Recommendation needed for accurate printing

I am working on some parts I need printed from some sort of glass filled nylon. Is it possible to get 3d prints to hold size within a few thou, or less? The shops I've dealt with so far claim they can but the the parts I get back are more like +/- .01" and of course any of the mating parts are .010" always seem to be in the wrong direction so I end up with .020" of interference. Machining these parts isn't really practical, and making a quick mold would be out of the question as I would need lifters on both side of the mold, slides, and multiple pass through cores.

These are small parts that can fit in your hand. I'd like to have them made from the glass filled nylon so I can do some basic durability testing. But if it's beyond the realm of that material to print something like this I would consider a more dimensionally stable material.

Dave Cross said:

I am working on some parts I need printed from some sort of glass filled nylon. Is it possible to get 3d prints to hold size within a few thou, or less? The shops I've dealt with so far claim they can but the the parts I get back are more like +/- .01" and of course any of the mating parts are .010" always seem to be in the wrong direction so I end up with .020" of interference. Machining these parts isn't really practical, and making a quick mold would be out of the question as I would need lifters on both side of the mold, slides, and multiple pass through cores.

These are small parts that can fit in your hand. I'd like to have them made from the glass filled nylon so I can do some basic durability testing. But if it's beyond the realm of that material to print something like this I would consider a more dimensionally stable material.

Click to expand...

3d printing is like inkjet or laser printing, there is an inherent raster resolution, ask them what is the resolution of their printer. Normal lasers are either 300 or 600 dpi some are 1200 dpi. So assume the worse and assume the error is half the resolution so for 300 dpi the best you get is +-0.0066" for 600 dpi half of that and so on. Any 3d printing shop that claims +-0.001" tolerance must be printing at least 2000 dpi. 1200 dpi would get you in the +-0.002 (50 micron) range. I would treat 3d printing the same as casting. It needs machining for tight tolerances. I have no for material recommendations. I know all the plastics are thermo plastics, and i do not know much about fillers. You can get 3D printed metals, but it cost a lot. You can get a reasonably priced printer and experiment yourselves. Cheap printers will be slow, so if it is production work, it may not be even feasible. it is like the old saying you can have it cheap, fast or accurate pick any two.

it is easy to get blindfolded by the claims of new technology.

dee
;-D

We're going to be molding this part after the design work is done. We do lots of machining here but these parts are really impractical to machine. I've always been skeptical that people can print and hold tolerances under a few thou, but I'm sure there are some competent folks out there who can.

Dave, you are correct in your conclusions. Frankly, when I do it, I make every effort to control things as well as possible,even paying attention to orientation, area of the build platform, and other nuances that can and will greatly affect one's success. At the end of the day though, I'll always run a drill or ream through critical holes to make sure... Let me know if I can be of any assistance.

Dave, with SLS it is going to be hard to hold better than the +- .010 as most companies run a .005" to .006" layer thickness and the nature of the process is the OD tends to be a bit larger and the ID tends to be a bit smaller due to a fine layer of powder adhering/bonding to the outer surface during the sintering process.
For higher accuracy the SLA process is a better solution as we build in .004" layers with our SLA machines and you don't have the 'growth' on the outer surface as you see with the SLS. If the nylon is not critical to the model than using the SLA with one of the newer epoxy resins would give you a more accurate model, and these newer epoxy resins are much more robust than SLA materials used prior.
Building smaller SLA models with +- .003" is very doable. To hold within a thou or two would require a test part to adjust the scaling factors used.
Hope this helps.
Don

Well that is kind of what I figured.

I guess I can keep doing what I am doing. Get the parts printed and then try and machine the mating features. It kind of works but machining one thing means wiping our another feature. It might just be one of those projects where I have to suck it up and start the mold, and just build it in a manner that allows for some adjustment. It's just a tricky mold to build, and I was hoping to cut down on the learning curve a bit