03-11-2014 10:50

JesseK

Re: pic: Team 3572 Chassis Project

I really like most of the ideas behind this configuration. There is only one major modification I would make: Cantilever the middle wheels and remove the bent bracket on the outside altogether.

This bracket looks like it will receive all of the load that the wheels will put onto the frame. Its one major weakness is that it will most likely flex/bend in the direction parallel to the drive wheels when the robot is fully loaded at 150lbs or pushing/resisting another 150lb robot. This can be mitigated by adding angle supports along the outer frame, yet a steel cantilevered axle will handle it just as well without the extra parts.

I'm still mulling over how other pieces would react to the various forces imparted upon them during a match, but nothing jumps out at me at the moment.

03-11-2014 10:53

nathannfm

Re: pic: Team 3572 Chassis Project

It's hard to tell from this angle, and I assume you checked this already, but it looks like the belts to the outer wheels might interfere with the CIMs.
Also, I can't really tell what's going on with the outer wheel axles, the hex looks like it just kinda merges with the 2x1 with no bearing, and on the oppisite side looks like it has been turned down to 1/4in or something.

...wait...

Those are bolts aren't they lol, I have never seen someone cantilever bolts before, you might get some wobble if the threads are going through a bearing because the diameter won't be exactily that of the bearing or the tube they are running through...

You might also consider adding a 3rd CIM in the center with a shorter belt and widening the pulley to accommodate.

Those plates give me a bit of pause, you may want to brace them, but its worth a try in the off season.

03-11-2014 11:25

MrBasse

Re: pic: Team 3572 Chassis Project

Design isn't totally done yet. We were thinking of cantilevering the front and rear wheels and running dead axles with a support block inside the frame rail to make it solid. We would also use smooth shank fasteners.

We have never done this before, so our big fear is building something with weak points. That's why we plan on building it now and refining until 2016. The brackets around the center drive portion are to keep everyone happy, the big debate has started with us between cantilevering the whole thing or keeping those in to make sure the hex shafts stay aligned.

The belts miss the CIMs by .100 gap.

A major goal of this was to be able to build it with what we have in our shop currently. It's killing me that we have to send the rails off somewhere to get them machined for us as we don't have capacity to do it ourselves (a drill press just isn't that accurate, and drills don't make very good holes for bearings).

03-11-2014 11:29

jwfoss

Re: pic: Team 3572 Chassis Project

Can you explain the theory behind your locations of your 8 wheels?
I'm curious about your choice in spacing the two center wheels so close together, thanks.

03-11-2014 11:35

Travis Schuh

Re: pic: Team 3572 Chassis Project

What is your belt wrap for the stage off of your CIMs going to look like? If that is one belt tangent to the outside of all 3 pulleys, then I would question if you have enough belt wrap on the driven pulley in that arrangement.

03-11-2014 11:35

Lil' Lavery

Re: pic: Team 3572 Chassis Project

I'd suggest increasing that clearance. There will be some motion in the belt while spinning, and it may end up touching the CIM with only 0.1" clearance.

03-11-2014 11:42

Scott Kozutsky

Re: pic: Team 3572 Chassis Project

I really don't trust the outside wheels. Our team made a minibot with cantilevered wheels on 3/8in bolts and despite it being substantially lighter than an FRC robot it could not handle the forces. Your wheels are also cantilevered out further than our minibot and are not on shock absorbing pneumatic tires leading me to trust this even less.

Also, you need blocks on the inside of the 2x1 to prevent the wheel bolt from just crushing it.

03-11-2014 11:44

MrBasse

Re: pic: Team 3572 Chassis Project

We have always been a 6WD team, and have gotten pretty good at balancing the robot right on the center wheels. The thought was that we would have less tip to the robot this way by keeping the distance between the outer and inner wheels as big as possible. And in keeping the middle two tight together we get direct driven gearing, and we can run like a 6wd that has a wider balance point and two extra wheels for more contact with the carpet.


We are going to run two 9mm belts from the motors, We tried the full wrap on a single belt and had a lot of engagement issues. Only four teeth were in contact on the main pulley.

a little belt slap is fun though right? We will be able to move the motor plates a bit in our final production. Those plates we will plasma cut in house so we can always make them taller and closer to the center drive unit.

03-11-2014 11:46

MrBasse

Re: pic: Team 3572 Chassis Project

Thanks for the tip, that is my biggest concern as well. We do plan to run solid blocks in the frame, either wood or UHMW as we have a big chunk left from last year. I want to learn as much as possible from this, so that our kids can learn as much as possible from this, so we are trying a lot of new things.

03-11-2014 11:47

nathannfm

Re: pic: Team 3572 Chassis Project

Oh, I didn't even consider he might be thinking this, don't do it! You want a very minimum of six teeth in total mesh and at least 60 degrees of belt wrap. I am pretty sure we had a bit more of both of those on our drive this year and even under extreme tension they would still ratchet. If you need a wider pulley I suggest http://www.vexrobotics.com/htdversapulley.html you can add slices to increase with width.

03-11-2014 12:01

MrBasse

Re: pic: Team 3572 Chassis Project

With two 9mm belts I think we'll be okay as long as we get the tension right.

03-11-2014 15:22

75vs1885

Re: pic: Team 3572 Chassis Project

looks like the cims would interfere with the bumpers, and they could get damaged if hit to hart from another robot. Is there an advantage to having them pointing outside opposed to inside?
Other than that it seems really solid.
btw, would the electronics board be mounted underneath?

03-11-2014 15:42

MrBasse

Re: pic: Team 3572 Chassis Project

Made a few updates. Looks pretty much the same but added in blocks for the front and rear axles. Also bumped the motors up and over to give .25" clearance for belt movement. Also I removed the braces on the outside of the frame to cantilever the wheels and added retaining bolts on the axles.

I'll try to add belts tomorrow so the clearances can be seen better.

http://imgur.com/Xw2xC0l

03-11-2014 15:45

MrBasse

Re: pic: Team 3572 Chassis Project

bumpers would mount on the outside of the CIMs, but mounts haven't been figured out yet. Putting them outside gives a huge amount of space to get all of our electronics down as low as possible in the frame. We are trying to get back to our 2013 design where our center of gravity was about 4 inches off the floor. Having the battery lay down in the middle of the robot makes that pretty easy.

We have always only used around 13-14 lbs of bumper, a more ridged attachment frame that eats up around 5 lbs of bumper weight could be all the protection the motors would need, I hope...

03-11-2014 16:52

Deke

Re: pic: Team 3572 Chassis Project

I like the layout, it is very creative.

It looks like the cim plates are 1/8" thick, if they are, I would suggest at least 1/4" thick with the Cims cantilevered off them. 1/8" works if both sides are supported, but a big bounce or hit will put a large load on that thin plate.

The middle drive pulley to the cims looks cantilevered from the frame tube, I'm not sure if would would be an issue or not, but it would help if you can support it with a bearing on both sides. Some testing of the current layout will let you know, just something to watch out for.

I would suggest the dead axles with blocks for the outside wheels like you suggested, we ran cantilevered dead axles last year and loved the performance over live axles. With your belt layout, it makes it easier to run dead axles versus live axles as well.

03-11-2014 17:18

JesseK

Re: pic: Team 3572 Chassis Project

This was my next comment since the CIM mount plates seem a bit tall. If the two mount plates were combined into a single mount plate, the design may be able to add strength without adding weight. It could also incorporate more belt wrap around the pulleys.

03-11-2014 17:24

Andrew Schreiber

Re: pic: Team 3572 Chassis Project

We (125) ran all last season with them dangling outside the frame like that. Had 0 issues with them. (1x Preseason, 4x Districts, DCMP, CMP, numerous off seasons)

Just remember that the wires exist and have minimum bend radii.

03-11-2014 17:31

Deke

Re: pic: Team 3572 Chassis Project

Running a support beam across between the plates would help too, just something like a hex shaft could be used, but it would be annoying having that hang over the middle of the bot. Even a 1/4" plate would make nervous without a gusset or additional support.

03-11-2014 18:46

MrBasse

Re: pic: Team 3572 Chassis Project

The motor plates are 1/8", but they are also steel. This would allow us an easier time adding supports to them. A bar across might not be a problem depending on where the electronics end up. I would like to avoid it if possible though.

The majority of our components were made of 1/2" conduit last year as it weighs much less than the aluminum of equivalent strength. Plus, I hate TIG welding. Once I figure out what I'm doing, I'm usually done with the project. If we need support we will most likely go back to 1/2" conduit. $2.50 a stick can't be beat with proper welding ventilation. Plus when you are done you can paint it aluminum color and nobody ever knows.

I'll keep an eye on the pulley for flex. If needed we can just extend the shaft and add another hole into the inner bearing plate. That was actually in the original design, but the shaft was cut short for prototyping and it seemed okay. Only building both sides and beating the snot out of it will tell if it will hold up though.

04-11-2014 00:22

asid61

Re: pic: Team 3572 Chassis Project

Looks pretty nice, but clean up the front/back axles in the cad. Either cantilevered live axle or dead axle will work, but decide on one pre-season.
Personally I recommend live axle, as it's easier to swap out wheels or gearboxes if something breaks.

Also, 12fps is too slow. Gear for around 15fps at least. We ran around 16.5fps at CalGames on 4 cims with no problems; we were running 12fps during the first half of the season, and the extra speed was very exciting.

04-11-2014 07:24

protoserge

Re: pic: Team 3572 Chassis Project

How do you adjust your belt tension and alignment?

Did you account for mechanical drivetrain loss when calculating your 12-13 fps?

I'd be curious as to the drive belt engagement given your configuration.

04-11-2014 08:21

MrBasse

Re: pic: Team 3572 Chassis Project

We can use the motor mounts to apply more tension if needed, just have to drill more holes.

Speed is based on JVN's calculator set to 81%. and double checked by my awful math and an 80% efficiency.

Belt runs can be seen here. After teaching drafting for ten years I realized I had never used the design accelerator in my life, so I had to play with it for a while to see what it could do. Ignore the fact that there is a 9mm belt on the wheel pulleys. That would be an 18mm belt there.

04-11-2014 08:35

MrBasse

Re: pic: Team 3572 Chassis Project

With two speed gearboxes last season we had our high gear running us around 18 FPS and it just didn't get used that much. It very much should have, but it didn't. Our drivers have always just seemed more comfortable with a more controllable machine at a lower speed.

04-11-2014 10:07

Scott Kozutsky

Re: pic: Team 3572 Chassis Project

They need more practice. There were teams geared to 22 fps this year and they weren't doing too badly once the drivers learned how to use it. Maybe try adding an easily adjustable 3rd stage like the versachassis WCD has so you can get your drivers up to the maximum speed they feel comfortable with. Our driver this year was geared to 16fps and wanted to go faster.

04-11-2014 10:13

asid61

Re: pic: Team 3572 Chassis Project

+1. It's better to have the ability to go fast is all I'm saying; that way you have a choice.

04-11-2014 10:17

hrench

Re: pic: Team 3572 Chassis Project

Looks interesting. I'd consider moving the motors lower and flipping them inside to lower the CG. Also, they look vulnerable where they are.

04-11-2014 10:55

MrBasse

Re: pic: Team 3572 Chassis Project

The height they are at is set by the pulley positions and the belts available. Putting them inside eliminates the added floor space gained by putting them outside. I'm also not too worried about them hanging outside the frame as the bumpers will almost completely cover them. We have been putting our air storage there for the last three years and have never had an issue, so I am fairly confident in the location of the motors. This is something we will keep an eye on to be sure though.

We will wait and see what the game is before we make gearing choices. If we gear up all we have to do is move two holes and spread the middle wheels a bit. We can get all the way up to ~18 fps by changing one gear. Or up to ~16 fps by changing a few pulleys in our primary reduction.

I agree on practice, but without the funds for a practice robot and typically finishing up with hours to go, it is hard to come by. Also, a location big enough with carpet that would allow real practice has long been a dream of mine.