Mecanum wear and dirt resistance

[Tom Ore]

Originally Posted by Ether

If the rollers are free to spin

The rollers aren't necessarily free to spin. Some spin fairly well, others are sticky, some are stuck. The newer AM mecanums seem to be more consistent than the older ones. Breakaway was rather rough on the mecanums - bouncing down the back side of the bump bent the screws holding the rollers and the rollers would stick.

[Ether]

Originally Posted by Tom Ore

The rollers aren't necessarily free to spin. Some spin fairly well, others are sticky, some are stuck. The newer AM mecanums seem to be more consistent than the older ones. Breakaway was rather rough on the mecanums - bouncing down the back side of the bump bent the screws holding the rollers and the rollers would stick.

If the rollers are not free to spin, then the wheel is no longer operating as a mecanum wheel.

... and if a swerve steering motor burns out or jams, the wheel is no longer swerving and the affected wheel will scrub.

BTW, don't use mecanum on dusty dirt, you are asking for trouble. The roller "bearings" will get contaminated and it won't operate properly any more.

[Tom Ore]

Originally Posted by Ether

If the rollers are not free to spin, then the wheel is no longer operating as a mecanum wheel.

... and if a swerve steering motor burns out or jams, the wheel is no longer swerving and the affected wheel will scrub.

BTW, don't use mecanum on dusty dirt, you are asking for trouble. The roller "bearings" will get contaminated and it won't operate properly any more.

The wheel can still function as a mecanum but at a reduced performance - only some of the rollers are sticky or stuck. With our Breakaway bot it still ran fairly well even though some of the roller shafts were bent.

[flameout]

Okay, so it sounds like they won't wear too fast (we won't run on asphalt bad enough to cause horrible control issues).

However, since it sounds like we'll have to replace them about once per year, do people have recommendations for how to replace the rollers? AndyMark sells them for $3-4 apiece, which is too much for us to spend each year (we have no problem buying replacements -- I would like to get the yearly maintenance cost down to $20 or less, which necessitates cheap roller replacements).

We've talked about fashioning rollers out of PVC ourselves, but I fear that it is too slick, negating the mecanum effect (by sliding just as easily along the axis of rotation as it rotates around that axis). Is there a less expensive way for us to make our own rollers (I know other teams have used urethane, but I recall that being above our budget)?

It sounds like the best option would be to buy a set of AM mecanums (or other COTS mecanums) and replace the rollers with something cheaper than AM's rollers each year -- does anyone disagree?

[PAR_WIG1350]

Originally Posted by flameout

Okay, so it sounds like they won't wear too fast (we won't run on asphalt bad enough to cause horrible control issues).

However, since it sounds like we'll have to replace them about once per year, do people have recommendations for how to replace the rollers? AndyMark sells them for $3-4 apiece, which is too much for us to spend each year (we have no problem buying replacements -- I would like to get the yearly maintenance cost down to $20 or less, which necessitates cheap roller replacements).

We've talked about fashioning rollers out of PVC ourselves, but I fear that it is too slick, negating the mecanum effect (by sliding just as easily along the axis of rotation as it rotates around that axis). Is there a less expensive way for us to make our own rollers (I know other teams have used urethane, but I recall that being above our budget)?

It sounds like the best option would be to buy a set of AM mecanums (or other COTS mecanums) and replace the rollers with something cheaper than AM's rollers each year -- does anyone disagree?

if you buy a full set of 12 rollers (for 8 inch wheels) it only costs $40 times 4 wheels means it would cost $160 per year IF you used AM rollers. To build new replacements for 12.5% of the cost of COTS rollers might be difficult, but there are other options that might be cheaper:

1) protect the rollers when driving on rough ground
Either lay out something to cover the ground or cover the individual rollers with a sleeve of some sort, the former is easy and quick, but a ground cover sturdy enough to drive on might need a lot of storage space. The latter would likely be time consuming and trying to keep the roller covers in place would be challenging, and performance might be affected.
2) have a different set of more durable rollers for driving on rough ground
Skateboard wheels come to mind, put 2 or three on an axle and shape them to the approximate contour of the mecanum rollers (the sponsors with machining capabilities might be able to help). Just switch the rollers when you switch surfaces.
3) instead of replacing the rollers, maybe new rollers could be cast over the old ones
This saves money on urethane, but the layered structure might fail, which wouldn't be good
4) this one is like number three, but the rollers are cast over PVC cores, this allows for more control over the structure of the wheel so that the rollers are more uniform, every time the rollers wear out, remove the remaining tread and cast a new one over the core.

I think the even numbered ones sound like they might work better, but if you could find rollers within your budget, that might be even better.

[flameout]

Originally Posted by PAR_WIG1350

Either lay out something to cover the ground

Unfortunately, this is not possible.

cover the individual rollers with a sleeve of some sort ... likely be time consuming and trying to keep the roller covers in place would be challenging

This is something I have thought of... however, the rollers are closely spaced IIRC, so we'd need to grind them down first, but I think it'd be doable. However, coming up with a suitable cover seems like the hard part... any ideas?

performance might be affected.

Performance is not very important, although reliability is.

2) have a different set of more durable rollers for driving on rough ground
Skateboard wheels come to mind, put 2 or three on an axle and shape them to the approximate contour of the mecanum rollers (the sponsors with machining capabilities might be able to help). Just switch the rollers when you switch surfaces.

Switching the rollers would be quite time-consuming, but I don't see us doing it more than a few times per year, so it's plausible. This would make dealing with gym floors easy -- we'd only switch for the duration of our presentations (the rest of the time we'd use our low-wear wheels).

I'm not familiar with skateboard wheels, but I'm worried that they might not have enough traction. Although we don't need a lot of traction, I think the kinematics of a mecanum drivetrain break down when the wheel isn't significantly more difficult to skid than it is to roll.

3) instead of replacing the rollers, maybe new rollers could be cast over the old ones
This saves money on urethane, but the layered structure might fail, which wouldn't be good

That's something I haven't really thought about... good idea.

4) this one is like number three, but the rollers are cast over PVC cores, this allows for more control over the structure of the wheel so that the rollers are more uniform, every time the rollers wear out, remove the remaining tread and cast a new one over the core.

That does sound like it would work.

Also, I don't see why we need to use urethane. Is there anything less expensive that would still do the job? I think I recall something a while back about casting holiday ornaments out of hot glue. What do you think about this idea?

EDIT:
Earlier I mentioned a concern about dirt -- it is possible to avoid traveling through the dirt (just a minor inconvenience). Therefore, in my opinion, it comes down to our ability to cheaply repair/replace the rollers.

[Tristan Lall]

I don't think you're going to achieve savings trying to develop a process for making replacement rollers. Assuming your team members' labour is worth anything (either in terms of money, or in terms of an alternative activity), you're going to be hard pressed to beat a cost of $160 per year. Plus you'll have to pay for raw materials as well.

If you really want to spend time developing better Mecanum wheels, that's another story. Then the R&D necessary to investigate new methods of roller production will be a good project.

How much mileage do you intend to put on these wheels in a year, anyway? Depending on the team, it could range from hundreds of metres to tens of kilometres.

[flameout]

Let's assume that team labor is free

I'm really looking into casting hot glue right now. We could make a mold using a pristine roller (we'd order an extra set of 12 to replace rollers that we mess up beyond repair) -- I believe there are commercial products for building reusable molds for stuff like this.

Further, by using hot glue, it might be possible to simply re-heat the roller (I'm not sure what the maximum temperature of AM roller before damage is done is) to remove the hot glue before re-casting the glue around the roller (we would use this as a roller-repair method).

An added benefit would be the knowledge of how to perform a casting.

Looking online, it appears as if enough hot glue to last us for many years can be obtained for $10, and most of the materials for making molds is around the same price, putting this process well within our team's budget (and freeing up funds for other projects).

We will be putting in perhaps a kilometer or two of mileage on these wheels on asphalt each year. The more I think about it, the less it seems like wear will even be an issue.

I'll propose that we simply buy AM mecanum wheels (probably 8-inch) and, if necessary, attempt to repair the rollers at the end of the year by casting hot glue around them. I'm not sure if casting hot glue like this would even work, but if it doesn't, then we can simply buy new rollers to replace the old ones.

Thanks for all the advice.

[Tristan Lall]

That hot glue technique sounds innovative, but definitely consider the heat buildup in the tread when a wheel spins in place (i.e. the wheel breaks loose before stalling the motor—generally a good thing for preserving motors). You don't want the glue to become soft and tacky as a result. Also, if you're running outdoors on asphalt, consider the ambient temperature of the surface, and the added heat due to insolation, especially in the summer. Again, you don't want a sticky mess.

Why not get in touch with AndyMark, and ask them about their durability tests on their rollers? (Andy might respond right here, if you're lucky.) See if they can estimate the wear for you, based on their tests and your usage patterns (e.g. mostly driving forward, with occasional sideways maneouvres, in moderate temperatures on asphalt, with a little bit of mileage indoors).

Another problem with soft rollers in general, and hot glue in particular, will be wheels picking up abrasive material from outdoors, and dragging it inside of buildings. You don't want tiny stones embedded in the treads, if you plan to transition from outdoors to a shiny, polished floor.

[tsaksa]

Originally Posted by flameout

Let's assume that team labor is free

I'm really looking into casting hot glue right now. We could make a mold using a pristine roller (we'd order an extra set of 12 to replace rollers that we mess up beyond repair) -- I believe there are commercial products for building reusable molds for stuff like this.


Thanks for all the advice.

I have used hot glue to cast a lot of things, but usually things that do not need much strength or abrasion resistance. If you have built many robots, or similar devices, you probably realize the importance of selecting rugged materials for wheel surfaces and the amount of wear even tough materials experience. And that is just with normal wheels. Mecanum wheels place even more demands on these surfaces. I expect you will find any rollers cast from hot glue to be a poor substitute for more conventional plastics.

If you do plan to try to fabricate your own rollers, try to think about all of the properties that will be important in the finished roller. How much weight will each roller need to support? Can you calculate the PSI at critical load bearing locations? Do you plan to lubricate each roller regularly or do you require some material that is mostly self lubricating such as Nylon or UHMW?

While I admire the desire to save money and try unique fabrication techniques, hot glue does have some significant drawbacks in a number of material properties. It may have good applicability if developing a rapid prototype, model, or even as a finished part if it is being used in a low stress application. But I doubt you will have much luck in this very demanding application.

[flameout]

Okay, so it sounds like hot glue doesn't have the durability to work as tread on the rollers.

In that case, I'll pitch the idea of mecanum wheels to our team with the caveat that it would be a yearly expense.

Thanks for all the help.

[tsaksa]

Originally Posted by flameout

Let's assume that team labor is free

I'm really looking into casting hot glue right now. We could make a mold using a pristine roller (we'd order an extra set of 12 to replace rollers that we mess up beyond repair) -- I believe there are commercial products for building reusable molds for stuff like this.


Thanks for all the advice.

I have used hot glue to cast a lot of things, but usually things that do not need much strength or abrasion resistance. If you have built many robots, or similar devices, you probably realize the importance of selecting rugged materials for wheel surfaces and the amount of wear even tough materials experience. And that is just with normal wheels. Mecanum wheels place even more demands on these surfaces. I expect you will find any rollers cast from hot glue to be a poor substitute for more conventional plastics.

If you do plan to try to fabricate your own rollers, try to think about all of the properties that will be important in the finished roller. How much weight will each roller need to support? Can you calculate the PSI at critical load bearing locations? Do you plan to lubricate each roller regularly or do you require some material that is mostly self lubricating such as Nylon or UHMW?

While I admire the desire to save money and try unique fabrication techniques, hot glue does have some significant drawbacks in a number of material properties. It may have good applicability if developing a rapid prototype, model, or even as a finished part if it is being used in a low stress application. But I doubt you will have much luck in this very demanding application.