A tougher extrusion test :-o

[illuminarti 18]

Since someone asked me to make it... I did. This is a revised version of the extrusion test cylinder that runs from 8 to 15mm³/s - a handy test for those folks whose printers can complete the standard version, and want to know where the real limits are. This cyclinder has a 50% larger diameter, to provide the same layer cooling time at the 50% higher speeds used here. (This one tops out at 150mm/s, with 0.25mm layers).

This is really fast extrusion - and arguably it's so fast that it's not actually a lot of use for anything. Printing real world shapes with corners and overhangs isn't going to turn out well at these sorts of speeds. And if your printer can't print any of this, then there's really nothing to worry about at all. Even 8mm³/s is more than you're likely to use in most real applications.

I tried a couple of test prints using Faberdashery filament (on a lazy susan, on the floor), with my factory installed (now 6 months old) UM2 feeder assembly. At 240º it started to fail pretty heavily at 15mm³, but was perfect up to that point. I upped it to 250º, and it only skipped once at 15mm³ (although that was enough to mess up the back of the print quite a bit, as it was never able to extrude properly over the missing part in the back. But it didn't actually skip again.

 

(There's also a skip at 12mm³/s which I'm not counting as a failure as it was due to my cat snagging the filament. Also, the first layer or two just above the indicator lines at the speed changes can start to look a little under-extruded because the lines are inset, and it's hard for the filament to re-establish a proper grip for a layer or two at that speed - that is a problem for many real-world prints, but isn't really a sign of having hit a limit for the extruder itself. Certainly there's no filament slipping or extruder clicks happening).

 

 

 

[illuminarti 18]

For the sake of completeness, here's another link to the original 3-10mm³ version of the test:

http://umforum.ultimaker.com/index.php?/topic/3976-almost-always-missing-layers-underextruding/?p=33427

 

[ian 32]

im going to give this a go tonight !

Thank you for the great work !

Ian :-)

 

[IRobertI 521]

Something I've been meaning to ask about these, does it account for acceleration? Since the cylinder is a bunch of short segments wouldn't that affect the actual throughput compared to a box shape?

 

[illuminarti 18]

Because it's a cylinder, the velocity change at each line segment junction is pretty small, so jerk shouldn't have much effect. (Whereas a box will speed up and slow down markedly). You probably lose a little bit of speed to acceleration and path planning effects - but I've timed it doing the circles and they're pretty close to the nominal speeds that are being asked for.

 

[cor3ys 28]

Hi Guys

You either love these tests or you hate them, i love them when they work and hate them when they don't...

A little bit of history:

when i got my machine i could print the 3 - 10 extrusion test then when my machine hit about 3 months old, i then had massive under extrusion and could only complete 5mm 3/s I obtained a new Teflon and was back in business, I could complete all the way up to 10mm 3/s

Over the last couple of weeks (another 3 months on) i was suffering massive under extrusion again and could only complete up to 5 -6mm 3/s, Today i received some more new Teflons and have fitted one and below is my results...

3 - 10mm 3/s using colorfabbs PLA at 230 degrees half a spool on back of the machine.

 

8 - 15mm 3/s using colorfabbs PLA at 240 degrees half a spool on back of the machine.

 

 

As you can see i started to get some under extrusion at 15mm 3/s my feeder was not skipping backwards though, i am more than happy with the results.. Later i will try again using Faberdashery PLA lose and upping the temp to 250 just like illuminarti to see if i can complete the 15mm 3/s perfectly as i was not getting any back skip on the feeder...

 

[cor3ys 28]

Ok i could not be satisfied with my test above, so i loaded my machine up with some faberdashery PLA and set the temp to 245 degrees and let her rip, as you can see i completed the test fully and still had no back skip of the feeder so i could most probably go some more :grin:

 

[DidierKlein 729]

So i tried it yesterday. The first test was done a couple of days after receiving my um2 and it went fine until 10mm3/s.

Here are the results:

 

I had some slipping of the feeder and resulting was some underextrusion.

The parts where this happened was when i left the machine do all by itself. If i helped the filament to be unrolled by hand and made sure it entered the feeder quite vertically it was all good.

You can see on the pics that the top part at 15mm3/s was ok.

I'm planning to put robert's feeder in place and redo the tests to see if it changes anything. I'm also planning to print his spoolholder after that.

The test was done with ultimaker gold PLA at 240°c

 

[IRobertI 521]

I'm planning to put robert's feeder in place and redo the tests to see if it changes anything. I'm also planning to print his spoolholder after that.

 

Probably not, but it has other advantages that can be worth the hassle of switching feeders

 

[skint 62]

Somebody needs to build me a machine where I can put all these stupid cylinders, melt them down and re-use the pla lol I think I have enough tested pieces to fill 10 complete coils !

 

[conz 0]

I'm interested in this topic too. Not just as a joke.

Recycling would be to much OT for this thread. I have started a new one here:

http://umforum.ultimaker.com/index.php?/topic/5550-extruding-your-own-filament/

 

[hummel 0]

My UM 2 did 10 mm3/s perfect but failed on 11 mm3/s.

 

[gr5 2,224]

If you can get 8mm^3/sec then I consider that to be a properly functioning UM2. If you fail at 3mm^3/sec or lower then that's a problem.

 

[123bob 0]

Is there a link somewhere to download the gcode file for this test? I can't see one in Illuminartis post. Am I missing something?

I would like to try it and see where my machine tops out.

Thanks,

Bob

[robson 4]

Would like to try this just for fun. Where can I download the test?

[BustBloodvessel 0]

where do i download the test cylinder ?

[Dim3nsioneer 558]

Somehow the gcode seems to have vanished. I put it on my dropbox (no guarantee how long the link will work!):

https://dl.dropboxusercontent.com/u/15273556/Ultimaker/gcode/UM2%20-%208-15mm3%20-%20Extrusion%20Test.gcode

I hope @illuminarti is fine with that. It was his gcode.

Edited April 10, 2016 by Guest

[ultiarjan 1,223]

Never did try this one before, but with the 2+ upgrade kit installed I thought let's see how the new feeder will cope

used 230c until 14mm/s, then turned in up to 240c.

Pretty amazing when you know it's also almost the last filament on the spool with a lot of tension in it...

Although these are crazy speeds to use with a 0.4 nozzle, it's nice to know the feeder can cope with a lot of volume when you're doing stuff like vases with the 0.8