3D printing/modeling advice

Last week I had one of those "I can build a better mouse trap" moments and after many lost hours of sleep, I'm working on a SketchUp model of my idea.

I'm planning to use Shapeways as my prototyping service, so I've been consuming all the knowledge I can about their service and limitations.

For the hive-mind: I need help figuring out how to deal with fit and tolerances. The thing I'm building is small: 24mm x 34mm x 9mm at the moment. It's got hinged doors in a couple places, and I would like them to close tightly and stay closed. I was thinking of locking the door to the face with friction-fit nubs or 'L's in the corners. Do I simply assume that the printing accuracy will magically work out right, or do I design in some slop in the size/location of the nubs vs the matching groove? How do I keep slop in individual parts from adding up? How do I keep the small part size and relatively poor printing accuracy from causing me headaches? (+/- 0.15mm is accurate until you look at the 2mm pin in my hinge...)

The material information is here.

Any rules of thumb, online resources, etc that you can recommend?

Much obliged!

Bump just because this is a great topic and I'm sure there is someone here that knows the answer.
BTW, MAKE magazine has a great article right now, covering printers from 300 to 3000 bucks.

You can't keep the slop from adding up, its part of the process.

You can do things to mitigate it though.

Either make the parts big enough, substitute more accurate parts like turned and ground dowel pins for the hinge pins, or go to a more accurate process.

If you want to make a hinge pin that you know will fit, you will have to model it as 1.85mm OD less whatever clearance you want in it. The Machinery Handbook has info on all the fits, locating, press fit, sliding fit, etc. Use those tables with the known accuracy issues of rapid prototyping to do your design work.

93gsxturbo wrote: If you want to make a hinge pin that you know will fit, you will have to model it as 1.85mm OD less whatever clearance you want in it. The Machinery Handbook has info on all the fits, locating, press fit, sliding fit, etc. Use those tables with the known accuracy issues of rapid prototyping to do your design work.

This is where I start to get lost.

This is the beginnings of my hinge.

The center pin is 1mm in dia, in accordance with the material's "min wire free". There's a 0.5mm gap for clearance (again, to match the specs), then a 1mm thick outer sleeve.

I don't really get to choose the fit from the Handbook (I have seen that referenced before), it seems to be dictated by that 0.5mm gap and the min sizes afforded by the material. Looking at the numbers now, I could make that inner pin 0.85mm in diameter, but that's not really solving the problem, is it? The overall "slop" is sort of limited by that 0.5mm required clearance.

(I guess I could build this as separate parts that snap together, but not sure 100% that would work)

As far as making a hinge pin that fits into a hole in one of my shapes:

This tutorial seems to show that small holes are wildly inaccurate in this process. Looking at the final chart at the bottom, would I be better off making a 1mm diameter hole 1.2mm deep in a face and then building a matching pin?

Now that you've made me think about it, I think I see the answer, but someone telling me I'm sane would be good right about now...

I design lab instruments in 3D CAD, make drawings from the models, machine, fabricate, and assemble the parts then make whatever "it" is work.

A colleague who has no experience with mechanical design or making things has bought a Makerbot, the a newer one. I've found for parts w/ very loose tolerances, like .040", it will sort of work. For comparison, I can mill to .001" and turn to .0005" relatively easily.

Speaking specifically to your case, I wouldn't try to make any feature smaller than about .080", and I'd drill all the holes smaller than that or ream if accuracy was needed. I've got a lot of design-for-manufacture experience and have mentored/taught manufacturability and machine design to engineering groups for a number of my employers - I'd be happy to take a look at your work and guide you toward the easiest path to success. It may be easier to machine the part from Delrin or aluminum than to print it. Or prove the concept w/ a 2x size part provided it's not rendered impossible due to some law of physics regarding fluid or gas density or strength-scaling of material properties.

When I've printed a part on the Makerbot my assumption is that I'll be scraping and sanding and filing to make things fit, and that it's for validation only. The assertions that "Quadcopters will be delivering goods that are printed at the point of delivery" are made by people with no background in manufacturing, I can guarantee.