I’m new to the Bambulab 3D printing community and have a few questions about my current 3D printing project.
A sprocket on an old back training machine in my home gym has broken.
The device is old and I can no longer get any spare parts for it. Now my idea was to replace the defective sprocket with a 3D printed sprocket.
The left-hand sprocket is the original sprocket that was fitted to the machine when it was delivered around 40 years ago. The right-hand metal sprocket has been reproduced over the years, but it never really ran smoothly. The right-hand sprocket was also once a plastic injection-molded part, just like the left-hand one. On the following photo you can see the original sprocket in dismantled condition.
I recreated this sprocket in CAD and then sliced it in Bamu Studio.
I sliced the sprocket with a filling level of 50% and 9 wall lines in the hope that this would make it stable enough. I then printed the sprocket at work in our X1C. I used Bambulab PLA CF. At home I only have a P1S, on which I have not yet installed the hardened nozzle and the extruder wheel. However, it is ready for installation.
Now that the sprockets have been in use for about a week, I have realized that my design is unfortunately not holding up. The tooth tips are breaking out of the teeth.
So I would like to ask the community what I can do to get the sprocket running. What settings can I change or what material can I use?
I wanted to avoid expensive materials like PPA-CF for the time being. Maybe someone has an idea how it could work with the cheap ones. After all, the original sprocket was only made of plastic.
Many thanks in advance for your help and suggestions.
Kaktuusch
PETG would be slightly tougher and less prone to breaking than PLA, but I suspect it would still have a short life.
The original sprocket was probably a type of nylon (a plastic). PPA-CF is probably overkill, but you can find PA6 for reasonable prices.
Hello, thank you for your quick reply.
I did a bit of research and found out that it was stupid to print the sprockets in PLA CF, as the CF content makes the PLA much more brittle.
My requirement for the sprocket is that it should last just as long as the original part (30 years xD). That’s why I would rule out PETG if you already describe it as short-lived.
I agree with you that PPA CF is too overkill, it’s also simply too expensive.
PA6 sounds like an option though. But which one? PA6 CF or PA6-GF?
If I understand correctly, Bambu does not have a normal PA6 in its range.
Can the CF and GF components in the PA6 filament also make the product more brittle and fragile, as is the case with PLA?
Those sprockets just let the chain roll over them or do they need those teeth to keep the chain in some kind of register? Or is there torque being transmitted to something we can’t see?
If there is no torque being transmitted and they just hold the chain, it may help to shorten or mostly remove the teeth so the chain rollers don’t hit them? Kind of like a bicycle derailleur?
The best material below €100/kg would probably be PA6-CF with post-printing annealing. Choose the right filament - 3D Filaments Guide for Bambu Lab Printers
However, wear will still be heavy and its not gonna last anywhere near as long as the original. Furthermore, it’ll release carbon fibres into the chain as it wears down, causing both erosion and corrosion of the chain link.
So you could try PA6-GF with annealing, which is much more difficult to print but solves the corrosion issue.
Ok, I know we all have a passion to create things and fix our own problems but I have one more suggestion to keep as a backup plan. If you’re very confident your STL is perfect PCBWay can print it in metal. If you upload it you can get an instant quote in several different metals. I would weigh the price of a roll of whichever filament you land on vs having them make one for you that might last way longer. Just a thought…
Well, others have already addressed the issues with filament so I’ll address the design aspects and what can work.
The best filament I found for structural parts that require both strength and rigidity is PC. It isn’t the easiest filament or the prettiest but at 100% wall loops, not to be confused by 100% infill, it produces parts that rival metal. Of course it’s not metal and will not work in your application but for prototyping before you send it off for metal print, it will give you a great idea of how the metal part will perform in terms of fitment and precision.
So if you change these settings below, I will get 100% layer coverage and the layer lines will be inline with the lines of stress. Try it for yourself and be the judge.
Try this out with PC and ABS. I found that ASA and PETG can often be too soft for gears. I have found that PC can be brittle at times but it is so rigid, I no longer use heated threaded brass inserts for my screws because I can directly cut tapped threads into the material itself and it is just as strong as brass inserts at no cost other than the filament. The trick is to use higher wall loops to ensure the holes are thick enough to cut a thread.
I have created another (beautiful xD) illustration of how this system works. You sit on the seat of the system and press against the roller with your back. This causes the eccentric disk (which looks like a horizontal D in the sketch) to move. This tensions the chain and the weights of the system move. The sprockets at the top are therefore actually only used for deflection. However, it seems to me that the left sprocket is under more load than the right one. This is because the angle at which the chain is deflected is steeper on the left-hand sprocket than on the right-hand one.
I suspect that the length of the teeth is important. If only for safety reasons, so that the chain cannot jump out so easily.
I am convinced of the quality of my STL. But why should I have this part printed from metal at PCB? If I’m already using a prototype manufacturer like PCB, then I can also have it turned or milled there in the conventional way. If I have the two sprockets 3D printed there, then 2 sprockets will cost me around 600€. If I have the two sprockets manufactured conventionally, it will cost around €100.
Just offering options, but yes, 600€ is a lot. The STLs I’ve sent them the quote was around $40. So getting them conventionally done for €100 doesn’t sound good?
Hello Olias,
Thank you very much for your great design tips and tricks. I will definitely take them into account for future projects and will read this article more often.
