#25 Chain
 

Hi Folks,

Typically we use belt for our drive train. This year with the bigger wheels we are decided that we wanted a stronger transfer method so we switched to #25 chain. We are running 8" AndyMark Pneumatic wheels, with a WCP Dual Speed Gearbox. Our low end is about 6 fps and the high is about 12 fps. We keep breaking chain. Is anyone else having problems with the large wheels breaking the chain? Some on the team want to switch to #35 chain, but that seems like overkill to me. I feel like we must be doing something wrong, not that the chain is too weak.

Thanks!

Re: #25 Chain
 

Alignment, alignment, alignment.

#25 chain needs to be aligned properly.

NO HALF LINKS!!!!

That is how we are using it this year and so far, so good.

BTW, #35 will work and is not overkill.

Re: #25 Chain
 

#35 Chain is definitely not overkill! CIA has used chain drives in the past and #35 is the way to go, especially if you get in a pusing contest with a defending robot. Also, how big are your sprockets? Using small sprockets increases the stress on your chain. Simply using bigger sprockets lightens the load on your chain.

Re: #25 Chain
 

The best way to fix #25 chain is to use #35 chain. People claim that if you use it correctly, #25 is perfectly fine. It's that first part that's hard. If you have the weight, I highly reccomend #35.


*Disclaimer: I hate #25 chain with a fiery passion because it cost us over 12 hours during RI3D this year. We would've been done a solid 12 hours earlier if we had used #35.

Re: #25 Chain
 

Yes, we made the switch to 35 about 3 weeks ago. While prototyping different size wheels, we kept breaking chain even without master links. Most of it was impact driven - as we hit obstacles, it shock loaded the chain. We've been driving a lot with 35 since and no problems - it's also more resilient to tension/slack.
At a scrimmage last week, we talked to another team that had switched. They had actually calculated the shock loading and it exceeded the 25 strength by a fair amount.

Re: #25 Chain
 

We started off running 9mm belts inside of our tubes running 8in pneumatic wheels which caused too much slippage so then we swapped to #25 chain like you and haven't had any problems since. The only time that one broke on us was when we had a power-pull get trapped in the drivetrain. Our drivetrain is even faster going at 20fps and it has been great. I'm not quite sure what is wrong with your setup but #25 chain should be able to handle the loads. Maybe a picture would help diagnose the problem.

Re: #25 Chain
 

We broke a #25 pitch chain this year for the first time.

Robot is geared for ~15ft/s, 8" pneumatic wheels, ~115lbs actual weight. We were coming over a defense dukes of hazard style and the driver slammed it in full reverse as the front wheels were landing and point loaded them, causing the break.

All of our chains were run chain in tube with dual bearing supports, however there was an issue found in the post mortem where aan additional 1/16th spacer got added, which caused 1/16th misalignment on an 8" CtC run.

Bot #2 (comp) was made with #35 pitch chain as an insurance policy since we had the weight allowance. We fixed bot #1 and left the #25 pitch on it. After a week zero event and ~5-10 additional hours of driver practice under extremely heavy defense, we haven't broken another #25pitch.

tl;dr - Make sure your sprockets are well support and well aligned and you shouldn't have an issue with #25 even in the most abusive drivetrain conditions. Want some extra insurance, run #35, but it still isn't an excuse to not pay attention to alignment and tension. It always pays to do the math though and compare it to the ultimate strength of your chain for edge cases.

Re: #25 Chain
 

At first blush, it seems that #25 chain should be able to provide more torque to the wheels than traction from the carpet (unless you are using teeny tiny sprockets) BUT, with drive systems, I don't mess around. There is nothing that lowers your draft potential as fast as driving in circles for a match because your robot threw a chain.

I recommend #35 chain unless you are completely painted into a corner and there is just no way to make it work.

YMMV.

Dr. Joe J.

Re: #25 Chain
 

Lots of things that could be going on here that could cause these issues, depending on a number of factors. What is the tooth count of your sprockets? How are you tensioning your chain? How long are your chain runs?

First things first, if you're breaking 25 chain in a drivetrain yes there's probably something you're doing "wrong" with it, but switching to 35 will give you a lot more fault tolerance so you don't necessarily have to solve that problem to keep going. In the interests of time and reliability, if you have the weight 35 chain might just be the easiest way to go.

I'm assuming you guys don't have any problems with alignment - that would be pretty easy to see.

If I had to wager a guess, if you're using 16T or 18T sprockets (the "standard" for 4" wheel WCDs for the past several seasons), but driving 8" wheels with them, this is probably part of the problem. Smaller sprockets with bigger wheels will be downstream from a larger reduction to get the same speed, so they are loaded with more torque. Shock loading is also more of a concern this year than other years. If you can make your sprockets larger, you will have less tensile force in your chain. In a WCD, you should be able to go up to at least 22T if I recall correctly.

Chain that isn't properly tensioned will have more problems - you're more likely to skip a tooth and shock loading will be greater, leading to a larger chance of chain failure. Chain on small sprockets connected to large wheels will stretch more quickly than chain on large sprockets or connected to smaller wheels, as a function of the increased load put on them. If you did exact center spacing with your chain to start, it might have stretched too much under load by now.

Re: #25 Chain
 

Search through the past Chief Delphi discussions about the relative merits of #25 and #35 chains and you will find plenty of facts, experiments, opinions and anecdotes.

One thing that should become apparent during your search is that your machine is a system made of many parts, that it is operating in varying environments, and that it is being asked (by its operators & software) to do more than one thing. That makes it easy to reach the conclusion that you have to ask the right question (one that takes into account all the important independent variables) before you can answer whether #25 or #35 is the right chain for your machine.

I don't say this to make the topic sound too hard. I say it to help you quickly reach a valid, complete answer. Answers to incomplete questions are incomplete answers, and letting incomplete answers churn your design is not good.

Bottom Line: As is often the case, the right place to start is by learning what has already been recorded; so that you can stand on the shoulders of the ones who have gone before you.

Blake

PS: Some engineers and shade-tree mechanics might just decide to do a quick experiment. That would be adding some dead weight and/or rotating mass to your robot to approximate the effects of switching to #35 chain+sprockets, and then (using the current #25) deciding if the robot is likely to perform well enough when it has to move the the more massive, and stronger, #35 chain+sprockets. If it does still perform well enough, switch to the stronger and "more forgiving" chain this season (before next season do more sophisticated experiments).

PPS: There is more than one type of #25 chain ..., and more than one type of #35 chain.

Re: #25 Chain
 

Sounds like you have a physics problem on your hands. We know torque = r x F, where r is the radius to the point where the force is applied, and F is the applied force.

No matter how you drive the wheels, if you desire a specific speed, the required torque applied to the wheel is a constant. Knowing this, we have a couple options to either reduce force on the chain or increase the maximum force the chain can withstand.

1. Increase the radius of the sprocket. This will reduce the force on the chain, and with a little algebra, you can calculate how much the force will be reduced if you so choose. Additionally, this will give more teeth for the chain to "bite" on, so the potential to put a ton of load on a single link pin is somewhat reduced

2. Upgrade to #35 chain. I say upgrade because if you have the weight for it, #35 chain is a fantastic solution for drivetrains. This will increase the maximum force the chain can withstand, and lessen the chance that the applied force on the chain will exceed the breaking point. This can also help you to lower your center of mass slightly, which, depending on the shape and height of your robot, could be a nice improvement.

3. Combine 1 and 2 by switching to #35 chain and increasing the size of your pulleys. Since you have to buy new sprockets anyway if you switch, why not get the largest radius that will work for you without creating clearance issues? You can decrease the force on the chain and at the same time increase the max allowable force for the chain.

You could also experiment with properly tensioned belts of appropriate width, spacing, sprocket size, etc, but I honestly do not know enough about belts off the top of my head to give a good suggestion for you.

Additionally, like other people have said, you will in any case do well to make sure that everything is properly aligned and tensioned without any half links. I know that when my team did some seriously tough drives during my high school years, we didn't even use master links, let alone half links. If you invest in a good chain tool, you can make complete loops of chain for your drive, and by removing master links you can remove another point of failure from your system, provided you properly use the tool to reassemble the chain.

EDIT: While I was typing this, some other people beat me to the punch. But nonetheless, if you can do it, #35 chain is awesome.

Re: #25 Chain
 

To be fair, your Ri3D robot had some massive, exposed chain runs that are exactly where using #25 chain would cause issues.

This is a good case example: #25 chain needs to be used with care and precision. If you aren't confident in your ability to provide that, or want to push your chain to its limits (small sprockets on big wheels, lots of torque or speed), you really should be using #35, just due to the fault tolerance.

This year lends itself well to very fault-tolerant, robust robots. While #25 can live up to that challenge, it requires a lot more care and effort to make it do so, especially after build season.

Re: #25 Chain
 

What is wrong with half links? Are they extremely weak?

Re: #25 Chain
 

I'd check your setup and your tension. Now I see all these people up here saying that it's an easy solution to switch to #35 chain and be done with it, but that's not fixing a fault of #25 chain, that's trying to cover up poor design and assembly, and in this endeavor, you should always work to have the best setup you can. Loads of highly successful teams use #25 chain in their drives and are likely using it more strenuously than you are. You can run #25 chain on tiny sprockets with sub-optimal tension and it's still going to run for the most part fine, as lone as your alignment is correct. Getting completely straight alignment of your sprockets is key. Use CAD to make sure you have correctly made spacers with little wiggle room to fit everything where it needs to be. Use the location of one sprocket in your CAD to drive the position of its partnering sprocket, and then design your spacers around that. After everything is aligned, make sure you have adequate tension. If you don't have tensioning blocks built in, you can go for other ways of tightening your chain (zip ties are a decent last resort to keeping tension in a system). If you're going C-C, use Paul Copioli's chain calculator in his Useful Calcs spreadsheet. You shouldn't just look at poor execution of #25 chain and think the solution is to move to #35.

I agree with Joe: I don't want an alliance partner that spins circles in matches because of lost chain. However I also don't want an alliance partner that covers up poor execution in design, because they will likely cover up poor execution in strategy and performance, which is the last thing I need in the finals.

Re: #25 Chain
 

We were having a lot of trouble with #25 chain during our Week Zero event in Merrimack, and we too have been thinking about making the change during some of our free time.

What we did to fix the issue for now is adding a 1/4" Lexan strip along the bottom of our chassis where the chain was hanging. We think it was hitting and catching on the defenses. After we did that, no more shearing. We also added some static tensioners, and are looking to put more long term dynamic tensioners on at some point.

Alignment is key for sure with #25 chain, or chain in general. Definately make sure you do your chain calculations right.

We have done #35 and #25 in the past on our drive trains, but none of the existing team has remembered a bot we made using belts. I do not know if that is a better idea, I know slippage is an issue with belts.

Re: #25 Chain
 

As a team that has successfully implemented #25 chain in numerous situations, including both drive-trains and serious arm mechanisms, we switched to #35 this year for our drive after snapping #25 a few times.

A few things I've seen in our own robot:

Alignment is key. A chassis/drive pod/drive assembly must maintain that alignment under all game conditions. This includes thrashing around on a defense. A chain that looks well-aligned statically may not be aligned at all when dynamic loading is considered.

Related: if you're not using beveled sprockets, you're going to have a bad time.

Shock loading is brutal this year. With dramatically inconsistent contact with traction surfaces during defense crossing wheels will spin up, then catch, then spin up again. This dynamic behavior can potentially overload chain.

If you use #25 chain, make sure you're using #25 HD chain. Typical #25 chain is quite weak.

As mentioned we changed our drive-train to #35 this year. We are still using #25HD on our main shoulder joint though.

Extremely weak... maybe... weaker... yes.

Re: #25 Chain
 

Brian,

I wanted to comment on a couple of factors that influence chain selection that had yet to be mentioned. Others have already hit on items like larger wheel sizes, sprocket size, alignment, tension, etc.

I know that in at least one past year, 2791 has run their drive gearboxes at the corners rather than the middle of the robot. This can be great for making space for a mechanism low down in the middle of the robot, but it also means more tension on one of your chains. Instead of having 1 chain carrying torque to each corner from the center axle, you now have a chain from the gearbox corner carrying the torque for BOTH the center wheel and the far corner wheel. Not sure whether this is in play with this year's bot, but I wanted to mention this.

In 2013 and 2014, Team 20 consciously used #35 chain to reduce part count. The machining resources were certainly more than capable of hitting the tolerances necessary to run #25 chain, but would still require a proper tensioning method, likely including cams and bearing blocks. Instead we chose to directly press our bearings into the 2 x 1 x 1/8" wall tubing and use dead spacing + 0.012" to take advantage of the extra lee-way afforded by #35 chain to reduce complexity and part count. In 2013, this was with 6" AM HiGrip wheels, and in 2014 this was with 4" colsons.

We threw one chain in 2013 at least in part due to a brain fart in design (my bad...) which led to using a half link to get the proper number of chain links.* But with a direct drive center wheel, even that wasn't a big deal and was easily replaced between QF matches. No issues were noticed for 5 events thereafter.

*When doing center spacing for a chain application, design for a C-C distance with an EVEN integer number of links. An odd number means you will be using half-links.

Re: #25 Chain
 

First, as someone who's admired Grasshopper Machines for 20+ years, I am going to tell you that you really need to give JamesCH95's experience strong weight.

Almost all my experience with successful #25 chain implementations on drive trains comes before anyone used conveyor belt for tire tread and certainly before pneumatic tires with sprockets that bolt on were only a few clicks away. It may very well be the case that with these new fangled grippy tire options, young'ins these days just have to go with #35 chain for their own safety.

Amen brother JamesCH95, amen.

Dr. Joe J.

Re: #25 Chain
 

I didn't see any reference between 25 and 25H in this thread. 25H is a far better solution without the weight of 35. It is still a bear to work with, but way better than standard 25. And yes, you have to run 25 tighter than you would think. 25 is very sensitive to alignment and tension. Somewhere I was told the "7 tooth minimum" rule. No idea if it is folklore or real but it works in general. Keep 7 teeth engaged at all times. This means no drive sprockets smaller than 14 teeth on a simple drive.

Re: #25 Chain
 

I build silly tiny electric cars and race them. I have run #25 chain for hundreds of hours with motors equivalent to 6+ CIMs to big (by FRC standards, 11") pneumatic tires. and it works just fine even under shock loads like getting put into the wall at 20mph. I even did some testing with 6.5hp brushless motors spinning a 20 tooth #25 sprocket and it worked fine for FRC time scales of less than a couple of hours of operation. With that said, I would swap over to #35 if at all possible. In my experience it is much more forgiving in fabrication tolerances and dealing with the unexpected situations like those that are common in FRC like getting repeatedly rammed, stalled motors, slight shifting of sprockets etc.

Re: #25 Chain
 

...

Re: #25 Chain
 

Personally, I like to overbuild everything. I don't like things breaking, and the degree to which I overbuild is directly proportional to how critical it is for success. The drive train is probably the most critical part of the robot for success. I would never use less than #35 chain for a drive train. Then again, since we made the switch a couple of years ago I'll jump through hoops to avoid using chain at all on the drive train - direct drive all the way!

Re: #25 Chain
 

IMO, #25 chain is great for mechanisms but unless you know what you're doing it's a pain for drive systems. Some teams can make it work, but most underestimate the level of difficulty to get it right. Essentially #35 chain has the benefit of added tolerance for mistakes, so as others have rightly pointed out, the quick fix for #25 chain acting up, is to just use #35 chain.

That said, issues with chains breaking, skipping, and maintenance are the main reason we no longer use chains on our drive system. Geardrive ftw. :rolleyes:

Re: #25 Chain
 

Thanks for this thread. It's very enlightening.

Does anyone have a reliable source for #25HD chain? It's a bit harder to find, from my cursory looks (the vendors we use don't sell it).

Thanks!

Re: #25 Chain
 

Re: #25 Chain
 

#25 chain is very sensitive to alignment, if one sprocket is 1/8" further out than the other, it'll bind up and have problems. Make sure you don't over-tension your chain, as this also makes it more likely to snap. There should be a little bit of slack in your chains, but not enough for the chain to be noticeably floppy. Especially with a game like this, #35 chain is definitely not overkill, and it's much less sensitive to alignment. We're using #25 for mechanisms, but #35 on the drivetrain is definitely a good option.

Re: #25 Chain
 

I would argue that if you can make a drive unit rigid enough to not mis-align gears, and precisely enough to enable proper gear meshing, then you can make a drive that can easily handle no-maintenance chain or belts reliably.

And, yes, no-maintenance (aside from lubrication, which gears need too) chain drives exist.

VexPro only sells #25HD as far as I can tell, by deduction in that they only sell #25 HD master links and state that only HD master links work with their chain. Comparing it side-by-side with other #25 chain sources makes it even more obvious.

I wish they'd state it on their website though...

Edit: boy, Dr. Joe sure knows how to make an engineer blush...

Re: #25 Chain
 

We are setup with 22T sprockets on 8" pneumatic wheels geared at 8 and 18 ft/s off of 2, 3 CIM shifting gearboxes, and #25 completely doesn't cut it.

We will be spending all of Thursday at week one tearing apart our drivetrain to replace all of the chain.

The spacing and tensioning and alignment of the chain was done absolutely perfectly, every chain had good tension without being too tight, it was by far the nicest drivetrain in team history, until we tried to go over defenses and chains started snapping.

Use #35 if at all possible, #25 really burned us this year.

Re: #25 Chain
 

I believe if you are running the 7.56" wheels in a total reduction that gets you 6 fps, I think that operates outside of the recommended range of #25 chain. We've been fine with it this year after beating up three different robots with it over the last 7 weeks but we are running a single speed north of 12 fps on 22T sprockets.

Re: #25 Chain
 

I've come to reiterate most of what has been said in this thread.

4901 ran #25 chain in tube with 17 tooth double sprockets for our drive system. If you were at Palmetto you saw first hand the time we had with it. We ended up taking the drive rails out 4 times.

What we learned is this, DON'T USE HALF LINKS. We ended up with a problem where we had to use half links to get our chain tension-ed without using a tension-er. We would break chain in 5 minutes after installing it again.

I will say if you use #25 chain is that making your chain runs continuous via Dark Soul tool or whatever other method you know of is a great idea. We managed to do so with the help of 1296 and never broke a chain again after that, of course we had to use our pre-planned tension-er slots after all.

Re: #25 Chain
 

Whaaaaaaat you don't have a 221/DarkSoul tool?
#25 is a total dream if you use it right, we used it wrong once after a couple former mentors tried to sell some fiction one year and of the 4 chain runs we had on that robot at Alamo in 2014, we threw all of them at least once. In 2014 because of snow we were allowed to bring in 45 pounds of withholding so we left the drivetrain out of the bag before Virginia and rebuilt the whole thing.

Every other time the chain has been properly tensioned and spaced and we have loved it.

Re: #25 Chain
 

We don't have one. We just bought one Yesterday. I was only brought into the light about that tool this week.

Re: #25 Chain
 

The recommended working load of #25HD is around 90lbf. This is fine to exceed for shorter lifespans, but with the tiny sprockets teams are running this year... :eek:

In round numbers:

8in wheel diameter
2in chain wrap diameter
µF = 1.3
Robot weight = 140lbf

Assume:
All weight on one wheel (possible, if not likely, during defense crossing)
Drive is traction-limited (certainly the case at 6ft/s with 3 CIMs per side!)

Chain tension:
140lbf*1.3*8in/2in = 730lbf

Breaking strength of #25 is around 780lbf, giving one a FoS of 1.1, and virtually guaranteeing yielding in the chain.

This case is quite conservative, but being conservative in drive-train stress considerations is probably a good idea.

For consideration: in our chain-snapping 2016 drive train the calculated tension was 550lbf. Changing to #35, with a breaking strength of about 1900lbf, did the trick. In prior years we've had chain tensions in the 150-200lbf range. Having a FoS of 3+ versus chain breaking strength is probably a good idea.

Re: #25 Chain
 

My philosophy, especially when it comes to drive trains, is, "do the math, then use the math to go for total overkill."

The importance of robustness in a drive train cannot be overstated, literally--if your robot can't drive around, you can't compete***, no matter how awesome your upper mechanical systems happen to be.

***Except last year, maybe. But let's not talk about last year, like, ever. Let's file it away in the Mental Folder of Things that Do Not Actually Exist, like HIGHLANDER 2. Ahhhh, that's better.

Re: #25 Chain
 

I think I recall a similar thread a couple of years ago which went in a different direction, with a lot of folks emphasizing the qualities of #25. My team has been using #25 for last several years with no problems, and we used it on our drive train this year. I guess we will see how well it holds up next week. Maybe I should bring sullied to do a "quick" swap in AZ if necessary.

Re: #25 Chain
 

Looking at this site here: http://www.partsforscooters.com/PFS-Chain-Chart

I was able to see a difference of thickness between #25 and #25H chain (roughly .24mm per plate, for overall thickness of about 1mm).

I am looking for a side-by-side tensile strength comparison to determine if it's truly worth making a jump from mcmaster #25 to a #25h change. If I find one, I'll post it here.

Re: #25 Chain
 

It is tough to find good manufacturer data for 25 series chain. Have a look at the table at the bottom of page 1 here.

The max allowable load column is synonymous with max recommended working load. Converted to lbf, it is 165 lbf for #25, 242 lbf for #25H. Average tensile strength is 992 lbf and 1323 lbf, respectively. These values are a little higher than I recall from past research, but at least gives you a side-by-side comparison.

We have used #25 chain successfully in drive and manipulator applications since our rookie year in 2004. I would definitely recommend #25H chain over #25 in all applications, the weight penalty is marginal for the added strength. I generally avoid #35 chain at all costs, due to weight, but if you are having failures with #25, I would say switching to 25H alone will likely be a marginal improvement - try to increase sprocket size as well. If you have the weight, consider #35 for the added reliability.

Drivetrain failures are are the worst possible failure point, as it leaves you useless on the field. Whenever possible, I encourage direct-driving at least one wheel, so that no chain is a single-point-of-failure to half your drive. If you must have chain directly from your gearbox, then I would consider #35, as failures there are unacceptable. This year, with 8WD using 8" pneumatic tires, we chose to direct drive one center wheel, use #35 to the other center wheel, and #25 to the outer corner wheels. The idea is a compromise between weight and reliability - no #25 chain failure will leave us unable to drive reasonably well, as it would only cut power transfer to one corner tire.

I think a huge part of the problem people are seeing is the dramatically higher forces involved with the larger tires paired with smaller sprockets. For years FRC drive trains have been evolving toward smaller wheels and less gearbox reduction, which pairs favorably with transitioning to #25 chain. In James' example case, if you were instead running 4" wheels, the max possible chain tension would drop to 365 lbf, and SF would increase from 1.1 to 2.1. I agree that his scenario is conservative, but I think most teams (running live-axle WCD, at least) are using even smaller sprockets than he calculated. Consider that the largest hubbed #25 sprocket Vex sells is 22T, with a pitch diameter of 1.75".

Re: #25 Chain
 

Where did you find it? Everywhere I've looked is sold out.

Re: #25 Chain
 

A google search for a "221 chain tool" (that's what we call it since all of ours are 221 branded) will lead you to 221's page for the item which they no longer sell....

...but the page has a link for the tool sold elsewhere and is available from that seller.

Re: #25 Chain
 

We usually use #25 chain in our drive, but not this year. We decided not to push our luck with 8" wheels and a field that promises to provide more drive train abuse than usual.

Re: #25 Chain
 

Thanks for the tip on how to google... Looks like Dave's Motors got them back in stock since I last checked their site.

Re: #25 Chain
 

Wasn't trying to push any buttons, just telling you how I got there since I didn't even know how to search for the current supplier's tool without knowing the name.

:)

Re: #25 Chain
 

I agree with your diagnosis but you lost me there on the recommended treatment.

Are shafts somehow immune to problems? Live axles bring there own problems. Gear teeth are not infinitely strong. And so on.

Make sure you have a reliable drive train. That is the take away. There are a lot of ways to get there, but you really need to get there.

Dr. Joe J.

Re: #25 Chain
 

True, though gears are slightly more tolerant to misalignment along the shaft axis than small chains (you just risk damage to the gears if the misalignment is too extreme). Our problems in the past usually stemmed from getting chain that had inconsistent tension (possibly due to using old chain?) and making shaft spacers for the wheels that were the correct size. As for lubrication, for external (not inside of a gearbox) gears we just spray on Teflon dry lubricant during assembly and don't worry about it again for the rest of the season.

Of course I never meant to imply no-maintainace chain systems weren't possible, just harder get right with smaller chain.

Interesting, I didn't know Vex sold #25HD. Guess next time I'll have to put an order in for some and replace all the stuff we have that's like 10 years old that used to come in the KoP. :rolleyes:

I think the point he was trying to make was that typically having a direct drive wheel eliminates total drive failure due to a chain or belt breaking. I also think that for most teams, you're FAR more likely to have a failure in a chain or a belt than you will with a gearbox output shaft (if you do, you've done something terribly wrong).

Re: #25 Chain
 

I was looking forward to switching to dry lube similar to this this year, given its the first time in many years that we are running open gearboxes. After bad signs of premature wear on the practice bot, we have played it safe by sticking with Red & Tacky, and adding covers to keep the grease from getting flung everywhere.

I'll have to make it a goal in the off-season to properly test a dry lube, so I can have confidence in it.


Chris pretty much has it dead on. Direct-drive of course does not eliminate all possible failure modes, but it at least eliminates one. If you are designing for live axles already then direct-drive is almost trivial.

Of course very reliable dead-axle drivetrains are possible without direct-drive.

Attention to design details, such as well-supported shafts, ridgid/robust frame construction, careful component selection (chain/belt type, sprocket/pulley diameter, etc.) and reliable method of chain tensioning are all still critical for success, no matter what drive style you choose. In this era where there are a selection of proven COTS gearboxes to choose from, I would generally agree with Chris that most teams will have more issues with assembly and implementation than with the gearbox components themselves.

We have run both live and dead-axle drivetrains, with varying degrees of success. In my time in FIRST, we have experienced:
- Bent drive shafts (2007, poor steel grade selection)
- Stripped gear teeth (2012, inadequate lubrication, fortunately happened post-season)
- Sheared shafts at poorly placed E-clip grooves
- Set screw failure (2007, never use set screws and D shafts in reversing load situations, and never in drive trains!)
- More chain tension issues than I can count
And many other drive failures ranging from nuisance maintenance items to crippling season-killers.

Experience can teach you what not to do, but it also proves that there are many ways to get it right.

Re: #25 Chain
 

We use, and love, Chain Saver as a chain and open-gear lubricant.

http://www.amazon.com/DuPont-Teflon-.../dp/B00KMMFE8Y

Can be found in may places, powered by the magical beast of Teflon. Spray it on, run things around a little by hand, then a few minutes later it's dried out and ready to go.

Re: #25 Chain
 

[quote=JamesCH95;1550502powered by the magical beast of Teflon[/QUOTE]

... and Moly B. Denum

Re: #25 Chain
 

Of course! How could I forget dear Moly...

Re: #25 Chain
 

Back on the Chain vs. Live Axle. For a lot of teams, the majority I think, Live Axle is just not going to result in a more reliable drive train. Yes, there are some advantages, I just wouldn't want to be driving people to make a poor decision for their team based on something that might be right for another team but a disaster for them.

Dr. Joe J.

P.S. I love chain for drive systems. JJ

Re: #25 Chain
 

When I saw the first four words of that sentence I had an immediate flashback to a time long long ago...

Re: #25 Chain
 

If you're worried about chain weight, a local team used #45 chain on their drivetrain in 2014... according to them, it weighed 20 pounds :yikes:

Re: #25 Chain
 

Can you give any examples of failures you get with a Live Axle system that you would not get by just using Chain on Dead Axles? In the 10 years I've been doing FRC I don't think I've ever seen a failure on a Live Axle, I have seen plenty of broken chains though.

What specifically do you see is the downside of a Live Axle system?

Re: #25 Chain
 

Bent axles are one failure I've had from past years. But I suppose that is more of a problem with cantilevered live axles than live axles alone.

The gear train to get the torque from one axle to the next is another point of failure. Many multidrive live axle systems need a bunch of gears betwixt & between the live axles. More gears, more axles, more gear to gear interfaces, more center distances to get wrong (or have to manage at a minimum).

Typically gears have smaller working radii where the force is transferred from axle to axle than chain (not always but often). This in turn means higher forces for a given torque, which isn't unmanageable (obvs) but can be tricky.

Back to failure live axle failures, the most memorable one for me was when in the Soccer Game where we had to go over the bumps (well you didn't HAVE to but we did). The team I was with then had an 8 wheel beast with an AndyMark shifter direct driving one of the axles on each side and chain in between as some have suggested.

We were in a good place with a win in our first match of the Quarter Finals at the Long Island Regional. During that match we went over the bump a little more energetically than we had up to that point in the tournament (think Dukes of Breakaway). We won the match and thought everything was fine - in fact we were thinking we had some great footage for our highlight reel at the end of the season.

As it turned out, match 2 didn't go so well because we just couldn't get going in high gear. Between matches we checked everything out, couldn't find a problem. We put the robot back out there for our 3rd match only to get the same result: High Gear was useless. It wasn't until we got back to the school that we realized that we had it so hard we had skewed our frame a bit which skewed the gearbox a bit which was enough to make the current go up in high gear to the point that our breakers kept tripping on that side. Loosening some screws, giving the robot a kick or two, and tightening those screws again had us running like a champ again (of course if we were going to compete again we would have had to solve that problem but it was our last tourney that season).

So, was that a live axle problem? I kinda think so but others may disagree.

Dr. Joe J.

Re: #25 Chain
 

I will certainly grant you that as an issue with cantilevered live axles, or really any cantilevered axle for that matter. This is also why I think teams that build WCDs are crazy, but whatever works for you I guess. ;)

I would agree that using live axles in combination with gears does add more points of failure depending on the drive configuration (our team has learned ways to mitigate this over the years to the point it's largely a non-issue). That said, it's also certainly possible to use chains along with live center axles to put some redundancy into the system without changing much in the way the chains are run.

I would personally say that using live axle gear-drives is NOT the best thing for most teams to do without careful research, planning, design, and precision machining. Using live axles on chain drives, however, is a very easy thing for just about any team to do (in fact the KoP chassis ships in this configuration, but with belts instead of chain) and, in my opinion improves the overall durability of the drive system.


2010 was a brutal year for drive systems. Going over the bumps between zones created problems even for us. We had to replace a number of the 3/8" steel shafts on the dead axles wheels in our drive system that year because they kept bending. Never had any problems with the two live axles though, so could have been an issue with your specific drive configuration or just bad luck. No drive system is completely without it's issues or immune from bad luck unfortunately. :rolleyes:

Re: #25 Chain
 

So suppose that a team did use #25 and suppose they HAD TO have half links, what are the strongest half links on the market?

Also can you replace #25 chain with #25H chain with out changing sprockets?

Re: #25 Chain
 

I don't know what the real answer is, but my team ran into this exact situation with our practice bot and we used half links from mcmasters as well as some half links from amazon. I would definitely recommend the mcmasters ones. No pesky cotter pins to deal with, and besides we broke the amazon half links the moment we began ramming into our wooden practice defenses. Meanwhile, the mcmasters half link held while another full-link failed elsewhere on the chain.

Re: #25 Chain
 

THIS TYPE of Half Link:



ARE MUCH BETTER THAN THIS TYPE:



Just don't even think about using the cotter pin type. Please.

Dr. Joe J.

Re: #25 Chain
 

Cosigned. Not even once.

Sincerely,

Lessons Learned In Week 0.5, inclusive of pre-bag time.

Re: #25 Chain
 

Hey Guys,

Can you please expound on "why". Where is the failure point with the cotter pin type.

Re: #25 Chain
 

I'll expound: the cotter pin fails, and then the whole thing flies apart spectacularly.

Okay, maybe not spectacularly--it just kinda dumps out the end of the tube. But at least on double-sprocket applications (like the 221 sprockets we're using), the cotter pins are more likely to hit something and fail. It's worse if you don't think about it and don't put the pins facing out each way, but the other style just avoids this problem entirely and makes things that much more idiot-resistant.

(I didn't say idiot-proof, because you know what happens then--just idiot-resistant.)

Re: #25 Chain
 

I'm understanding you to say the failure mode is external impact damage to the cotter pin. It's not wear or normal shock loads or fatigue failure. Is that correct?

Re: #25 Chain
 

Since we run chain-in-tube, it's tricky to get a definitive answer here. But from what we can hear and make educated guesses on, external impact damage is a very real possibility. Runner-up would be shock loads; generally speaking, we didn't get enough runtime on the failed ones to make me believe it's fatigue.

We're running a versachassis with 6 8" wheels on 25 chain with standard sprockets (18t?) and haven't had any issues after a full competition with plenty of defence crossing and hard defence. We did have one failure during practice due to a tensioner backing out but it was an easy swap.

Re: #25 Chain
 

The cotter pin is part of it, but the link is just not as strong.

The zig-zag plates in an "offset link" are set up to have complex loads on them (unlike standard links that have plates that are in very simple tension loads by design - assuming your chain sprockets are aligned properly). Anything you can do to help support that load is going to make the chain stronger.

With the cotter pin type offset link, the pin is a slip fit (the pin that carries the load, not the cotter pin). This type of joint gives less support to the zig-zap plate than in the other type of offset link, where the pin is smashed/mushroomed so that the pin and the plates are more effectively joined. Again, that better joint between the pin and the zig-zag plate translates to a stronger offset link.

And this is completely separate than from the whole cotter pins are subject to mis-application when they are so small and jewelry like.

Dr. Joe J.

Re: #25 Chain
 

@ Billfred, Dr Joe: thanks for the added detail.

Re: #25 Chain
 

Very confusing. Both pictures you show are under the same part number? #6261K105

Re: #25 Chain
 

Yes, sometimes McMaster links to a family of parts page rather than a particular part page. No have why.

But, when you get to the point where you actually put the part in your cart, you can usually look at the CAD link and see what the part really looks like.

The photos are there so folks can know what I was talking about without following links. I did not intend to put this or that part number into play.

Dr. Joe J.

Re: #25 Chain
 

I don't know where Dr. Joe got the pictures, but 6261K105 is the one you want. It has the former's arrangement, even if the picture is of the latter. (I should know, I ordered 15 of them. Whether they arrived in time for Palmetto, welllllllll......)

Re: #25 Chain
 

Moly is a magical woman

Re: #25 Chain
 

Shall we do a post-mortem on our drivetrains this season?

We ran into issues with our #25 drive chain and switched to #25H, no problems since.

Re: #25 Chain
 

We (4901) had issues with our #25 chain early in the season (at Palmetto) thanks to poor quality half links. We switched to continuous chain runs made with the dark soul tool with no half links and manually tensioned it with bolts and tensioner slots. Didn't break a chain since then. I should also mention our controls mentor Ryan implemented a SLEW control onto our drivetrain to make it so that the drivers couldn't slam full forward to full reverse instantly. That was a suspected issue causing chains to snap.

The second team I worked (5632) with continued to use half and master links. They used the "good kind" with the non-cotter pin links. They broke chain very often. They tried upgrading chain type within the spectrum of #25 chain and it only made the problem happen less often instead of solving it completely. We started with Vex Pro chain. It didn't even last 5 minutes from what I was told during testing. We upgraded to a Browning (I think) HD chain and that lasted 10+ matches normally.

Re: #25 Chain
 

Repeatedly people keep mentioning keeping chain tight, which is especially important for #25 chain.

I've looked for a COTs chain tensioner that works like the tensioning device for a chain-saw bar and never seen such an item sold. I want to be able to loosen a locking nut, tighten the chain with a screw, then re-tighten the locking nut.

Does someone make a COTS robot device that can move the axle forward or back by turning a screw? Motorcycles have a similar set-up.

If you google 'roller chain tensioner' you get mostly all idler-wheel add-ons.

This would be an easy item to design and offer for a vendor and for 'dead' axles, you'll have to buy two for each axle bolt that needs adjusted, so potentially, eight on a six-wheel robot. That adds up profit fast.

Re: #25 Chain
 

The vex tensioner blocks do this with a cam. I would recommend this method as it's very easy to use.
http://www.vexrobotics.com/vexpro/al...gblocks-g.html

Or you can just wrap a zip tie around your chain run in the middle to tension it. No, I am not joking about this. It usually lasts for entire competitions as long as you are using quality zip ties.

Re: #25 Chain
 

5188 ran a #25H chain drive, with 8" pneumatic wheels and no tensioners, and it was bulletproof. We ran on a four CIM drive with a 11:72 reduction, then a 18:42 chain run to the wheel. We also added 0.018" to all the chain center-to-center distances. Chains were continuous runs made with the Dark Soul Chain Tool.

Re: #25 Chain
 

We ran straight #25 in a pretty bad way; direct center-to-center with sprockets outside the tube. That being said, we only threw a single chain the entire year, at SVR, due to the looseness getting extreme because of some screw holes widening (haven't checked it out for sure yet though). We did not break any chains. We used the Dark Soul chain tool to rejoin it, which I think helped a lot; we've killed masterlinks in the past.
That being said, if I had to do it again I would use sliding bearing blocks for tensioning and run 25H or 35 chain, just from fear. I was in a constant state of fear. :D

Re: #25 Chain
 

6 CIM 2 speed 8" pneumatics 6WD Versa WCD with 25H chain on 18T sprockets. Zero issues.

Re: #25 Chain
 

After our fifth or sixth event, we found a few sprockets starting to fail. Chain was probably under-tensioned for awhile contributing to this.

25H chain, 22T sprockets, 6" wheels. It's the defenses that loaded them like this really.

Going to look into belts next year, since our failures weren't so much chain strength as much as not keeping up and constantly adjusting tension. Belts don't stretch.

Re: #25 Chain
 

They do slip and break though... :rolleyes:

Re: #25 Chain
 

A very important lesson that we learned this year is that Belts DO stretch, but it is a little nuanced.

Belts DO NOT stretch appreciably over time. ie. if your spacing is setup properly you will not need to add tension over time because the length of the belt will be constant.

Belts DO stretch instantaneously. If you shock load the belt hard enough it will flex outward in such a way that it can stretch slightly. Possibly enough to skip some teeth. I have some high speed video of this if you want to verify.

Chain DOES stretch appreciably over time. It will need to be cared for.

Chain DOES NOT stretch instantaneously in an appreciable way. As long as your mounting is done well, the length of the chain should not change during shock loads.

If anyone cares to contradict the above, please do. Right now these are all anecdotal, and I would love something more concrete.

Re: #25 Chain
 

1073 ran both belts and chains in our drivetrain this year. We had #25 chain going from the gearbox to the center wheel (it was placed above it due to space issues) and belts running from the middle wheel to the outer wheels. We didn't have any problems with the belts stretching, but we did implement some static tensioners on the #25 chain due to there being quite a bit of slack. No problems with stretching or snapped chains/belts at all

Re: #25 Chain
 

I'll put in a good word for timing belts. On 1768, we used 23 belts on our 2016 robot with great success. We used exact center-to-center distance runs with no tensioners and had zero failures all season long (5 events, 86 official matches). We used a mixture of 6mm, 9mm, and 15mm wide belts, depending on the strength required in the application. We used 15mm belts on our 8WD (picture) with 8" wheels and 60T versapulleys and never skipped a single tooth.

While chain certainly has its place (arms are a great example) and its advantages, properly used timing belts are lighter, quieter, safer, more efficient, require no lubrication, and are completely maintenance-free. Any ratcheting or failure is indicative of a design flaw, not of some unavoidable problem with belts. Whether used at exact distances like we did this year, or with tensioners as we have in the past, timing belts are fantastic for a variety of FRC applications.

Re: #25 Chain
 

What tolerances are you running for bearing placement on your drivetrain?

Re: #25 Chain
 

The OMAX waterjet our sponsor uses holds tolerances around +/-0.003". I'd say somewhere tighter than +/-0.010" is necessary to give exact distance belts proper tension. I've heard that some teams add a few thousandths to their exact distances to compensate for their manufacturing tolerances, though I've never done this.

Re: #25 Chain
 

Agreed.

We used a #25 chain, six 8" pneumatic wheel, 22 tooth Vex sprocket drivetrain this year and had no major problems. We worked hard to make sure that all of the sprockets were aligned. We also used the Dark Soul chainbreak tool to assemble the chain in order to avoid using links and it worked well. There was some slack that showed up and one chain slipped off so we had to improvise a chain tensioner using large zip ties and encountered no further problems. We were quite pleased with the results.

Now I must also add that our robot was well under the weight limit (90 lbs) so I think that helped.

In the future, depending on the nature of the game and if we are close to the weight limit, I think I'd go with #35 just for peace of mind.

Dan

Re: #25 Chain
 

I don't have the high speed camera to investigate this but I'd be curious if this was the belt stretching or a combination of, pulley shafts deflecting under high load, the teeth on the opposite pulley compressing, and the teeth on the skipping pulley compressing and bending. The way the belt is constructed I'd be much less surprised if it was the neoprene teeth deforming rather than the fiberglass tensile member stretching. I don't think they stretch much (more then negligibly) before failing but I don't know the exact numbers (elongation at break) on that.

Re: #25 Chain
 

In your belt runs, do you add anything to the nominal C-C distance? We had designed in about a 0.002 in our shooter belt (3mm GT3 belt), but since it was waterjet out, I can't be certain it held to .002 anyway. Regardless, it worked out fine for us. I've just noticed the west coast products calculator adds a bit.

Re: #25 Chain
 

The fixed c-c runs in our drivetrain were long by quite a bit, I think we've had a few runs were we've added as much as 50 thou. That was early on though and on runs of about 10 inches, when we first tried belts we tested a few different things and went with what we thought was good. Recently we've been trying to go closer to no addition, I don't know what the current situation is on that.

The way our drive has been set up in the past with 22T pulleys I don't think it's actually possible for the belts to skip without breaking teeth.

As far as gearbox belts, we've always built in some way to tension them, the current multiple mounting hole pattern has each pair vary by 5 thou I believe.

Re: #25 Chain
 

• ANSI 25: 1/4” pitch, 115 lb working load (McMaster-Carr)
• Chain stretches/wears over time
• If lengthening is significant chain will skip teeth
• If this happens, chain will need to be tensioned
• Higher tension causes chain lengthening to occur faster
• If tensioning is not used, it is important to space wheels such that a
whole number of chain links are need to span the distance
The amount of tension in the chains is largely effected by
sprocket size
• Assuming torque is fixed, a larger sprocket is equivalent to a
longer moment arm, and therefore less force (tension)
• Max torque is dependent on the amount of friction, which is a fixed
value
• To minimize tension, choose the largest sprocket that provides
enough ground clearance between the playing surface and chain

Bennett, Ben. "Drivetrain Design." Drivetrain Design (n.d.): n. pag. www.simbotics.org. 22 Oct. 2011. Web. <www.simbotics.org/files/pdf/drivetraindesign.pdf>.

Simbotics did an excellent analysis of many common drive trains used in first and the though process behind a reliable design, I encourage everyone to give it a look

Re: #25 Chain
 

Does anyone have any good info on how much #25 chain stretches over time? As chain stretch is actually wear of the roller axes, it will presumably plateau after a break-in period. That is, once it has stretched, what is the effective pitch length for the free runs between the sprockets? 0.2501"? 0.251"? 0.26"? 0.27?"

Has anyone ever tried "pre-stretching" chain as an alternative to dynamic tensioning? If so, how much tension and time is needed to get on that plateau?

Re: #25 Chain
 

We've used previously used chain before with no problem. The working life of chain is much longer than a single FRC season. However, after a season, the chain is basically as long as it's going to get.

Re: #25 Chain
 

The lengthening due to wear of a chain is calculated by the following formula:
% = ((M- (S*P)) / (S*P) * 100

M = the length of a number of links measured
S = the number of links measured
P = Pitch

Green, Robert E. et al. (eds) (1996), Machinery's Handbook (25 ed.), New York, NY, USA: Industrial Press, ISBN 978-0-8311-2575-2.

Re: #25 Chain
 

That seems like some extremely interesting information that might not exist at this point. I wonder if someone with access to a good tensile testing machine (like a University) could do some tensile tests on lengths of chain. That would yield both theoretical data if you pull it till the chain snaps, and practical data if you used a max load that might be something you'd see in FRC applications.

Basically, I would love to spend a day with #25 and #35 and a big hydraulic tensile testing machine.

Re: #25 Chain
 

Chain doesn't stop stretching (roller axes don't stop wearing), but it does plateau after a little bit. Teams have tried pre-stretching chain before, but you want to actually run the chain while doing this, not just pulling on it or whatever, since chain stretch is really the pins wearing down. Not as simple as linear force + time. There's also some sprocket wear that can be a factor in chain tension.

Re: #25 Chain
 

We abused #25 chains and sprockets this season. During design and assembly, all looked good and worked smoothly.
The 8 wheel cantilever axle design was extremely hard on the bearings and the 2x1x1/8 aluminum chassis tubes. After a day of qualifying at Alamo, the axles had excessive axial and lateral clearance. The bearings were no longer tightly fitting in the chassis tubes, causing a wee bit of chain/sprocket alignment error. We ended up letting our alliance down in the semifinals as we had destroyed chains and sprockets. (sorry about that)
The first step towards a fix was to replace the C-clip axle retention design with an end bolt design. This allowed for a preload on the bearings, keeping the axles parallel to each other and the sprockets aligned nicely. We also started replacing all of the chains and sprockets after qualifying. Future drive systems will not need this kind of maintenance.