3D printer uses

[elf]

The main drive timing belt pulley on my rose engine is 3d printed, close fit on an aluminum hub, and secured with machine screws. It's been working flawlessly for more than four years. Spur gears printed in ABS or Nylon don't usually need a metal hub, but there's no reason it can't be done.

[RMinMN]

Maybe an amateur now, but you'll get better with practice.

Is it carved in stone that there must be a press fit holding two pieces together? I wonder what would happen if there were some holes, or other features at the OD of a rim such that the plastic could be printed directly on to the rim, with those features incorporated as some sort of anchoring points. Tossing that out as an idea.

I think this is an example of mission creep. I think it would take some tweaking of the gcode sent to the printer to make sure the head didn't want to move where the hub was mounted. Then locating the hub in exactly the right spot and securing it to the print bed with something that can withstand the heat and still be able to remove it when the print was completed would complicate it.

[tom_d]

Probably not, I am a very amateur CAD designer but someone with some experience probably could do so easily. The plastics used in the printing will determine how good of a fit you will get and how much of a press fit is necessary. It may be that no press is needed even if a suitable adhesive were to be used. If the width of the gear isn't critical a mounting flange might give extra adhesive area even.

Maybe an amateur now, but you'll get better with practice.

Is it carved in stone that there must be a press fit holding two pieces together? I wonder what would happen if there were some holes, or other features at the OD of a rim such that the plastic could be printed directly on to the rim, with those features incorporated as some sort of anchoring points. Tossing that out as an idea.

[The Metal Butcher]

Solidedge is good. I use it at work.

[loose nut]

I successfully had a 2-part collar modeled in CAD, 3D printed including accounting for shrink, and cast in bronze. It made an unavailable part available again. And, it was very satisfying.

I have chosen not to acquire a CAD package/3D printer, though, because although the 3D printer is now affordable the CAD package IMO is not. Also, pre-retirement I worked for companies that sold enormous CAD programs to semiconductor manufacturers, and I have a very limited willingness to learn another huge CAD tool.

metalmagpie

Try Solidedge, they give away a full version to hobbyists and it is a very capable CAD program. You don't have to learn everything just enough to do the modelling and export it as STL files. Minimal learning curve since you have experience.

[dalee100]

I successfully had a 2-part collar modeled in CAD, 3D printed including accounting for shrink, and cast in bronze. It made an unavailable part available again. And, it was very satisfying.

I have chosen not to acquire a CAD package/3D printer, though, because although the 3D printer is now affordable the CAD package IMO is not. Also, pre-retirement I worked for companies that sold enormous CAD programs to semiconductor manufacturers, and I have a very limited willingness to learn another huge CAD tool.

metalmagpie

Hi,

And there are a couple of other choices besides FreeCAD. Fusion360 or OnShape have limited but useable free options. FreeCAD is open source and runs locally. And rather than "calling the mothership" to use, FreeCAD is yours to do with as you choose and to install on as many computers as you want with no costs. But it does have somewhat more of a learning curve than say Fusion360.

While you can do CAM with FreeCAD, 3D printers don't use CAM software as CAD/CAM is traditionally thought of. A Slicer is used to generate proper 2D gcode for printing. Design your part in 3D CAD, export the file in the proper format, (generally .stl), and slice. Slicer software is free and open source.

Still I've dropped OnShape, Fusion360, and SolidWorks for the freedom of use that FreeCAD provides. Cost is no longer a reason to avoid 3D CAD.

[thin-woodsman]

Raspberry Pi custom cases. I want to make a case with a fan on one side and a fan on the back.

That's a great example! Electronics cases are always wrong-sized, but easy to design and print (not to mention a ton of existing ones available for free download).

I whipped up an ESP32-CAM with a DHT-11 temp-humidity sensor to stick inside the enclosure (and connect to Octoprint). Making a case for it is pretty much the first job the printer will have when it is back online. That ESP32 stuff is fantastic, I might start making game cameras with it.

[RMinMN]

I successfully had a 2-part collar modeled in CAD, 3D printed including accounting for shrink, and cast in bronze. It made an unavailable part available again. And, it was very satisfying.

I have chosen not to acquire a CAD package/3D printer, though, because although the 3D printer is now affordable the CAD package IMO is not. Also, pre-retirement I worked for companies that sold enormous CAD programs to semiconductor manufacturers, and I have a very limited willingness to learn another huge CAD tool.

metalmagpie

Have you priced out FreeCAD with its integrated CAM package. The biggest cost I see with it is the time investment needed to learn a new interface.

[AD5MB]

Raspberry Pi custom cases. I want to make a case with a fan on one side and a fan on the back.
A stacking sub case. Three 18650 batteries, a BMS, 120 VAC to 12 VDC to 5 VDC converters. An UPS for my network and RPis.
A pocket for glasses and remotes. I have a wall of 9 monitors in one room, 8 in the shop. I need a way to store the remotes in the same physical arrangement as the monitors. Arrange them like Anderson Power Pole connectors, mating slides on the top and sides so you stack them once and they stay that way.

[metalmagpie]

I successfully had a 2-part collar modeled in CAD, 3D printed including accounting for shrink, and cast in bronze. It made an unavailable part available again. And, it was very satisfying.

I have chosen not to acquire a CAD package/3D printer, though, because although the 3D printer is now affordable the CAD package IMO is not. Also, pre-retirement I worked for companies that sold enormous CAD programs to semiconductor manufacturers, and I have a very limited willingness to learn another huge CAD tool.

metalmagpie

[RMinMN]

Can you design the gear so that it sits on a metal rim, maybe with spokes. An old gear of appropriate size where there are a few missing teeth would make a good donor. Machine off the remaining teeth from the rim, then print out the gear "tire" with the needed tooth count and press in place, maybe a few drive pins to help keep it in place.

Probably not, I am a very amateur CAD designer but someone with some experience probably could do so easily. The plastics used in the printing will determine how good of a fit you will get and how much of a press fit is necessary. It may be that no press is needed even if a suitable adhesive were to be used. If the width of the gear isn't critical a mounting flange might give extra adhesive area even.

[epicfail48]

The reason for this is that I could not get prints produced to exact dimensions (a 1/4" hole, for example, might be 0.213, or the edges of a dovetail might be 0.010 oversized and therefore not fit the qctp)\.

This is actually something very important that i feel should get more attention than it does. As fantastic as 3d printing is, on the hobbyist side of things, its not really a high-precision operation, least not how people on this site view the term. Before anybody jumps down my throat, yes, it is possible to get fantastic tolerances on 3d printed parts, even on hobby machines. It is not, however, easy. Trying to hit a +-.005" tolerance on dimensions for a part is going to require a LOT of calibrating the printer and redesigning models to account for things like thermal expansion during the print. Calibrating the printer is a one-time thing, fortunately, but tweaking the model is going to have to be done for every different model and every different type of filament you use. Its doable, but not fun

Bottom line, reasonable expectations are needed. If youre trying to print tight-tolerance parts, either be prepared to spend a crapload of money up front on something like a Form Labs printer, or be prepared to spend more time calibrating the printer then actually printing. Hobby printing is very much a low-tolerance process (again, going by machinist terms). Your standard FMD printer is probably going to be good for +-.015" for any given dimension, something like an SLA printer is even closer, but getting subtractive manufacturing tolerances is a little much to expect out of consumer machines

Course, no reason you can split the difference if you need high-tolerance parts by post-machining the prints. Thats what i do when i need something better than what my printer can spit out

[mattthemuppet]

thanks, yes accelerometers. bottom left you can see a tab with four holes - accelerometers get mounted there, one on each pendulum (as I call the swinging part). I flash a high powered led as a triggered strobe and have been able to balance down to 1 gram cm irrc. or was that .1 gram/cm? can't remember but its pretty low. That's the rotor from a tool post grinder I was fine tuning. Still working on the project overall. 3D printing really improved by taking weight out of - less inertia. The Mark I was all metal...scraped in favour of plastic

that's really cool, thanks for the info!

[Dan Dubeau]

Has anyone here made tooling for presses and brakes using their 3D printer? That's one aspect that really interests me, being able to trial and make custom tooling out of PLA for a fraction the cost of conventional tooling.

It's on my list to try. Just ahead of it on my list is to fix the pump on my press .

There's a few youtube guys who've done it with success.

[garyhlucas]

My problem isn’t the extruder, it extrudes just fine. It isn’t the G-code I read the G-code just fine and I tried printing a file used many times before.