CNC motor Coupler Death and Repair

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Toymaker

Well-Known Member
HMEM Supporting Member
Joined
Jan 26, 2022
Messages
747
Reaction score
402
Location
Thailand
Thought I would share this little repair in case anyone else runs into a similar problem.

The Plum shaft coupler on the Z axis of my DIY CNC lathe quite literally went to pieces last week; the red plastic insert between the coupler's two halves (shown in left photo below), broke up into lots of small pieces seen in the photo on the right below.
1669002788127.png
1669003383823.png


I immediately ordered a replacement (eBay) but decided to try to repair the broken one so I could continue to use my lathe while I waited the expected 3 weeks shipping time for the new plum coupler to arrive.

I wrapped masking tape most of the way around the two halves, leaving only the top section open. I then filled the void between the two metal halves with "steel epoxy" (similar to JB Weld) and let it harden.

I've been using the lathe for the past week and so far the epoxy is holding. I suspect that since the glue isn't actually holding onto the metal halves, but merely acting as a bushing between the halves, it should last a very long time.

1669004848801.png
 
Clicking noises and backlash...sure enough a look under two of the covers revealed that I have the same issue. :cool:
I guess all the rubbers need to be replaced. Wonder how the stuff can "disappear", are there bugs that eat rubber? (you found at least some remains, I feel there are not a lot of crumbles) Can rubbers be printed with those rubbery high end filaments on the 3d printer? (just a stupid brain glitch, I wount try by myself)
Need to find the rubbers or maybe complete couplers. The complete machine needs to come apart....motors for z and y are inside the castings.
I am confident, that the alignment between motor and ball screw of my machine is so "good", that something will go horribly wrong with a too rigid coupler....

Hope you get your spares quickly.

Greetings Timo
 
Clicking noises and backlash...sure enough a look under two of the covers revealed that I have the same issue. :cool:
I guess all the rubbers need to be replaced. Wonder how the stuff can "disappear", are there bugs that eat rubber? (you found at least some remains, I feel there are not a lot of crumbles) Can rubbers be printed with those rubbery high end filaments on the 3d printer? (just a stupid brain glitch, I wount try by myself)
Need to find the rubbers or maybe complete couplers. The complete machine needs to come apart....motors for z and y are inside the castings.
I am confident, that the alignment between motor and ball screw of my machine is so "good", that something will go horribly wrong with a too rigid coupler....

Hope you get your spares quickly.

Greetings Timo

If I use a pliers to squeeze one of those little red pieces, which are the remains of the "bushing", there is noticeable flexing,..not too much,...just a little. So I used a 5 minute epoxy,....which when it hardens, also has a little flex. At the moment, I'm inclined to leave my epoxy repaired coupler in place until I see signs of trouble. Fortunately for me, the repaired coupler is easy to see and access,...but replacing it will mean disconnecting at least one end of the Z ball-screw.

Good luck with your couplers.

BTW, maybe I found pieces because those rubber-eating bugs aren't indigenous to Thailand :cool:
 
I have made small couplings of that type. I made a steel mould and cast the 'spider' in urethane. I used a 70 shore rubber compound, but a harder compound would have been better. If too soft, the squeezed rubber creates an axial force.
 
I have made small couplings of that type. I made a steel mould and cast the 'spider' in urethane. I used a 70 shore rubber compound, but a harder compound would have been better. If too soft, the squeezed rubber creates an axial force.

I can only guess at what shore hardness my epoxy spider might be. Pushing a center punch into the epoxy using only my hands, no hammer, the hardness seems to be about the same as Lexan. The epoxy is substantially harder than the pieces of original spider material,....BUT the original spider material seems to be falling apart from either age &/or too much exposure to WD-40 which I use as a cutting fluid, and which frequently gets splashed onto the coupler.
 
what you want is Lovejoy couplings, they are much better built, and at least here in the USA not expensive at all. lovejoy couplings
Very "disturbing" name I guess the poor sales guy will have to listen to a lot of weird comments (like the one you are reading now :).
Can you just buy rubbers from them?

Greetings TImo
 
what you want is Lovejoy couplings, they are much better built, and at least here in the USA not expensive at all. lovejoy couplings

Thanks for recommending a dependable brand. The coupler that died lasted for 7 years of hobby level use. I paid about $14 for it, so not a bad lifespan considering all the WD-40 that was splashed onto it over those 7 years. At the moment, I'm inclined to leave my epoxy repaired coupler in place until I see signs of trouble. Fortunately for me, the repaired coupler is easy to see and watch for signs of breakage.
 
The black spiders are BunaN in Lovejoy couplings, with a Shor hardness of around 80A. The Urethane spiders have a Shor hardness of 55N.

Here's a link to some TPU filaments rated in order by hardness. If you are after the 55D+ hardness of urethane spiders Crystal Flex is a 65D, but a 1/2 kg spool is $66 USD plus another $12 in shipping, could be cheaper to just buy replacement spiders.

https://x3d.com.au/blogs/tips-and-t...est-tpu-filaments-according-to-shore-hardness
Most of the usual TPUs appear adequate for replacing BunaN material for hardness, provided your application doesn't mind a bit harder spider. As most CNC machines have rigid stepper mounts and a bearing on the driven end of the lead screw this is unlikely to be an issue. This tends to also be in favor of leaving a cast in place epoxy repair alone as it works, unless wanting to play with TPU. The spider in most stepper applications is just slightly smoothing out the stepper pulses rather than allowing quite slight (0.5 degree max) misalignment of shafts if that.

As TPU is a flexible polyurethane it is reasonably resistant to most oils and the like, although without knowing the exact make up of the filament this would be a generalization too broad for any life or safety critical application without digging deeper. I sure would be willing to try printing a few replacements if any of my spiders ever break down. Replacement spider prices and / or shipping have gotten just silly lately, more than I paid for entire couplings just several years ago.

Cheers,
Stan
 
These JAW type couplings will destroy the elastic spiders if they are out of alignment. They can't take more then a few thou of axial or radial alignment errors. If you can figure out how much misalignment you have between the motor shaft and load shaft, and correct this. You may have to shim the motor end assuming it has slotted NEMA mounting base, so you can align it in both axis for radial error. Axial make sure they are in line with each other.
The differential movement is what will eat the elastomeric material.
 
couplers like that are designed to accommodate any misalignment between the shafts, perhaps the reason the rubber deteriorated is that it was being continuously flexed by misalignment, perhaps the epoxy replacement will shift the deterioration to a more expensive to repair part (either the lead screw and nut or the stepper motor itself) (and, no, epoxy does not flex like rubber). perhaps you should have filled the coupler with silicone rubber or shoe-goo or ...
 
From my Automotive experience of elastomers, they are all good for something, but few are good for more than 1 or 2 characteristics. As they are organic, all elastomers suffer from rapid ageing and deterioration with some form of organic solvent. In cars you have 4 or more main groups of solvent where it is critical to select the right elastomer. Engine oil, ATF, Brake fluid, long-life coolant (glycol) battery acid, windscreen washer (ethyl alcohol), petrol, etc. All degrade some elastomers. I guess your CNC is not built for WD40 pollution of the joint rubber.
But.... and this is the key, ALL elastomers will age and perish naturally. Car tyres and vibration mount rubbers do Not meet original specification tests after about 5 years, even though we use them at 15 years old! Brake hoses really should be changed before 15 years old. Engine oil seals may last 20 years, but may deteriorate in 10 years.... On a 12 year old SAAB I changed the broken rubber star on the transmission shifter linkage into 4 springs made from coat hanger wire, which lasted 3 years before I scrapped the car. (It worked like new, but was going to rust in a few years).
Rubber turning crumbly into small piece is common ageing, from fatigue stress cracking throughout the material, so may not be from the WD40, or may be the WD40 accelerated the ageing. I would expect WD40 to make a soft gooey or sticky surface if it was attacking the rubber.
The Epoxy resin repair may last some time depending on usage - fatigue from the number of cycles of bending within the shape of the piece. Polyurethane as a basic molecular material is very fatigue resistant, unless over-stressed. Normally used for long-life flexing components. (The best material for the foam seat cushions to last up to 12 years... Bouncing back-sides give million's of flexing stress cycles, but eventually it just turns to dust with age. Neochloroprene lasts a longer time, but not with cyclic stress.).
Hope there are no inaccuracies in this and it helps you understand the fixes for your aged and broken coupling? Change them all when you service the machine, or have years of replacing individual parts as they progressively fail.
K2
 
Very "disturbing" name I guess the poor sales guy will have to listen to a lot of weird comments (like the one you are reading now :).
Can you just buy rubbers from them?

Greetings TImo
I am reasonably certain you can by the flexible inserts, though I douibt that they are made of rubber (or latex), there are a lot of the couplers on ebay, I even have an odd one listed there that is MT2 on one end and 1/2 inch shaft on the other end... no idea what that was originally for
 
couplers like that are designed to accommodate any misalignment between the shafts, perhaps the reason the rubber deteriorated is that it was being continuously flexed by misalignment, perhaps the epoxy replacement will shift the deterioration to a more expensive to repair part (either the lead screw and nut or the stepper motor itself) (and, no, epoxy does not flex like rubber). perhaps you should have filled the coupler with silicone rubber or shoe-goo or ...

Epoxy flexes more like plastic than rubber,...but it does flex. You can easily see how flexible your favorite epoxy is by mixing a small amount on a plastic bag, or some other thin plastic container, or packaging, like a soda bottle. Spread the epoxy onto the plastic in a thin layer no more than a few millimeters thick. Once the epoxy is cured it's easy to demonstrate it's flex limits before it breaks.
 
From my Automotive experience of elastomers, they are all good for something, but few are good for more than 1 or 2 characteristics. As they are organic, all elastomers suffer from rapid ageing and deterioration with some form of organic solvent. In cars you have 4 or more main groups of solvent where it is critical to select the right elastomer. Engine oil, ATF, Brake fluid, long-life coolant (glycol) battery acid, windscreen washer (ethyl alcohol), petrol, etc. All degrade some elastomers. I guess your CNC is not built for WD40 pollution of the joint rubber.
But.... and this is the key, ALL elastomers will age and perish naturally. Car tyres and vibration mount rubbers do Not meet original specification tests after about 5 years, even though we use them at 15 years old! Brake hoses really should be changed before 15 years old. Engine oil seals may last 20 years, but may deteriorate in 10 years.... On a 12 year old SAAB I changed the broken rubber star on the transmission shifter linkage into 4 springs made from coat hanger wire, which lasted 3 years before I scrapped the car. (It worked like new, but was going to rust in a few years).
Rubber turning crumbly into small piece is common ageing, from fatigue stress cracking throughout the material, so may not be from the WD40, or may be the WD40 accelerated the ageing. I would expect WD40 to make a soft gooey or sticky surface if it was attacking the rubber.
The Epoxy resin repair may last some time depending on usage - fatigue from the number of cycles of bending within the shape of the piece. Polyurethane as a basic molecular material is very fatigue resistant, unless over-stressed. Normally used for long-life flexing components. (The best material for the foam seat cushions to last up to 12 years... Bouncing back-sides give million's of flexing stress cycles, but eventually it just turns to dust with age. Neochloroprene lasts a longer time, but not with cyclic stress.).
Hope there are no inaccuracies in this and it helps you understand the fixes for your aged and broken coupling? Change them all when you service the machine, or have years of replacing individual parts as they progressively fail.
K2

I'm truly at a loss to definitively understand why the original coupler failed as it did. At first blush, the out-of-alignment explanation makes the most sense, but when I spin the now repaired Z axis coupler with my fingers, I cannot detect a stiff spot in the rotation, which is what I would expect to find if the stepper shaft was out of alignment with the ball screw shaft. As mentioned previously, the failed coupler has been on my lathe for 7 years, but of course I have no idea when the rubber spider part was made nor how long it sat on a shelf in a warehouse waiting to be sold. And there's always the possibility the failed spider was made from a bad batch of rubber compound. So the reason for failure remains a mystery.

I've made a note my build-log-book as to when I repaired the Z axis coupler, so if it fails again, I'll know exactly how long my epoxy repair lasted.

1669179552223.png
 
Last edited:
From the evidence presented, I propose fatigue from repeated compression and relaxation as the coupling has been rotated. You may expect a million or so cycles to do this, but a greasiness fluid such as WD40 can reduce the fatigue life by tens or hundreds, so a few thousand or tens of thousand rotations may cause the total failure you experienced. How may hours running and at what rpm to generate the first failure?
K2
 
To temporary repair those, I assemble in place as it would be normally make a little tape mold around it to keep the distances and inject some of those high strength non silicone adhesive/sealant through a small hole on the tape with a vent on the other side. They stay flexible.
 
To temporary repair those, I assemble in place as it would be normally make a little tape mold around it to keep the distances and inject some of those high strength non silicone adhesive/sealant through a small hole on the tape with a vent on the other side. They stay flexible.

Can you provide more information on what adhesive/sealant you're referring to,...an eBay or Amazon, or some other online link would be most useful.
 
Here in Portugal they call it "Seal and Glue". It's a very high strength acrylic sealing/gluing substitute for silicone and can be painted. It's used in the construction business. The Amazon example is just one of the brands.
https://amzn.eu/d/fQpSj7g
 
Back
Top