels for 3 events and counting.

[philso]

In addition to the videos that Todd linked to, you may want to study the "Strategic Design" YouTube video posted by Karthik before you really get started. It may also be of benefit to study game videos and compare the driving characteristics of robots with long chassis vs. that of robots with wide chassis, especially when traversing the scoring platforms at an angle. You may also notice that many of the fast HP stackers (i.e. the Einstein winning alliance) did not need ramps tethered to their robot (or a third robot).

Originally Posted by Tweedy399

One thing that my team and many others teams did was add an RC "stabilizer". This would grab the RC initially and follow it all the way to the top of the stack to secure the RC and the whole stack. This can even not be powered but just ride with the tote manipulator and the stack. We have found great success in this and I would highly recommend it.

This echos what a mentor from one of the Einstein winners said in the 13th post of the following thread. We added an RC stabilizer after our first Regional. If we had done our strategic analysis properly, we would not have added it. We would have had it on our robot from the beginning and we would have done better at our first Regional.

http://www.chiefdelphi.com/forums/sh...d.php?t=136979

Originally Posted by Dunngeon

Reiterating what Todd and Eric said, it really seems like you're wanting to cherry pick things that worked for other teams. Looking at the successful designs is great, but what do you gain in the long term from building a good, robust robot for THIS game?

One of the biggest mistakes I see teams (FRC and FLL) making is that they copy without understanding what they are copying. It is good to study another teams design to add to your repertoire of design techniques and mechanisms. When you copy only one piece of someone's design without understanding the whole, you may miss copying another part of their design that is crucial to the success of the part you are copying.

The reliability that you are after has as much to do with the quality of construction as the design of the robot. Your team may want to look at the construction practices (mechanical and electrical) in the off season and make any necessary improvements in the off-season. Your off-season robot can be the final test for any training you do. Monochron's comments about the lift are spot on. Or lift mechanism was built crooked and when it was first installed, there was so much friction it had to be forced up and down. I highly recommend the "Electronics Tutorial" videos published by Code Orange. We had ZERO electrical problems this year through two Regionals and Champs.

http://www.teamcodeorange.com/tutorials-electronics.php

[Sperkowsky]

hello everyone,

so we have ordered all of our parts and I have a few questions but first what we ordered

Vex $125

Flanged Bearing - 13.73mm (1/2" Tight Press ThunderHex) x 1.125in x 0.313in (11-pack)
1/2" ThunderHex Stock (3 feet) x2
#25 Sprocket w/ Hub - 16t - 1/2" Round ID x6
Clamping Shaft Collar - 1/2" Hex ID x4
#25 Heavy Duty Master Link x10

Competition robot parts $50.00

$50 dollar version including rollers and roller hardware

Pitsco

Tetrix .25 chain
Tetrix max chain breaker tool

Andymark $430.00

Planetary GearMotor Bracket (am-2197) x2
PWM Connector Kit (am-2817)
Logitech Extreme 3D Pro USB Joystick (am-2237)
Assorted Heat Shrink Tubing Kit, Red, 127 Pieces (am-2861)
REV Robotics 1" Extrusion, 6ft (am-2901) x3
Assorted Rev Extrusion Brackets x36
REV, 3/8" bore V-groove Ball Bearing (am-2914) x4
#25 Connecting Link for Roller Chain (am-0371) x10
040.5-DP Bimba Air Cylinder w/ nut (am-0292)

Grainger $800
I dont have the full list but it included
1x2 aluminum tubing 6ft x6
1x1 aluninum tubing 6ft x4
1/4 inch rivets
lazy tong rivet tool
and a crap ton of heavy duty steel corner braces for structure.


is all that good enough to build the robot specifically the chain parts as we dont have many parts spare in the shop.

I saw screws that went through chain to connect it to the elevtor what is that called.

We are trying to something similar to team 3940.

thanks

[asid61]

If you're using #25 chain, buy the Dark Soul chain tool from team 221. It allows you to break and reconnent #25 chain without using hardware.

[GeeTwo]

Coming in a bit late to the party, but I'll try to check everything:

From OP, for future reference, in lengths more than about four inches, bar stock should only be used for tension elements. If the item will be under any sort of compression or torsion, go with angle or channel (or, though I've not designed with it, sheet metal with appropriate bends).

I'm not entirely sure where the "initial robot" ends and the "new robot" begins. For sanity's sake I'll assume everything after "We are using a spare chassi[s]" is all about the new robot.

You refer to rotating the mecanum wheels sideways, and Kevin Leonard advised negatively. Unfortunately, the graphics provided do not really provide any idea as to how the wheels/gearboxes/bearings were attached to the frame in the first place, nor how they are mounted now. Perhaps he has more info. 3946 used Mecanums for Aerial Assist (we vastly underestimated defense!), and the key elements to make mecanum work from my experience are:

• (As with any drive system) make sure that there is a solid support structure. The wheels must support the axles (and/or gearboxes), the axles/gearboxes support the beams, the beams support the joists, and the joists support the rest of the robot. The key thing is NOT to think so much about the robot chassis as supporting the gearbox/axle, as to think about the gearbox/axle supporting the chassis. (OBTW, the same is true of airplane fuselages and wings, which is where I learned this many years ago, before FRC existed; wings that fold up are no good for flying. I remember cringing when I saw the F4Us in Baa Baa Black Sheep/Black Sheep Squadron, and marveling when I later found out that yes, those were real WWII planes!)
• Keep the axles horizontal. We mounted AM TB2 gearboxes on the main beams with no bearings on the ends of the axles, and had to stiffen up those mounts considerably.
• Provide enough flexibility in the chassis and/or a suspension so that weight is equally distributed among the wheels. This one I would consider more of a "good idea" than a "hard requirement". Our practice robot for Aerial Assist (made of lumber, and the product of slamming together a wooden pickup chassis and makeshift mecanum mounts with an external control board that was almost a trailer, but was eventually "twisted around the umbilical" and "screwed to the right side" of the robot) had about four times as much weight on the right rear wheel as the front left wheel, but the drivers learned to work with it fairly quickly.

I can't figure out this quote - the bottom line is that bar stock cannot support compression or torsion, only extension. If the concept of swapping out the bar stock with cable gives you the creeps, bar stock is probably not the right answer.

For the elevator assembly we are going to take pieces of bar stock with 2 bearings on the sides gripping the tubing. With 4 sets of 2 bearings we believe it will slide nicely with any weight load and again the assembly is held together with 2x1 rectangular tubing.

Ramp: can't help you there first hand - we decided early to go landfill. If you do manage to get your lift faster than your HP can feed through the chute door, I recommend the long low slope ramp used by 254 (Cheezy Poofs) on EinsteinCarson; it elegantly coupled this impedance contrast without requiring a "Robin" style (see team 148, RoboWranglers) active stacker.
What do you mean by "webbing"? 3946 used chain, as we did not have ready access to long loops of timing belt, but we knew that we could extend 10' lengths of chain with master links as long as we needed. We used two chains to provide torsion control in the "roll" dimension. With sufficiently robust rollers to handle this torque, this would not be necessary. We did not trust our ability to make such a system, and decided to use two chains.

Sorry, but the Ri3ds seem like years ago now. Even after I went back, I'm not sure exactly what you're talking about with Ri3d Indiana.