General tolerances for part design
 
Notifications
Clear all

General tolerances for part design  

  RSS
Lintsi
(@lintsi)
Active Member
General tolerances for part design

Hi,

I ordered my MK2 printer last week and now while waiting I'm already designing some models to be printed. While founding many answers online to my various questions one question raises above others. What tolerances and design/manufacturing methods(or tips) have you used for creating small tolerances between different parts (Slic3r settings, print detail level, materials, general dimensions)?

For example I'm now designing an assembly where I need to make a bearing housing for a traditional ball bearing and an axle which goes through the bearing. The bearing's true outer diameter is 21,95 mm and inner 6,95 mm. Naturally I want those fittings between parts to be as tight as possible. In this article ( http://makezine.com/2013/12/11/top-ten-tips-designing-models-for-3d-printing/ ) she advises to use 0,2 mm for tight fits and 0,4 mm for lose fits. Does these values apply while printing with MK2? Before designing my parts any further I would appreciate if you could share your experiences and tips about this matter.

Br,
Lintsi

Napsal : 05/10/2016 4:22 pm
PJR
 PJR
(@pjr)
Antient Member Moderator
Re: General tolerances for part design

Hi Lintsi

This really does depend on many factors, build quality, calibration, model design and slicer parameters.

For me, I generally use a 0.2mm difference to get a reasonable fit; I guess 0.15 would be tight and 0.3mm loose. But that is after more than 18 months experience and playing around.

Peter

Please note: I do not have any affiliation with Prusa Research. Any advices given are offered in good faith. It is your responsibility to ensure that by following my advice you do not suffer or cause injury, damage…

Napsal : 05/10/2016 5:38 pm
JoanTabb
(@joantabb)
Veteran Member Moderator
Re: General tolerances for part design

Hi Lintzi,
I see Peter beat me to the answer... 🙂

There a lot of items to take into account when designing parts for printing,
I will start a list with some and no doubt others will add their own items:
1, your printer will be inherently slightly inaccurate. so all things being the same, if you print a vertical 20mm cylinder, and a vertical 20.2 hole for it to go into, and your printer is well setup. it is likely that you will get a tight fit (because the same inaccuracies affect both items to a similar degree.. (this may alter if you rotate the relationship of the cylinder and hole, because there is probably a small difference between the X and Y axis movements on your printer, and also the single sided part cooling fan can cause differential cooling...)
if the edge of the 20mm hole was on the build plate and / or the end of the cylinder was on the build plate, it is probably that you will have some distortion which will cause problems with the 'Fit' because it is probable that there will have been some spreading of the model where it meets the build plate... some of this will be due to the squish on the first layer which helps the plastic stick to the build plate. putting a small chamfer on the cylinder model and the hole model will often mitigate this issue.
however if you then replace your 20mm cylinder with a 20mm machined part, it is probable that your fit will be 'different' simply because of the different tolerances between the printer and the lathe or whatever machined the new part.
2, different plastic filaments will have different expansion and contraction rates, which can result in different sized parts. so one print mat be ok, and the same 'G'code with a different filament could be unsatisfactory.
3, if you print your cylinder and your hole with their centre-line horizontal instead of vertical, it is probable that your items will end up distorted especially if the curved surface of your cylinder is resting on the build plate, (at best this area will have a flat on it, at worst the whole part will end up oval instead of circular.
4, it is inevitable that there will be some surface rippling on your plastic parts. the degree of rippling can be adjusted by changing feed rates, temperatures materials, cooling etc.
5, ambient temperature can affect the model dimensions.
6, Surface blemishes can cause problems.

so it is probably a good idea to allow for test print and revise model activities within your development process...
alternatively, you could purposefully make the bearing housing undersize and ream it out,,, similarly the shaft could be made slightly oversize and finished down to a good fit...

there is some scope with lower melting point parts, to heat the bearing and force it into position, allowing the bearing to re form the plastic surface, or purposely making the 'fit' loose, and using loctite retaining compound http://www.loctite.com.au/3320_AUE_HTML.htm?nodeid=8802648195073 or similar, to give the final fit.
(you would have to assure that this material doesn't react with your filament, (i understand that Polycarbonate can become brittle when affected by other materials.))

I hope this helps.
Regards Joan

I try to make safe suggestions,You should understand the context and ensure you are happy that they are safe before attempting to apply my suggestions, what you do, is YOUR responsibility. Location Halifax UK

Napsal : 05/10/2016 5:46 pm
Lintsi
(@lintsi)
Active Member
Topic starter answered:
Re: General tolerances for part design

Hi Peter and Joan

Thank you for your replies!

I will bear these answers in my mind when designing parts and printing with my (future) printer setup. Trial&error method is a quite good way to learn in this field on manufacturing. 🙂

// Lintsi

Napsal : 05/10/2016 6:44 pm
gz1
 gz1
(@gz1)
Estimable Member
Re: General tolerances for part design

Uhm, for starters, do not use loctite with thermoplastics.

For "holes" for things like bearings, a lot of things will affect your accuracy.

1. Arc compensation http://reprap.org/wiki/ArcCompensation
2. XY calibration - if your machine isn't perfectly orthogonal, it will reduce the effective ID of a hole
3. Layer registration - if your machine doesn't register layers accurately, it will reduce the effective ID of a hole
4. Hole orientation -- vertical hole vs. horizontal hole makes a difference

So basically most inaccuracies tends to reduce the effective ID of a hole you try to produce.

As a general rule, going roughly by the arc compensation guide, for D above 3mm, I usually add 0.2mm for a "normal" fit. It also helps to reference all holes by the center instead of their edges so you can change hole tolerances without screwing up your design.

If you think about #1, you realize that it's also really hard for printers to produce perfectly sharp internal corners, like if you're doing a square and you need the corner to be sharp, it can't do it. So where possible, I turn the corners into imperceptible holes, kind of like this:

http://www.thingiverse.com/thing:963466

Where a Lego piece with sharp corners was expected to be pressed into that face.

The other tricks will be shape dependent.

Napsal : 05/10/2016 7:09 pm
Share: