Printing speed

[continuous_david 1]

Hello, im brand new to the community and new to the Ultimaker, although I have used other 3D printers before. So far my experience is positive, but i'm still testing out some of the settings.

What settings to you use for the printer speed? 100%? More? I assume that 100% is equal to the speed that was set in Cura, and the default is 50mm/s

What are some of the drawbacks of higher speeds? Do you need to raise the temperature when you increase the speed?

Any advice is welcome!

-David

[gr5 2,230]

Quote

I assume that 100% is equal to the speed that was set in Cura, and the default is 50mm/s

Unfortunately the defaults change from release to release. You are correct about the 100% part. I believe the recent defaults have been 50mm/sec for printing shell and 80mm/sec for printing infill which will significantly reduce quality due to the speed change.

There are 2 issues - quality and throughput.

Regarding throughput, the UM2 and UMO can print about 10mm^3/sec through the .4mm nozzle when PLA is on the hot end of things (230C). This is really pushing the limits of what it can do -- we are talking over 100psi of pressure in the nozzle and about 10 pounds or 5kg of force on the filemant by the feeder. I recommend 5mm^3/sec max.

To calculate the volume multiple .4mm nozzle by layer height by speed. So 50mm/sec at .2mm layer height would be (easy to do in your head) 4mm^3/sec or a bit above my recommended limit of 5mm^3/sec unless you go up to 240C (less viscous). .1mm layer height you can print faster with no underextrusion.

Regarding quality, any speedchange will result in over or underextrusion for a few milliseconds resulting in lower quality - for example on corners of a cube - the printer has to slow down for the corner but will never go slower than the "jerk" speed of 20mm/sec.

So in summary - if you want extra beautiful quality - go with 20-35mm/sec (and cooler temperatures).

If you want speed here are my recommended top speeds for .2mm layers (twice as fast for .1mm layers):

20mm/sec at 200C

30mm/sec at 210C

40mm/sec at 225C

50mm/sec at 240C

The printer can do double these speeds but with huge difficulty and usually with a loss in part quality due to underextrusion. Different colors print best at quite different temperatures and due to imperfect temp sensors, some printers print 10C cool so use these values as an initial starting guideline and if you are still underextruding try raising the temp. But don't go over 240C with PLA.

[2go 6]

Sorry to revive an old topic but this is a simple question so no need for a new thread. Let's say I have sliced my STL with all the speed options set to 20 mm/s (printing speed, infill speed, etc). However travel speed is 150 mm/s. Imagine I turn printer's knob during the print and select speed to be 300%. My question is does it affect travel speed? Cause you know, all speeds become essentially 60 mm/s but travel speed - a whopping 450 mm/s. Or maybe changing speed in printer's menu ignores travel speed?

[neotko 1,417]

There's are hardcode limits set by the firmware on the rom of the machine. By default AFAIK they are set at 300mm/s for X/Y

You could set your max travel speed, let's say 200mm/s with

M203 X200 Y200 ; Set's max travel speed for x/y

M500 ; Store's the new info into the eprom of the machine firmware

You can edit this on any text file with the extension .gcode

Remember to add a 'return' after the M500 line or the command won't be executed.

[gr5 2,230]

First to answer the question.  The gcode requests a specific speed, that is then multiplied by the percentage you refer to.  Then lots of things happen - for one there is a max speed for all 4 axes (yes, Extruder is an axis - x,y,z,e) and all moves are done linearly (a straight line in 4 dimensional space - no curves allowed!) so if the next move involves any of those axes going over the max speed then it is restricted.  

For example if max XY is 300mm/sec and max Z is 10mm/sec and your move involves moving Z by 1mm and X by 10mm then it will move X at 100mm/sec because any faster and Z has to exceed it's max speed of 10mm/sec.

Then to make things more complicated there are maximium "jerk" and maximum acceleration settings for all the axes.  If your move only involves XY then yes it's much simpler.  But still the axis may never get up to 300mm/sec if the acceleration isn't high enough.

To reduce the Z scar many people like to increase the Z acceleration.  Increasing the max velocity makes no difference (because it's only moving typically .1mm and doesn't have time to accelerate up to full speed) but increasing acceleration makes a big difference - your Z axis movement can go from making a "bzzt" sound to a much faster "tick" sound.

300mm/sec was chosen I believe not because of servo or hardware restrictions but because that's the most pulses the arduino can put out per second if all axes are moving and still have some cpu cycles left over for important things like planning the next deceleration.  But that assumes the steppers on a typical UM which I think is something like 200 steps/mm?  I forget but that would be 300*200 or 60000 steps per second for each axis moving!  Another reason to get the hell off the wimpy arduino and onto a more expensive $10 cell phone cpu.

Edited March 26, 2016 by Guest

[leonardo-aguiar 1]

On 7/1/2015 at 1:11 PM, gr5 said:

To calculate the volume multiple .4mm nozzle by layer height by speed. So 50mm/sec at .2mm layer height would be (easy to do in your head) 8mm^3/sec or well above my recommended limit of 5mm^3/sec unless you go up to 240C (less viscous).

0.4 x 0.2 x 50 = 4

no?

[gr5 2,230]

Oops.  So not so easy to do in my head I suppose.  I corrected it.  Now instead of "far above" the recommended speed the user is only a little above my recommended speed unless they raise the temp to 240C.  But of course, the quality of print will go down if they go that hot.

[leonardo-aguiar 1]

Nozzle Diameter (mm)	Layer Height (mm)	Print Speed (mm/min)	Print Speed (mm/sec)	Throughput (mm^3/sec)
0.40	0.05	1000.00	16.67	0.33
		2000.00	33.33	0.67
		3000.00	50.00	1.00
		4000.00	66.67	1.33
		5000.00	83.33	1.67
	0.10	1000.00	16.67	0.67
		2000.00	33.33	1.33
		3000.00	50.00	2.00
		4000.00	66.67	2.67
		5000.00	83.33	3.33
	0.15	1000.00	16.67	1.00
		2000.00	33.33	2.00
		3000.00	50.00	3.00
		4000.00	66.67	4.00
		5000.00	83.33	5.00
	0.20	1000.00	16.67	1.33
		2000.00	33.33	2.67
		3000.00	50.00	4.00
		4000.00	66.67	5.33
		5000.00	83.33	6.67
	0.25	1000.00	16.67	1.67
		2000.00	33.33	3.33
		3000.00	50.00	5.00
		4000.00	66.67	6.67
		5000.00	83.33	8.33
	0.30	1000.00	16.67	2.00
		2000.00	33.33	4.00
		3000.00	50.00	6.00
		4000.00	66.67	8.00
		5000.00	83.33	10.00

[leonardo-aguiar 1]

I'm planning on testing speeds for my printer so I did the table in the above post. Is it safe to say that since the throughputs are low when printing with lower layer heights, then I can increase the speeds drastically?

[leonardo-aguiar 1]

On 7/1/2015 at 1:11 PM, gr5 said:

So 50mm/sec at .2mm layer height would be (easy to do in your head) 4mm^3/sec or a bit above my recommended limit of 5mm^3/sec unless you go up to 240C (less viscous). .1mm layer height you can print faster with no underextrusion.

Shouldn't it say "a bit below my recommended"?

[gr5 2,230]

2 minutes ago, leonardo-aguiar said:

Shouldn't it say "a bit below my recommended"?

Not really because the temperature wasn't mentioned, right?  I just skimmed it.  But 240C is not normally recommended.  It's a bit below if he prints at 240C I think?  And if he prints at the default temp of, (not sure but let's say) 210C, then he's printing too fast and should drop to 30mm/sec (almost half speed).

 

It's actually all more complicated.  It depends on the feeder and the quality you want.  Overhangs and certain other details (corners) suffer a bit at higher temps and higher speeds.

 

It depends on what your goals are.  I have 6 printers and usually I'm not in a rush so I usually print at 25 or 30mm/sec.  If I'm impatient and don't care about quality I can go as high as 100mm/sec and/or 0.4mm layers with 0.8mm nozzle.

 

Basically the speeds I published in my table are achieved with about 5 pounds (about 2.5kg) of force and the UM2 black feeder can usually put out about 10 pounds (barely if you are lucky and the teflon part is brand new).  The UMO easily puts out 10-12 pounds much more consistently and the UM2 "plus" feeder and the UM3 can put out I think probably about 15 pounds.  Bondtech even higher.  But the UM2 black feeder has the most trouble so I recommend speeds that can handle about 5 pounds typical force.  Your printer might have no trouble with 10 pounds and so you can print double the speeds I posted.

 

[leonardo-aguiar 1]

Thanks for helping gr5.

 

I upgraded my UM2 with the e3D Titan extruder and hotend.

 

My goal is to optimize print time depending on quality.

 

0.8mm nozzle x 0.4mm layers x 100mm/s speed = 32mm^3/s

 

That's way above the 10mm^3/s throughput you talked about for the 0.4mm nozzle. Is there a way to calculate the max throughput for different nozzle sizes?

 

It's finally good to know that the UM2 black feeder should be kept around 5lbs. I think I might have been printing too fast for many of my prints. Too many failures made me buy the e3D upgrade.

[gr5 2,230]

Well I've printed 0.8mm nozzle with 18mm^3/s on a UM2 with a black feeder.  That's about the limit.  0.8mm nozzle can handle a much higher throughput because the larger bore means less pressure needed to squirt out the plastic.

 

Yes the titan feeder is excellent.  Just as good as the UM2.  Particularly the 1.75mm.  e3d concentrate mostly on 1.75mm so their 1.75mm products are excellent.  The 3mm products not so great.  For example they have this all metal hot end (not sure if titan) that doesn't do well with PLA.  You really need some teflon in there for PLA but I think only the 1.75mm version has teflon in it.  Without the teflon you often get clogs/jams because the molten PLA sticks so well to the metal.

 

Anyway, I've done MANY extrusion tests and the best one is to print a cube in cura with 100% infill and no walls.  Ignore the bottom layer.  Set print speed to 100mm/sec for convenience.  Then you can adjust the speed in % and it will equal the speed in mm/sec.  (or slice it at some other speed and simply do the math to calculate your actual speed).  If you are doing 100% infill and you see gaps then you are underextruding.  Print slow at first so you know what a perfect layer looks like then slowly increase until you see gaps - then slow it down until you don't.  Keep note of temperature, speed, nozzle diameter, layer height.  Change temperature and repeat the experiment on the next layer of the cube.  Keep good notes.  Go all the way down to 180C as that's the region where you get the best quality but are forced to print very slow.  Repeat for other nozzle sizes.