Check Out Our New Custom Frame!

[brennerator]

See below. Combined duplicate posts from separate threads into a single thread.

-Madison

[Joel J]

Don't make the mistake of leaving your cantilevered wheels out in the open. It cost 60 in 2004, and every "too exposed" cantilevered robot at least once in 2006. At the least, make it so that bumpers can be attached to the robot, to sort of "protect" the wheels.

Moderator note: This is relevant to the posts below it and was taken from another thread. I can't change the timestamp.

[brennerator]

Hi,

I started a custom vs. kit chasis thread a few days agao and we decided to go ahead and build a custom frame and here is what we came up with. If any of you are feeling especially motivated, feel free to check out our CAD files as well. We had a few more questions, any responses would be very much appreciated.

1) Our axils jut out 2.25 inches. With the force of the chain, will our axils bend?
2) Note the front and rear axils and that they have only one sprocket and that we have to use a spacer. What would be a good spacer to use?
3) Our wheels are exposed and we are using a 1/2 inch keyed shaft. Our wheels are high traction wheels from IFI Robotics In what ways could the wheels be damaged and are there any ways to prevent this besides putting a beam or bumpers on the outside.
4) Can we use use e-clips or spacers?
5) Any other flaws with our robot that we should know about before we go on and CAD it?

Thank you very much for your responses. They've been really helpful thus far.


http://www.pinewoodrobotics.com/files/chassis.zip

[Lil' Lavery]

If you're going to use cantilevered wheels like that, I'd suggest supporting the axles at two points, rather than just one.

[Madison]

Originally Posted by brennerator

1) Our axils jut out 2.25 inches. With the force of the chain, will our axils bend?
2) Note the front and rear axils and that they have only one sprocket and that we have to use a spacer. What would be a good spacer to use?
3) Our wheels are exposed and we are using a 1/2 inch keyed shaft. Our wheels are high traction wheels from IFI Robotics In what ways could the wheels be damaged and are there any ways to prevent this besides putting a beam or bumpers on the outside.
4) can use use e-clips or spacers?
5) Any other flaws with our robot that we should know about before we go on and CAD it?

Thank you very much for your responses. They've been really helpful thus far.

What sort of manufacturing capability is available to your team and to what extent are you proficient at using it? There are no photos of your past work in the gallery that allow us to get an idea of what experience your team has had in the past.

Some questions and concerns I have based on your CAD model:

• It appears as if your gearbox is driving the center axle through one chain, while a second chain connects all three wheels. As shown, that second chain engages the center sprocket across a very small arc along its bottom. I believe that, without additional wrap around the center sprocket, you'll find that the chain will not remain positively engaged at all times, resulting in complete loss of power transmission to the outer wheels. What's worse is that wrap is along the bottom side, meaning that the introduction of any slack in the chain will immediately find its way to the bottom. Don't underestimate the nuisance of gravity
• It's difficult to discern what components are custom-made and which are off-the-shelf. Is there particular use or reason behind the elaborate pillow blocks? They're complex to machine and their length through bore seems excessive, making the cantilevered load on each axle much further from the frame than it need be.
• Again, it's difficult to tell based upon only the rendering (and I don't have Inventor installed, so I can't open the files you linked to), but are the hubs meant to hold the wheels custom made? I think custom-made is the way to go here so that you can make certain the length through bore is equal to or greater than that of the IFI traction wheel you order. The off-the-shelf hub available from AndyMark is too short and has a diameter too small for the IFI wheels and would result in part of the wheel being completely unsupported.
• It looks like there's no provision for tensioning chain built in. Consider placing slots to allow the pillow blocks to be moved and locked down rather than discrete holes at fixed distances.

I think what you have is a good start, but that it needs refinement. Some of the other questions you've asked -- about spacers and e-clips and the like -- suggest to me that, overall, you might not be ready to tackle a drivetrain as resource and design intensive as this. It would make a fantastic project for the summer or off-season, though, and allow your team to gain a lot more experience and perspective on what it takes to thoughtfully design and build a custom drivetrain and frame.

Edit: I split the discussion about this drivetrain from the thread titled "Standard vs. Custom Frame". All discussion about this design should happen here from now onward.

[Billfred]

Have you tried the IFI wheels while cantilevered like that? It may be fine for all I know, but I don't think anyone's posted any information--good or bad--about the wheels in that configuration.

[Lil' Lavery]

Originally Posted by M. Krass

• It's difficult to discern what components are custom-made and which are off-the-shelf. Is there particular use or reason behind the elaborate pillow blocks? They're complex to machine and their length through bore seems excessive, making the cantilevered load on each axle much further from the frame than it need be.

I believe that those are off-the-shelf bearings. They look similar in size and proportion to the needle roller bearings 116 used in 2005.

[brennerator]

Im one of the engineers on this project.

So first of all one of the sprockets on the gearbox (yes there are two; the black one and our lazy CADDED silver one) drives the back wheel. The other sprocket drives the middle wheel. The middle wheel then drives the front wheel. This assembly worked last year; my only concern is the tension on the cantilevered axels.

Second: as to the question of eclips or spacers, last year we used shaft collars to space things out; this year we are looking for a more effective spacer. The spacer will be used ON the axel on the front and back axels to align the sprocket. The question was phrased incorrectly.

Third: the hubs are custom made.

Fourth: our frame is made of aluminum tubing .125" thick. Therefore i do not know if we can make a slot to lock down the pillowblock bearing (off the shelf). We were thinking of using different material to make our chassis out of, but we had some success with this tubing last year. Can anybody offer insight on what material we should use?

Last year we used a suitable tensioner that fixed right onto the chain.

We are using the 1" thick 6" IFI traction wheels. The thinner width should help lower the cantilever force, while the size will preserve the gear ratio. Please note that the sprockets are CADDDED as juxtaposed discs. The gear ratio has been calculated correctly however.

We are more concerned about structural aspects since our chassis drivetrain worked fine last year, but our chassis got physically damaged in the 3rd round.

[Madison]

Originally Posted by brennerator

So first of all the gearbox drives the back wheel which drives the middle which then drives the front. This exact setup worked for us last year.

Ah, that makes more sense. It's difficult to tell which plane the gearbox sprocket lies in. You mention below that the sprocket models are not accurate, so I presume that the differing diameters is one aspect that's incorrect.

Second: as to the question of eclips or spacers, last year we used shaft collars to space things out; this year we are looking for a more effective spacer. The spacer will be used ON the axel on the front and back axels to align the sprocket. The question was phrased incorrectly.

I don't have too much experience with retaining clips in high torque applications, but I'm sure someone like Cory will be along to tell more about their experiences. Anecdotally, I'd say they're probably an effective alternative.

We've used thin-walled Aluminum tubing as a spacer material in the past. It's effective but cumbersome.

Third: the hubs are custom made.

As long you're certain that the length through bore is greater than that of the IFI traction wheel, I think you'll be okay. I'm not sure how well the new iteration of the IFI wheels will take to being hit by other robots from the side, however. Last year, some folks had problems with the sidewalls collapsing.

Fourth: our frame is made of aluminum tubing .125" thick. Therefore i do not know if we can make a slot to lock down the pillowblock bearing (off the shelf). We were thinking of using different material to make our chassis out of, but we had some success with this tubing last year. Can anybody offer insight on what material we should use?

Creating the slots for tensioning is a pretty straightforward operation if you have access to a mill.

If you remain concerned about cantilevering load so far from your frame (and you don't have the capacity to properly calculate the effects of that on your materials), consider switching to a wider frame rail. Inverting the pillow blocks into the frame rail would reduce the distance between the wheels and the frame.

Speaking generally, it's very important that you provide as much information to us as possible so that we can make informed comments and answers to your questions. It's difficult to critique a design when critical pieces of information are missing.

[Jonathan Norris]

Originally Posted by Joel J.

Don't make the mistake of leaving your cantilevered wheels out in the open. It cost 60 in 2004, and every "too exposed" cantilevered robot at least once in 2006. At the least, make it so that bumpers can be attached to the robot, to sort of "protect" the wheels.

I remember a match against 254 in 2005 at the Sacramento regional where we were playing defense against them and damaged a few of their exposed wheels. Its definitely a design choice when using live axles like that. I would also recommend that you design for the use of bumpers with exposed wheels like that.

[brennerator]

I apologize,
A different team member wrote up the first post.
Thanks for all the help!

I especially liked the idea of inverting the pillowblock bearing.

If theres anything else you see please dont hesitate to critique us! Its a lot better now than after we CNC the thing.

[EricH]

This is a good start, but I have two questions.
1) Isn't it better to wait and see what the GDC throws at us before designing something? Admittedly, this is easy to change now, but nobody knows what's coming.
2) Will the area with no gearboxes be strong enough? I'm not sure that the side rails will stay straight in that configuration.

[Lil' Lavery]

I'd be afraid of using hollow wheels, such as the IFI wheels, in an exposed position like that. Something like the AndyMark performance wheels, AndyMark traction wheels, or some variety of colson wheels may be a better option. Regardless of wheel choice, I would also advise having plenty of spares (even if they don't break, if one tread gets worn out etc., due to the ease of switching live axle wheels, spares are often advantageous). If possible, you're going to want some form of exterior protection (typically a bumper), and not just for the safety of your wheels. 254 had trouble last year when teams discovered that if they hit 254 at high speed, they could often "hop" on top of 254's wheels, making it next to impossible for 254 to move effectively. A good example of this can be found in Newton Quarter-final 2-2.

[Dick Linn]

What are the dimensions of the rectangular tubing? Is it aluminum? We did a chassis awhile back with cantilevered wheels using a 2" x 4" aluminum frame member (a kit part at the time). Can you say "flexible"? Beware. I vastly prefer axles supported on two sides by frame rails. I am somewhat cautious, but in 5 years, we've had ZERO drivetrain failures, and very little maintenance. When you can't afford spare parts, you tend to do it right the first time.

[EricH]

Originally Posted by Lil' Lavery

If possible, you're going to want some form of exterior protection (typically a bumper), and not just for the safety of your wheels. 254 had trouble last year when teams discovered that if they hit 254 at high speed, they could often "hop" on top of 254's wheels, making it next to impossible for 254 to move effectively. A good example of this can be found in Newton Quarter-final 2-2.

968 had the same problem at LA. Their only loss was when a team somehow got on top of their back frame and slowed them down, to say the least. They later added a bumper, IIRC.