Looking for tolerance design guides for cylinders to hole fits. Also screw bosses and id for screws that work well. I’m sure it depends on the material or is this all trial and error by each material type? I was going to prototype some smaller sections so I don’t have to print the whole housing for an automatic cat feeder. An input would be great help.
What I do to test new fittings is I’ll design a small part that has what I need to test; holes for rods, bearings, or nuts and I over-size the holes in increments of 0.1 mm. I’ll print 4 of them and test parts for fitment before printing the entire project.
@Adam_Steinmark thanks for the tips that was what I was thinking. I will give it a try.
well at least to get some sizes for nuts and bolts and diameters for the holes to make threads do some lists come in handy so you dont have to meassure them - https://www.engineersedge.com/screw_threads_chart.htm - having a proper drill set is also a blessing with the in between sizes like fe 3.5mm/5.5mm and so on 
My approach, if I were being particularly “scientific”, would be to first establish the target sizes for the hole and shaft. That would be based on guidelines for sliding, clearance, or force fits. Those clearances, IIRC, are based on the diameters, so that the difference between the shaft and hole diameter are based on the diameter. The Machinist’s Guide ( http://www.penntoolco.com/morse-pocket-size-machinist-guide-1001/) I believe has that information (US measurements). The Google Books preview ( https://books.google.com/books?id=sdcjCAAAQBAJ&pg=PP3&lpg=PP3&dq=machinist+guide&source=bl&ots=pKRk55k8Ei&sig=MazqEQsGkRYCqoO0Nw9y-5FLsO4&hl=en&sa=X&ved=0ahUKEwjqzNiahsvXAhWhi1QKHeqHByYQ6AEIkQEwDQ#v=onepage&q=machinist%20guide&f=false) which I think is the same book, shows page 22 as having a press fit and slide fit chart. I’d combine that information with knowledge of the performance of my printer either already established or trial and error to get the fit desired.
HTH.
http://www.penntoolco.com/morse-pocket-size-machinist-guide-1001/
Thought a little more about it. See https://www.engineersedge.com/general_tolerances.htm which gives classes of fit (loose to press) and leads to tables including dimensions and recommended clearances. They do vary with the diameter.
Page 4 of this, http://www.ucc-udb.com/V01-V07-udb.PDF, has tables of fits as well.
This, http://www.amesweb.info/FitTolerance/FitToleranceImperial.aspx , discusses the impact of tolerances on fits and necessary clearance, so speaks quite directly to 3D printer performance.
@Googolplex_Goku @Kleinfeld_Technical This works all well and good with machining but when 3D printing thermoplastics you run into the issue with the material deforming slightly while it cools. There may be different sizes for preferred clearance between ABS and PLA. I’ve had fittings for nuts designed for parts printed in ABS be too loose when printed in PETG.
@Adam_Steinmark Absolutely right that the materials involved affect the clearances needed. That’s why my first post mentioned the performance of the printer (and by implication the materials). The machinist’s tolerances are a starting point.
@Adam_Steinmark kk good to know i only printed pla because of no enclosure for my printer and i listen to reviewers didnt want to be the 1000th user with material problems because of the missing housing but its almost done the housing so i can finaly try other materials yay ^^ but such tables come in still handy
I find that large parts printed in ABS will shrink about 0.5-1% in the XY plane after cooling. They don’t really shrink in the Z plane as the material is placed down at a set height even if the previous layers have shrunk down a minuscule amount (though this could be partially responsible for the cracking that can occur on tall prints when not using a skirt or heated enclosure)
As for circles, part of the problem with tolerance is that most models will be printed in STL format and arcs will be converted to line segments. So this add to the difficulty of predicting the final inside diameter depending on the resolution of the arc. Ie. The measured difference in dimensions when measuring say an octagon from “point to point” vs. measuring “between the flats” will change depending on the number of sides in the polygon. Though the dimensional difference will be less the higher number of sides used.
Add to this the effects of the “pull towards the inside” effect of printing small diameters, plastic shrinkage, slight layer misalignment, and under/over extrusion and you can see where it is difficult to guarantee fit without using good old trial and error.
Generally I find adding about 0.1mm to the diameter of my holes when designing helps a ton. If you need utmost precision, professional appearance, and longevity then heat-set inserts are a good option too. Metal sleeves for axles, and bearings can often be set into the plastic in the same way.
Ankit
I always give myself a clearance of .2mm normally works