Filament tolerances

Help with printing with specific plastic filaments.
robajohn
Posts: 135
Joined: Sat Oct 15, 2016 9:36 am

Filament tolerances

Post by robajohn » Tue Dec 13, 2016 8:42 am

Ok I am about 2 months into this hobby and on another thread about accuracy problems I was told to put a caliper to my filament and put that exact number in Cura.

Cool no problem.

So I started doing that and I was shocked at the variations- I ranged from 2.9 to 3.13 mm in the same roll sampling every meter or so. Ok well that was the cheap stuff - you get what you pay for, right? So let me check this $36 per kilo roll stuff I have set aside for special projects... pretty much the same variation! (and Cura was set/preloaded to 2.85 so that would seem to be an issue)

So two questions...

1) What is would be a "normal" range? Are the rolls I checked "normal" at +/- .10 mm?

2) I assume I should average the numbers to get a value to feed into Cura? Would you more experienced folks agree?
Last edited by robajohn on Tue Dec 13, 2016 9:26 am, edited 1 time in total.

nopick
Posts: 1029
Joined: Mon Sep 09, 2013 11:14 am

Re: Filament tolerances

Post by nopick » Tue Dec 13, 2016 9:12 am

It is normal to have a little variation in size. Taking an average over a couple meters is the best option.

User avatar
PHugger
Posts: 77
Joined: Sat Feb 27, 2016 7:10 pm

Re: Filament tolerances

Post by PHugger » Tue Dec 13, 2016 12:58 pm

I would also suggest practicing with the caliper.
It's very easy to measure at an angle which will give inconsistent and inaccurate results.
With practice you should be able to get results that vary only by a few hundredths of a millimeter. Take the average.
As you have noted - cheap filament is less consistent, but this has been getting much better for the past few years.
You may also find filament that is not perfectly round where you get a consistent max and mix - take an average.


Best regards,
PCH

robajohn
Posts: 135
Joined: Sat Oct 15, 2016 9:36 am

Re: Filament tolerances

Post by robajohn » Tue Dec 13, 2016 1:06 pm

PHugger wrote:I would also suggest practicing with the caliper.
PCH
I guess a decade as a machinist did not properly prepare me for using calipers. Damn! ;)

User avatar
PHugger
Posts: 77
Joined: Sat Feb 27, 2016 7:10 pm

Re: Filament tolerances

Post by PHugger » Tue Dec 13, 2016 1:24 pm

I didn't mean to insult you, just sharing what I've learned from my experience learning how to do it.
Variations of 2.9 to 3.13 mm seems a bit much (again from my meager experience).
I make it a habit to recheck the loaded filament each time I get ready to slice/print something.
I mark the roll with a good average reading so I can tell if something has changed drastically.


Best regards,
PCH

robajohn
Posts: 135
Joined: Sat Oct 15, 2016 9:36 am

Re: Filament tolerances

Post by robajohn » Tue Dec 13, 2016 2:59 pm

PHugger wrote: I mark the roll with a good average reading so I can tell if something has changed drastically.
I love that idea!

GJG
Posts: 72
Joined: Sun Oct 09, 2016 10:53 am

Re: Filament tolerances

Post by GJG » Tue Dec 13, 2016 6:11 pm

Hi
I have a bit different take on dimensional tolerance in 3D printing than most people do.
I've had items printed by a number of people over the past 3 years or so.
My results for being able to assemble things without trouble is always consistent, no mater what machine printed it.
In the case of my relatively new TAZ6, I look at the STL line.
I use Cura as my slicer, seems to work pretty good for me.
It seems to me, as though there is only 1 STL line of travel for the extruder.
The sliced file always follows the STL line.
It appears that when doing an outside parameter, the filament squish out accounts for about 0.2mm of larger than the STL line.
When doing an inside parameter, the filament squish out accounts for about 0.3mm smaller than the STL line.
When considering that the extruder makes a 0.5mm noodle when it's extruding, this makes sense.
..
CURA and maybe every other slicer does not accommodate for inside line and outside line.
My TAZ6, using standard profiles Cura Lulzbot edition slightly overextrudes.
I don't see that as a problem, at least for now.
Keeping the STL line of travel in mind, I always modify my drawings to account for the fact of STL line that
I'm either imagining, or is truly a fact.
My parts always fit pretty much exactly how I design them.
..
Sorry about my long windedness here, this is just my 2 cents worth.
I hope others keep chiming in here, as this is a very interesting topic.

robajohn
Posts: 135
Joined: Sat Oct 15, 2016 9:36 am

Re: Filament tolerances

Post by robajohn » Wed Dec 14, 2016 8:43 am

GJG wrote: Sorry about my long windedness here, this is just my 2 cents worth.
Are you kidding? That is a GREAT post!

I am starting to kick myself here as I started a thread under hardware about dimensional tolerances and what I learned there led me to make *this* post about the variability of filament (as one of the factors listed in that thread) but your post would have been perfect in that thread. (and it is still awesome!)

I have been goggling and reading for 2 days now and learned a LOT but this is the first time your perfectly logical line of thought has come up!

It is extremely helpful! Taking that tidbit of knowledge PLUS what I had already learned about the first layer having extra squish to boot PLUS the variation in filament I can now account for the problems I am having getting parts to fit together. More on that in a bit!

robajohn
Posts: 135
Joined: Sat Oct 15, 2016 9:36 am

Re: Filament tolerances

Post by robajohn » Wed Dec 14, 2016 8:54 am

PHugger wrote:I would also suggest practicing with the caliper.
PCH
So just because I hate being one of those people who ask for help then don't listen when they receive it I went home last night and took the caliper to the roll I have on the printer now again got pretty consistent readings over a meter. I did not want to unroll to much so only a meter. this is a el-cheapo roll. Since I was running a consistent 2.9 I thought, "HMMM, maybe PCH was right I was just sloppy with the caliper the night before?"

Also I did NOT take the good stuff out of the vacuum bag it is stored in

So I kicked off a 20 mm cube calibration run and about 10 layers up saw the extra squish out (part getting wider) paused the printer and now we are 3.1. AHA! Un-paused and went about 10 more layers and the "bulge" around my cube eased back in. Sure enough, the filament was back to 2.9

Very informative!

Now bear in mind this is $8 a roll stuff we are talking here. So "you get what you pay for". What I need to do now is get my $36 roll out and do the cube again. and see how that goes.

ttabbal
Posts: 298
Joined: Wed Nov 11, 2015 4:20 pm

Re: Filament tolerances

Post by ttabbal » Wed Dec 14, 2016 10:57 am

Sadly, tolerance on filament isn't what we would like. It becomes more of a problem when you have parts you want to fit together. It's difficult to get things to fit properly sometimes. I find it helps to make everything else as accurate as I can. Calibrate the e-steps and set the idler tension properly so you get the right length extruded. Calibrate the Z-offset so you get the right height, particularly for the first layer. The default Cura profiles over extrude the first layer, turn that off. I also find under extruding slightly helps offset the filament tolerance. Don't go too nuts there or you don't put down enough plastic and cause other problems. I find I have to do a few tests on each new roll, but once I get it dialed in it's pretty consistent for most rolls. I also like to test a meter or so in 4+ places every print or two. It helps catch the odd ones that have a big change in the middle. It doesn't seem to happen that often to me, but it does happen.

Post Reply