What's new
What's new

3D Printing Aluminium Gearbox Housings

Xylord

Plastic
Joined
Nov 18, 2016
Hey all!

I work for a lab where we need to get specialized rugged power tools made. I was wondering if anyone had tried making a gearbox housing which would traditionally be cast using 3D printed aluminium. Feels like that's becoming pretty affordable these days, especially for small runs (1-5 parts).
Would love input from people who've worked with 3D printed aluminium in general as well.
 
To forwarn, I'm not quite what you asked for; as I've only worked with things 3d printed from titanium and inconel and not a gearbox case specifically.

A gearbox housing should be fairly trivial to design to be printed. Just make sure you follow the design rules for whatever printing technology you are using. There are specifications for max overhang angle, max unsupported span, and other such things. Most print houses will have specifications on how to design the part for printing.

If you are prototyping, 3d printing would be a good way to go; as you could test your design before you get it put out for production. Or if you just wanted a one-off it is also a reasonable choice. Depending on your requirements, post process strengthening is sometimes done. Like hot isostatic pressing, this usually yields a bit more strength, and eliminates any possible voids if those are an issue.

In general with DMLS printers, Z height adds some cost. As adding more powder is the slowest part of the process. Taking up more area of the print bed can also increase price as they can't put as many other people's parts on the run. So the prices can get high even if you arn't actually using that much metal volume.

If you are making one-offs, you can also just do titanium if that would be an advantage for you. If I remember right, aluminum might be a little bit more expensive than titanium, or at least close in cost to print. As it rates on the more difficult side for them to print with.

I know a few people who have gotten an aluminum print done, and they seemed happy with them. The AlSi10Mg alloy has a pretty low elongation before break for aluminum just as an fyi.

Just for another alternative for your radar, you can also get casting molds directly printed in sand.
 
Em ..
if the gearbox is 1-2-5 hp+ in power, and has no other exotic design constraints like extremely low weight no matter the cost,
it is likely a fabbed steel or ci housing will be 10x cheaper.

My "feel" is that 3d printed metal precision housing for a gearbox about the size of a shoebox == 20.000$.
A fabbed steel housing 2-4k$, depending.

Endless ways to make the fabbed parts cheaper, if more than 1 is desired.

Power/size/mass has little effect on fabbed cost, but 3D printing can be 2-10x the cost for just a bit more power or precision.
Power three in time for size in volume and time, and larger assys can only be done by very few makers, if precision and mechanical robustness is needed.

A smaller gearbox than 2hp might be quite a bit cheaper, and as a proof for later metal casting the 3d route might be right, perhaps.

If the gearbox goes into exotic use, plane, space, balloon the extra cost might be worth it.

Nevertheless I feel that any gearbox housing 3d printed will need secondary machining and probably/perhaps adjustable bearing mounts of some kind.

The OP wanted metal-printed housings indicating direct use of the housings as-printed, with not specified post-op work, accuracy, or rigidity, or use.
Likewise volumetric size, power, work envelope, qty was not indicated.

For example;
A BMW 330i gearbox of == 200 kg / 250 Hp peak, volume half the size of a small dorm refrigerator.
Housing == 50-80 kg in alu.
This would likely need thick walls of honeycomb, maybe 2" thick.
Maybe 30 kg printed.
Likely cost 500k $.
 
I am mainly working on small handheld tooling, so we are talking closer to 1/2 hp than 1 hp in most cases. Your intuition more or less aligns with mine, we'd go fabbed for larger gearboxes, for sure. I also think post-printing machining will be necessary for locating features, like bearing locations, mounting, o-ring/gasket grooves where necessary. We'll have to design the parts such that getting a datum to locate other features from is not too difficult.
The sizes we're dealing with are closer to a closed fist than a closed fridge, so hopefully we're not getting a 500k bill :D

Thank you for the insight on the manufacturing costs, and for the info on post-processing. I wasn't aware of the possibility of isostatic pressing. In your experience, do metal DLMS parts have issues with sealing or leakage due to their porosity? Also, from the quotes I've gotten so far, aluminium seems to be a fair bit cheaper than titanium, but the design wasn't modified to take into account the better properties of Ti, so the amount of material could be reduced, bringing the prices closer.

If you know of any shops in NA that do small production runs and that you'd recommend, I'd love to know about them, since you have a hand in the industry!

Getting a sand mold and casting would be a good solution as well, we could probably whip up a little Al foundry easily enough, we'll have to consider it.
 
I just keep thinking of how many I could make from solid in the time it takes to 3D print one

3d printing is cool tech for things that are not easily machined, or where low volumes and purchasing cycle speed make it faster[IOW engineer can run one overnight on his desk rather than wait for purchasing to get 3 quotes]

I mean, I can turn 50 bucks worth of aluminum to chip in about 3 minutes.............

In the 2 1/2 axis world where I imagine most gearbox designs live, I cannot even imagine casting for low unit volumes and or high value items
 
If you have to do any post machining it may well be cheaper to just machine it from solid. Trying to hold and locate a printed part would be a PITA.
 
Thank you for the insight on the manufacturing costs, and for the info on post-processing. I wasn't aware of the possibility of isostatic pressing. In your experience, do metal DLMS parts have issues with sealing or leakage due to their porosity? Also, from the quotes I've gotten so far, aluminium seems to be a fair bit cheaper than titanium, but the design wasn't modified to take into account the better properties of Ti, so the amount of material could be reduced, bringing the prices closer.

If you know of any shops in NA that do small production runs and that you'd recommend, I'd love to know about them, since you have a hand in the industry!

Getting a sand mold and casting would be a good solution as well, we could probably whip up a little Al foundry easily enough, we'll have to consider it.

You would probably want to post machine, or at least polish up sealing faces. The surface finish on each face can vary quite a bit depending on the orientation of the part as it is printed.

There are not really much for porosity problems, or sealing in the part itself. It is nearly solid. The iso-static pressing just evens everything out in the part, fixing minor imperfections, as well as evening out any accumulated stress. It really isn't necessary in anything but the most high performance of applications. If you do end up running into issues, it is an option on the table.

It has been a year or so since I last got a quote, and for that part at least; the aluminum was about the same price in the scheme of things.

All of the DMLS shops do small production runs, since the technology doesn't really lend itself to mass production. So you wont run into minimum order requirements very often. I can't really say much on experiences with different companies as I've only worked with a handful of parts, most of which I didn't do the buying. Stratasys Direct, and protolabs are the ones I've gotten quotes from.

Where 3d printing really shines is complicated internal features, like complex oil delivery passages. The application we used it for was printing liquid rocket combustion chambers. As you could print the hundreds of cooling channels directly into the engine and save many hours of machining time with a less complicated end product. So if your gear box is just a simple case. Conventional machining is likely going to be a faster and cheaper route. Since you have to machine bearing seats and other stuff like that anyway. So after the setup time, the time to machine the rest of the gearbox isn't that many more hours.

To prototype a casting, the 3d printed sand casts would likely give you a more relevant result to the final production method. As there are plenty of things you can do 3d printing that wouldn't work so well casting it. The 3d printed casting would save quite a bit of time, as you don't have to invest in tooling up a production run for something that might be bad.
 
learned long time ago say you want to machine a small plastic part and its ends up costing $100. each
.
then you find a mold cost $10,000 to have made but if you mold 20,000 plastic parts cost ends up less than $1. each
.
sure you can make one part and pay 100x more than if part made by a more efficient process. engineering and design for manufacturing efficiently is a science. those who under stand it are the ones manufacturing in massive quantities
.
sort of like custom car made for $500,000 and a car made for $20,000 thats made by the 100,000's each every year
.
take a simple 3/4" wrench open and closed on each end. that 3/4" wrench of forged and heat treated steel and polished and chrome plated. if you costumed made the wrench can easily be over $1000., you can buy a wrench set of 10 wrenches for under $50. usually
 
The size OP suggests and the fact it's wanted in aluminum sure makes it sound like milled from solid to me. And there are folks elsewhere on this board offering time on fancy 5-axis DMUs for $60/hour - so if it's a 3-axis part as is likely it shouldn't be too hard to machine. (Unless it requires the kind of internal voids or passages that wrrock refers to.)

I'll note that some of the best uses of 3-d printing I've seen (very limited experience) are either patterns for castings, or making short runs of parts where you really wanted it in some kind of plastic - and really don't want to make anything like an injection mold.
 
Have you gotten separate quotes from service bureaus to quote 3D Printing and any required post processing and a 2nd quote for Machining the casings from billet. I suspect if there is not a design element that requires 3D printing, having the casings machined will be significantly less expensive.

As for setting up capability in house, the learning curve with printing metal can be 6 months long according to several sources both in service bureaus and manufacturers I've spoken with. Post processing can also be tedious and expensive. If you need post machining... You may find setting up a HAAS mini-mill, Tool Room Mill or even a VF instead of a printer at this time provides a very solid path to success. You may already have machining capability, but if not, I think it's a great first option.
 








 
Back
Top