Bearings vs Bearing Blocks

[Knufire]

Originally Posted by Monochron

Maybe you could shine a little light on how center to center behaves with a chain that stretches over time. For instance, we have attempted center to center with success in the past, but we have no where near the tolerances the list as "needed". Would better tolerances cause less stretch in the chain? Basically, I'm still skeptical that you can run chain without ever planning to tension it. I have heard of it being done, I'm just not sure how.

5188 ran a chained drive with exact center-to-center distances this past year. We calculated the hole distance for an even amount of chain links, then added 0.018" to compensate for chain stretch. This number was suggested by Paul Copioli and confirmed by testing by 2363.

[DampRobot]

Originally Posted by Monochron

Maybe you could shine a little light on how center to center behaves with a chain that stretches over time. For instance, we have attempted center to center with success in the past, but we have no where near the tolerances the list as "needed". Would better tolerances cause less stretch in the chain? Basically, I'm still skeptical that you can run chain without ever planning to tension it. I have heard of it being done, I'm just not sure how.

My personal experience with c-c designs is limited, and a lot of the info I have is from talking to friends on other teams.

The systems I felt comfortable running exact c-c were all reasonably low torque, in manipulators. They were also overpowered, so efficiency was not a major concern. This eliminated two of my biggest worries about c-c designs.

First concern: That chains/belts could slip, or "ratchet" as they stretched over time. Basically, if they chain is a little loose, and you apply too much torque to the system, the angle on the teeth in the sprocket will push the belt/chain away from the sprocket. If the belt is loose enough and/or you apply enough torque, the chain/belt will actually fully disengage from the sprocket, and the system will slip. I was OK with this in the applications I used c-c for, because we never expected to see large torques in the system, and if we ever stalled the system, it wouldn't be the end of the world if the belt/chain slipped. Of course, repeated slipping is bad for the life of the chain/belt, especially belts. As you're often going to be stalling your drivetrain, need it to have a lot of torque, and really, really don't want your DT to break, I don't like the idea of using exact c-c DTs.

Second concern: loss of efficiency. If you're tensioning super hard on a chain/belt, there's going to be more friction. If you have sliding bearing blocks, you can dial this tension in, but if it's an exact c-c system, what you see is what you get. I was OK with potentially having a lot of friction in the system because it was overpowered for what it needed to do.

If you're belt/chain is too loose, you run into concern one. If it's too tight, you run into concern two.

Maybe I'm misreading your questions, but c-c tolerances don't directly effect chain stretch. Sure, if your c-c distance is too big, your chain will likely stretch over time, but that won't necessarily be a bad thing if your system is overtensioned.

Basically, tighter tolerances get you closer to the goldilocks zone of between concern one and two. If your application is very demanding on both sides (like DTs), you will need better tolerances. If you're OK ratcheting sometimes or losing efficiency (like in some types of intakes, for example), a c-c solution may make sense.

I don't mean to blast c-c designs. If your team can pull them off for DTs, awesome! They can be much lighter, and certainly are more simple. When I built them, I really liked them. I just didn't trust 100 to be able to pull off a perfect c-c DT when I was on the team, and doubt that the risk/reward calculus makes sense for most teams in FRC.

[Chris is me]

In terms of viability of an exact center drive design, belts and chains can't be directly compared. 25 chain absolutely does stretch over time (and sprockets wear) and thus an exact center chain drive is not always viable. In a WCD, the small sprocket sizes use combined with the loads involved make exact center chain drive a bad idea.

Exact center belt drives are a lot more viable. Belts will not stretch in an FRC robot's lifespan. If you can machine with decent accuracy, you can hit the tolerances required. If I had to make up a number, I would say +/- .005", but really it's just never been a problem for my (former) team. We just CNC the drive tubes to exact center distances and it's good enough. Basically, if you have a CNC mill, there's no reason you can't do an exact center belt drive if you wanted to.

In fact I think it's easier to mess up tension with a sliding block belt drive than an exact center drive. Exact centers are probably better than the adjustment you can do by hand, and it's easy to over or undertension a belt. I think sliding tensioners for belt drives are almost strictly worse than exact centers in my experience. Counterintuitive, I know.

It is possible to overload a belt, causing ratcheting or belt failure. A rule of thumb is for 24T pulleys or smaller in a drivetrain, you will need 15mm wide belts. The combination of 24T pulleys and 15mm belts has served my (former) team well for several seasons, not once ratcheting, failing, or otherwise ever needing maintenance at all.

Other than for retention purposes I don't think the bearing holes have to be within .002" of perfect to work for exact center belt drives - that tolerance is probably a bit tighter than required. Still not hole saw tolerances though.

[asid61]

Originally Posted by Chris is me

In terms of viability of an exact center drive design, belts and chains can't be directly compared. 25 chain absolutely does stretch over time (and sprockets wear) and thus an exact center chain drive is not always viable. In a WCD, the small sprocket sizes use combined with the loads involved make exact center chain drive a bad idea.

Teams that do chain-in tube use fixed C-C regularly. The way the sprockets are set up to force the chain against the side of the tube may account for this, but I think it's application-specific.

[1493kd]

I can attest to the simplicity and robustness of exact c2c belt drive. Chris's old team 2791 was kind enough to walk our team thru its construction and design this past season. We made use of RPI's cnc and turned out by far our best drivetrain we have ever had. Trust me in the past 1493 has built some of the worst drive trains in the history of FRC and I dont think we will be changing from belt in tube c2c for awhile.

The ability to get the hole spacing correct is 99% of the challenge.

[Monochron]

Originally Posted by asid61

Teams that do chain-in tube use fixed C-C regularly. The way the sprockets are set up to force the chain against the side of the tube may account for this, but I think it's application-specific.

Yeah this is what I had in mind when I asked. How is the chain not stretched so much that it needs to be replaced?

[DampRobot]

Originally Posted by Monochron

Yeah this is what I had in mind when I asked. How is the chain not stretched so much that it needs to be replaced?

I have yet to see any really good explanations of chain stretch. That doesn't mean no one understands it (if someone does, please jump in!), but that for the most part the FRC community is interested in phenomena instead of root causes. (Never use scissor lifts, never use steel, always active maintain control of game pieces, etc. are examples of phenomenological "rules" in FRC. We are usually content with fuzzy understanding of the physical motivation behind things we observe to be true. Before anyone gets super mad at me, I'm pretty guilty of this too.)

From my experience, chain stretch has three components. One dominates over the short term, but is dominated by the other two effects over the long term.

First, there can be actual physical stretching in the chain, in the sense that there's some very large spring constant to the chain. You can probably stretch chains just a tiny bit as you're putting them together, but barring any extreme loading, this is a very small amount.

Second, the chain itself will wear over time. Grease in the tiny bearing surfaces in the chain links will dry up, and the chain links will physically wear so that the link to link distance increases very slightly. Unlike component one, this is a long term effect, and isn't reversible.

Third, the sprockets around the chain will wear. This doesn't have anything to do with the chain itself, but will manifest itself in the same way. Over time, the steel chain will wear aluminium sprockets, and the chain will become looser as the sprockets become very slightly smaller. In the absence of hard evidence, I would guess that this is the dominant effect behind long term "chain stretch."

In any case, I can't see a way in which a c-c design would stretch chain any more or less than a tensioned system set to exactly the same tension. The only difference is how you compensate for the stretch. In a sliding bearing block system, you just re-tension and you're good to go. In the case of c-c designs that asid was talking about, the wall of the tubing retains the chain, which keeps it from coming off or ratcheting on the sprockets, which makes the stretch less of an issue.

[Monochron]

Originally Posted by DampRobot

In the case of c-c designs that asid was talking about, the wall of the tubing retains the chain, which keeps it from coming off or ratcheting on the sprockets, which makes the stretch less of an issue.

Hhm, this could be a key part of the root cause. Any ratcheting increases stretch. Therefore a system that never ratchets would stretch much slower overtime than one that ratchets ever 50 operations or so.

[AdamHeard]

Originally Posted by DampRobot

I have yet to see any really good explanations of chain stretch. That doesn't mean no one understands it (if someone does, please jump in!), but that for the most part the FRC community is interested in phenomena instead of root causes. (Never use scissor lifts, never use steel, always active maintain control of game pieces, etc. are examples of phenomenological "rules" in FRC. We are usually content with fuzzy understanding of the physical motivation behind things we observe to be true. Before anyone gets super mad at me, I'm pretty guilty of this too.)

From my experience, chain stretch has three components. One dominates over the short term, but is dominated by the other two effects over the long term.

First, there can be actual physical stretching in the chain, in the sense that there's some very large spring constant to the chain. You can probably stretch chains just a tiny bit as you're putting them together, but barring any extreme loading, this is a very small amount.

Second, the chain itself will wear over time. Grease in the tiny bearing surfaces in the chain links will dry up, and the chain links will physically wear so that the link to link distance increases very slightly. Unlike component one, this is a long term effect, and isn't reversible.

Third, the sprockets around the chain will wear. This doesn't have anything to do with the chain itself, but will manifest itself in the same way. Over time, the steel chain will wear aluminium sprockets, and the chain will become looser as the sprockets become very slightly smaller. In the absence of hard evidence, I would guess that this is the dominant effect behind long term "chain stretch."

In any case, I can't see a way in which a c-c design would stretch chain any more or less than a tensioned system set to exactly the same tension. The only difference is how you compensate for the stretch. In a sliding bearing block system, you just re-tension and you're good to go. In the case of c-c designs that asid was talking about, the wall of the tubing retains the chain, which keeps it from coming off or ratcheting on the sprockets, which makes the stretch less of an issue.

You also have all the backlash in every interface reacting the chain tension being pulled out (which can be a reasonable sum depending on the assembly).

In addition to that, all the parts reacting chain tension have a deflection of some amount, which for a "U" shaped frame could be pretty appreciable (or greatly cantilevered shafts).

[Jared]

Originally Posted by DampRobot

First, there can be actual physical stretching in the chain, in the sense that there's some very large spring constant to the chain. You can probably stretch chains just a tiny bit as you're putting them together, but barring any extreme loading, this is a very small amount.

Second, the chain itself will wear over time. Grease in the tiny bearing surfaces in the chain links will dry up, and the chain links will physically wear so that the link to link distance increases very slightly. Unlike component one, this is a long term effect, and isn't reversible.

Third, the sprockets around the chain will wear. This doesn't have anything to do with the chain itself, but will manifest itself in the same way. Over time, the steel chain will wear aluminium sprockets, and the chain will become looser as the sprockets become very slightly smaller. In the absence of hard evidence, I would guess that this is the dominant effect behind long term "chain stretch."

In any case, I can't see a way in which a c-c design would stretch chain any more or less than a tensioned system set to exactly the same tension. The only difference is how you compensate for the stretch. In a sliding bearing block system, you just re-tension and you're good to go. In the case of c-c designs that asid was talking about, the wall of the tubing retains the chain, which keeps it from coming off or ratcheting on the sprockets, which makes the stretch less of an issue.

I don't beleive the sprocket wearing is the main cause of chain stretch. Our elevator and drive chains both stretched over the course of the season, and in both cases, the black anodize coating wasn't scratched off or worn at all. Also, I've noticed that replacing old chain with new chain causes a pretty large increase in tension.

It appears to me that he chain stretch comes predominantly from the fit between the pins and the bushings loosening up over time. If you play with a new and old length of chain, you can feel the difference in flexibility.

[AdamHeard]

Originally Posted by Jared

I don't beleive the sprocket wearing is the main cause of chain stretch. Our elevator and drive chains both stretched over the course of the season, and in both cases, the black anodize coating wasn't scratched off or worn at all. Also, I've noticed that replacing old chain with new chain causes a pretty large increase in tension.

It appears to me that he chain stretch comes predominantly from the fit between the pins and the bushings loosening up over time. If you play with a new and old length of chain, you can feel the difference in flexibility.

With #25 chain, you have a good deal of links even in relatively short run.

.0001" of wear per pin quickly adds up.

[DampRobot]

Originally Posted by Monochron

Hhm, this could be a key part of the root cause. Any ratcheting increases stretch. Therefore a system that never ratchets would stretch much slower overtime than one that ratchets ever 50 operations or so.

I think I confused you on cause and effect. Ratcheting is caused by loose chain, which is a consequence of chain stretch, wear, etc. Chain-in-tube drives help alleviate some of the symptoms of poorly tensioned chain. If they're designed so that the chain is very close to the inside of the tubing, the chain physically can't ratchet without interfering with the wall of the tubing. Since ratcheting is the thing you actually care about in drives (along with efficiency and to a small degree backlash), if you solve the problem of ratcheting you don't have to worry quite as much about the problem of chain stretch causing a loss of tension over time. It's addressing the symptoms, rather than the underlying condition, but that doesn't mean it's not an effective solution.

Ratcheting certainly causes sprocket wear (and a lot of tooth wear on belts), and it's possible that it stretches chain more as well. However, I've almost always heard it described as an effect of loose chains/belts, not visa versa.

[FrankJ]

Chain stretch is a slight misnomer. The links don't stretch. Or more accurately the load needed to deform the links is very close to its breaking load. What you are seeing is wear in the pin & bushings. During break in the high spots wear quickly which gives you the rapid initial stretch. You take a length of worn out chain and lay it on a flat surface and push in & out you will see a lot more movement than in a new chain. Sprocket wear also looks like chain stretch.

Ratcheting is caused way too loose chain, not enough chain wrap around the sprocket, or the sprockets being in adequately supported (allows the CC distance to shrink under load.

[Monochron]

Originally Posted by DampRobot

I think I confused you on cause and effect. Ratcheting is caused by loose chain, which is a consequence of chain stretch, wear, etc.

I was just suggesting that frequent ratcheting could cause chain to stretch faster as it is a big shock load on the chain. Basically a system already in a bad state, accelerating further into that state.

If they're designed so that the chain is very close to the inside of the tubing, the chain physically can't ratchet without interfering with the wall of the tubing.

Ah, so the chain could be very loose but it doesn't matter in that case.

[Munchskull]

So from my understanding C to C chain is a bad idea?