Printed parts incorrect sizing

Hi,

I am currently building a Prusa Mendel.

I have printed off all the plastic parts on a Stratasys Vantage SE and begun assembly. I have noticed that some of the parts "seem"wrong. I say this because for eg:

On my x-end idler, where an M8 nut is supposed to sit in the hexagonal recess the M8 nut is loose. This seems to be occuring on all of my parts.

In both Autodesk Inventor and Solidworks and Rhino the parts measure identical to the physically built parts. Are they designed to be loose????

I have seen videos of people assembling their parts and they seem to be a tighter fit.

Any explanation most appreciated.

Attached is an STL of the file I built.

Thanks,
Mark

How loose is loose? There should be some oversizing but still close enough to act as a wrench and hold the nut captive. Trying to print things directly to size can be a hassle because it leaves no room for shrinkage or "artifacts" that occur during the printing process and also bolt machining tolerances. M8 hardware should be 12.73 - 13.0 mm across the flats (someone correct me if I'm grossly mistaken)

How big are your openings? I'd say if its sufficiently close enough to hold the nut, its correct.

JW

---

Josh

//++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
//Prusa / RAMPS_1.4 & ATMega_2560 / Slic3r > Pronterface > Marlin / Luck
//++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

I suspect your printed parts are very accurately made, whereas the stls are made to be printable on worse printers with larger tolerances, so it is better to have people like you have parts that are slightly too big than it is to have a load of people not be able to insert pieces. As long as the nut won't rotate, it'll be fine, but you could always just glue a thin strip of something to the outer edges of the nut for a tighter fit.

I had the same problem, when i printed out the Prusa v2 x-ends the hole for the z-axis leadscrew nuts were way too big... so much so that the nuts could spin all the way around... I posted a topic about it here [forums.reprap.org]

It wasn't a problem with my printer, that was just the size that the .stl's were made. I changed the size in the .scad file and re-compiled it, then reprinted them and they came out perfect.

1st of all I would like to thank everyone so far for there quick response to my problem. I think this forum is awesome and will try to help out everyone I can whilst I am using it and hope you guys continue to do the same.

Josh: I'm talking rather abnormally loose. Just about as loose as daufhammers but probably not quite. The M8 nut does not turn when the threaded rod turns through it. On other parts however I have m3 nuts that are doing full rotations. I may have to glue a strip of something on the sides of the nut as James Glanville has provided as a solution.

My question then is, will it matter if the nut is not completely centred? Could this potentially effect my print quality?

Should I try re-sizing the holes in inventor and rebuild?

Should I perhaps glue the nut and then bog around it so it is completely locked in place before threading?

CHeers,
Mark

Just a thought... could I use a locknut on each side of the hexagonal cavity to hold the threaded rod in place?

I'd say that if you have the ability to fix the models yourself, I'd do it. CAD seems to be one of those things that completely escapes me. If you can correct it now and get things printed properly the first time instead of trying to shim and force it to work, you'll be better in the long run. Gluing it might also add some unforseen permanance that will be twice as hard to correct later. Its really committing to something that you know isnt right in the first place. Otherwise if you have any problems later down the road, it will be the first thing that you start to question. I've been using some "scrounge" steppers and any time I run into the smallest problem with printing, its the first thing that pops in my head. Someday I'll just pony up and buy some standard issue steppers. Ounce of prevention...

Out of curiosity, whats the source of your models youre printing? Are they standard models that everyone else has printed and assembled a thousand times? If you're using a model source like https://github.com/prusajr/PrusaMendel I would start to question something else. But if youre printing some variation or derivative from Thingiverse, I might suspect the model first.

JW

---

Josh

//++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
//Prusa / RAMPS_1.4 & ATMega_2560 / Slic3r > Pronterface > Marlin / Luck
//++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++