Dangerous! Snapped chain!

Hi all, first run on the maslow cutting the sled template and my chain right snapped. Broke at the link inside the carriage with the cotter pin. The left chain held but this created a swinging router pendulum of death, cutting through my workpiece, taking out chunks of the frame, a little wall, luckily no humans or power cords were in the way. Really dangerous situation as I didnt have a good way to cut the power- cpu wall outlet was behind said pendulum of death. Router had been hanging and calibrated almost a week before so perhaps there was something wth the router bit catching a little on the cut (was watching the whole time and didnt see any issues like that, running ar 12krpm through 3/4" ply) 1) why is the chain this weak? Anyone else havd this happen? Is there something I did wrong attaching to the sled? Did the carriage/cotter pin attachment aid in the breakage? Where can I order new chain?

3 Likes

Yikes! Seems like I read about this happening once before. The chain is pretty stout, and should be able to hold way more than that little sled. Wouldn’t think you would need to replace the chain, just remove a link or two from the end, and do a load test of some kind. You might need to get a chain breaking tool, or go to a bike shop. The tool I bought will handle several sizes of chain (bike chains are a bit larger, I think).

I guess we all should be wary of this possibility.

Good luck!

ebay is a good source for chain fairly cheap

per http://www.ocm.co.jp/en/pro/roller/03_01_02.pdf #25 chain should have a
working strength of 154 pounds and a yield strength of 1058 pounds, so it should
not have broken, can you get us a picture of the broken link?

David Lang

2 Likes

This is what I experienced. This is caused by connecting to the chain link that has no pin in it.

4 Likes

ahh.

Yes, that is a way that the chain was not designed to be pulled, so without a pin in it, that’s a weakness.

the top mount kit doesn’t have this weakness as the pin goes through the normal chain pin holes.Both the wood triangulation kit and the ring kit have this problem.

the only way to avoid it with these kits is to put the chain in a bit more so the pin goes through a solid link. This will add one link to the rotation radius of the triangulation kit.

2 Likes

So @dlang, perhaps there should be a general notice to everyone with a ring to use the next link. This is not a weakness, it is an error in the assembly guide, and a lack of communication about a serious safety concern. I personally do not know that much about chains, and assumed step 6 in the assembly guide was correct…

I am willing to bet that most everyone is following this instruction to the letter.

With a caliibrated system, it would be simple to pin the next link on each side, and move the chain on each sprocket accordingly, right? And with an uncalibrated system, I believe the claibration routine will deal with the change in the rotation radius when running the Triangular test cuts.

Thanks!

2 Likes

It’s not that you misunderstood the instructions, it’s that this hasn’t come to
our attention before, so nobody though about this weakness.

I never thought of this, but as soon as I saw the picture, it was obvious to me
what went wrong

I agree, this is worth a safety bulletin being sent out by @bar to everyone who
purchased a kit.

David Lang

4 Likes

Save Fingers, Save Lives With A No-Voltage Release For The Shop | Hackaday should be build in as well :blush:

2 Likes

Picture is not clear wha to do to me. Another angle please.

Bummer that this happened and I’m glad no one was injured.

The most important lesson here is to have a safe and quick way to cut power to the router at all times!

Personally I have a power strip with cutoff switch wired only to the router at my workstation/laptop desk. I recommend something similar.

Additionally, considering the recent report of someone running over their Z-axis wires while cutting, please be sure that all wiring is secured in a way that prevents any similar thing happening and above all:

" ALWAYS MONITOR A RUNNING MACHINE, NEVER LEAVE A MACHINE UNATTENDED. "

good luck!

1 Like

I think doing what you say (move a link on the sprocket) would work, but it might be slightly off depending on what the end of your chain looks like. In the picture you posted, the end of the chain is a bushing so moving it won’t make a difference. If the end of your chain is a pin, then moving it to the next link in will put the cotter pin up against a bushing. Therefore, the rotational radius would be changed by the thickness of the bushing. Nevertheless, I suspect it’s too small to make a difference.

Right now the cotter pin passes through the last link of the chain. The cotter pin presses against the bushing but because it’s the last link, there’s no chain pin inserted through it. The hypothesis is that because there’s no pin, the bushing fails. So the fix is to feed another full link through the opening and place the cotter pin through the gap of the second-to-last chain link. That way the cotter pin is pushing against a bushing with a chain pin inserted through it.

My first thought for a simple temporary fix until it is addressed by @bar will be to slide a small finishing nail through that last link bushing hole and then bend the ends over and clip off most of the excess. This should help strengthen the bushing in the last link without needing to change the calibration or reconfigure your chain length.
Just my 2 cents
Thanks
Jim

4 Likes

That is frightful! Should we change the recommendation to be to put the pin through the second link?

If you send an email to info@maslowcnc.com we will send you a new one right away, but I agree that until then you should be OK to remove that last link

2 Likes

This would be a minor difference in the rotation radius, a couple mm will make a
difference.

But the chain should always end in a bushing. To end in a pin, you either have a
detachable master link installed, or after they push the pin out to break the
chain, they would have to go to the effort of putting the pin back in.

David Lang

YES!!! (20 characters :slight_smile:

That’s what I was wondering about.

Yes, this is the same thing that happened to me. Thanks everyone for the fast input.

The router kill switch is definitely going to happen, going to Rockler today. I was going to put this on the side of the Maslow but now I know that is not a safe enough place to put it. It must be in front of the cutting plane, ideally right next to the operator’s spot. A related question though-.if the router was cutting while moving and I kill the router power then the bit gets caught while the sled can keep moving. Seems like it could ruin other things (calibration and/or the motors). Has anyone hacked Ground Control/Arduino to add a physical stop switch also so the motors stop also?

On the chain strength- I’ll move to the second link but have three questions: a) When re-calibrating chain lengths, should we add one chain link (use second link) each time since Maslow is assuming we are using the first link? b) what is the rated strength of the chain?, and c) is the last link actually weaker because the pin is not in the link?

Trying to answer my own questions:
a) Yes? Right?
b) Some googling shows tensile strength of #25 roller chain should be around the 800-1000 lb (lbf?) mark (www.usarollerchain.com). But sites like McMaster show standard #25 chain having a “working load” of 85 lb. (https://www.mcmaster.com/6261k171). Obviously these are different ratings, but I’m not sure how each is measured. Assuming the router and sled weighs ~20lb (90N) and doing some trigonometry, the tension on the chain should really be less than 3x this (about 60 lb). A factor of safety of 5 would be appropriate as there may be additional forces (Bit grabbing, unequal chain tensions, etc) at play. So if the strength of the chain is indeed ~1000 lb - it should be fine. But if the failure is 85lb “working load”, then it is totally a concern.

c) I haven’t quite wrapped my head around the force diagram of the chain/pin/roller/bushing yet. But I think it’s related to the question above in that there could easily be different failure ratings for tension on the chain (the pins take the load) vs the force going over a sprocket (the rollers take the load). The cotter pin through the roller in the carriage seems like the same loading case as the roller over the sprocket - crushing the rollers. And I would assume both locations should be nearly equal given the tension on the chain is everywhere. The pin in the second link might not be add strength since there’s a tiny slop in the roller/pin/bushing mechanism. However it might be an excellent “backup” mechanism if the roller were to fail. Or the roller might be bending a little bit, removing the slop and indeed stiffening the joint with the pin.

As a second failure mechanism, from the attached picture you can see that the linkage side plates splayed (looks a little like a “Y”). It could be that happened as the roller tore out, but it could also be that the side plates splayed first due to torsional forces and the assembly slid apart. This probably would be helped by the roller side pins in the second link, but not significantly as those are not supposed intended to take shearing force, either.

I don’t know the answer, but all I’m saying is perhaps this area needs more thought before assuming moving the link up a spot is the solution.

1 Like

When calibrating, you don’t need to do anything to compensate for using the second link of the chain for the cotter pin. The routine will modify the rotational radius to account for it. I think it will shorten it by 6.35 mm.

The router kill switch is definitely going to happen, going to Rockler today. I was going to put this on the side of the Maslow but now I know that is not a safe enough place to put it. It must be in front of the cutting plane, ideally right next to the operator’s spot. A related question though-.if the router was cutting while moving and I kill the router power then the bit gets caught while the sled can keep moving. Seems like it could ruin other things (calibration and/or the motors). Has anyone hacked Ground Control/Arduino to add a physical stop switch also so the motors stop also?

On the chain strength- I’ll move to the second link but have three questions:
a) When re-calibrating chain lengths, should we add one chain link (use second
link) each time since Maslow is assuming we are using the first link? b) what
is the rated strength of the chain?, and c) is the last link actually weaker
because the pin is not in the link?

Trying to answer my own questions:
a) Yes? Right?

when calibrating, the system figures out the rotation radius, and that is just a
bit different if you hook the sled one link further up (i.e. you don’t need to
worry about it)

b) Some googling shows tensile strength of #25 roller chain should be around the 800-1000 lb (lbf?) mark (www.usarollerchain.com). But sites like McMaster show standard #25 chain having a “working load” of 85 lb. (https://www.mcmaster.com/6261k171). Obviously these are different ratings, but I’m not sure how each is measured. Assuming the router and sled weighs ~20lb (90N) and doing some trigonometry, the tension on the chain should really be less than 3x this (about 60 lb). A factor of safety of 5 would be appropriate as there may be additional forces (Bit grabbing, unequal chain tensions, etc) at play. So if the strength of the chain is indeed ~1000 lb - it should be fine. But if the failure is 85lb “working load”, then it is totally a concern.

the reference I find says a working load of 150 pounds and a yield of 1500
pounds

c) I haven’t quite wrapped my head around the force diagram of the
chain/pin/roller/bushing yet. But I think it’s related to the question above
in that there could easily be different failure ratings for tension on the
chain (the pins take the load) vs the force going over a sprocket (the rollers
take the load). The cotter pin through the roller in the carriage seems like
the same loading case as the roller over the sprocket - crushing the rollers.
And I would assume both locations should be nearly equal given the tension on
the chain is everywhere. The pin in the second link might not be add strength
since there’s a tiny slop in the roller/pin/bushing mechanism. However it
might be an excellent “backup” mechanism if the roller were to fail. Or the
roller might be bending a little bit, removing the slop and indeed stiffening
the joint with the pin.

the chain is not designed to have force on a roller without a pin in the chain.
you would always have more than one link around a sprocket, so this isn’t
something that’s rated

As a second failure mechanism, from the attached picture you can see that the
linkage side plates splayed (looks a little like a “Y”). It could be that
happened as the roller tore out, but it could also be that the side plates
splayed first due to torsional forces and the assembly slid apart. This
probably would be helped by the roller side pins in the second link, but not
significantly as those are not supposed intended to take shearing force,
either.

the exact order of failure doesn’t matter. what matters is that the chain is not
designed to take force in this configuration.

David Lang