Specimen with 100% infill underextruding towards seamposition

[breakkid 0]

Hello guys, my name is Damian and this is my first post. There is a problem that I can't solve.

 

I'm printing a lot of impact strength specimen which have the requirement to be solid (100% infill), having a line width of 0,6 mm and a top/bottom & infill line direction of 0°. Other speciment with different setting I had to print were fine, but all those specimen with 0,6 mm line width and top/bottom & infill line direction of 0° were just underextruding. At first I checked max volumetric flowrate, which looked absolutely fine. After that I reduced the print speed from 60 mm/s to 40 mm/s. The lines seemed to look better, but they were underextruding towards one side of the print. I've leveled the heatbed manually. Still the same problem. Then I reduced even to 20 mm/s print speed. Now only 1/4 of the specimen is underextruding. After changing the seam position from back right to back left I discovered that the underextrusion is always moving towards the seamposition. And from this point on I don't know how to continue, Can you guys please help me?

 

I've attached two pictures and the .3mf-file. On the first you can see the underextrusion under a microscope on the second you can hopefully see the gaps growing towards seam position. Sorry for my camera quality (2nd from top printed with 20mm/s). Unfortunately it doesn't get better.

 

I'm slicing with Cura 5.6 and printing on an Ultimaker 3 Extended. I'm using the standard brass nozzle for ultimaker with a diameter of 4 mm.

UM3E_impact strength specimen.3mf

[Slashee_the_Cow 280]

4 hours ago, breakkid said:

I'm using the standard brass nozzle for ultimaker with a diameter of 4 mm

I'm going to assume you mean 0.4mm 🙂. 4mm would be one hell of a nozzle.

 

Also: infill isn't the problem here. At least not yet. The bottom layer is skin.

 

Does it start when leaving the Z seam or on the way back?

 

You're not at 100% flow, but that should affect things equally. I normally set my initial layer flow to 105% to aid adhesion, but if adhesion isn't a problem, don't worry about it.

 

Although I can't help but notice: you're printing PLA at 250°?!?!!?!!?!?!!!?!!?!!!?!??? I've never seen it done that hot. The PLA+ I get at do at 210, the regular PLA at 200 and silky at 205.

 

Now looking at the first layer...

Looks like it shouldn't be a surprise you're getting gaps.

 

Turning on a second bottom layer and looking at it:

It's a little bigger and the skin is being printed thicker... so does that mean?

Yes, it does. So if we disable the negative initial layer horizontal expansion (it's in the Walls section) what does the first layer look like?

Looks like the second skin layer - if you had one - would have been printed. I can still see tiny gaps, but they might get filled in as it oozes. And I am looking straight down, zoomed in, which doesn't do it any favours.

 

I tried a bunch of other settings and nothing seemed to have a significant impact.

 

Slashee's Conclusion:

• Don't bake your PLA. Turn the temp down. If it's that much hotter than it's supposed to be I'm not surprised it's getting a bit thinner and blobbier.
• You're making your initial layer a bit smaller, and it's making the lines a bit smaller.

[breakkid 0]

Hello, thanks for your fast response.

 

Quote

I'm going to assume you mean 0.4mm 🙂. 4mm would be one hell of a nozzle.

Yes, a small typo. Sorry. haha

 

Quote

Does it start when leaving the Z seam or on the way back?

 

It appears after leaving the z seam.

 

Quote

You're not at 100% flow, but that should affect things equally. I normally set my initial layer flow to 105% to aid adhesion, but if adhesion isn't a problem, don't worry about it.

 

The flow was calibrated, so the specimen can stay in tolerances I set before for my experiment.

 

Quote

Although I can't help but notice: you're printing PLA at 250°?!?!!?!!?!?!!!?!!?!!!?!??? I've never seen it done that hot. The PLA+ I get at do at 210, the regular PLA at 200 and silky at 205.

 

Yes, actually the Silk PLA by Colorfabb has a recommended printing temperature from 225-235°C. With temptowers I checked all the temperature settings possible and at this setting it still looked fine. Not even overextruding. And for my design of experiment I need to have two temperature settings (a "colder" and a "hotter" one).

 

The initial layer horizontal expansion was set to a negative value to remove the elephant foot i'm getting while printing seven of those specimen.

 

So far on the first layer with this setting all lines where connecting greatly.

 

I've tried now to print them with the value 0 and the bottom layer is connecting now not even at all.

 

 

[breakkid 0]

EDIT: Forgot to mention that with 0,3 mm line width I don't have these problems at all. By turning the angle to 90° no problem aswell. The 0,6 mm line width became better at 90° after reducing the print speed and increasing the flow of the infill which cura usually sets a few percent lower than your "general" flow rate.

[gr5 2,178]

Looks like pretty bad underextrusion.  It could be the feeder or the hot end. 

 

feeder

 

To test the feeder, fight it.  Use the MOVE command in the UM3 menu to position the filament well above the print core so it is free to move in the bowden and away from the print head.  Then move forward and at the same time fight the filament with your hand by pulling it the other way.  It should pull with 10-18 pounds of force (>5kg).  If it's less than 10 then something is wrong with the feeder.  Maybe you disassembled it and put it together wrong (a common problem).  Or maybe the tension isn't set to the middle setting.  Or maybe you printed CF filament and ruined the gear that pushes the filament through the feeder.

 

hot end

 

Make sure the front fan (not the side fans!) is spinning.  Heat the hot end to above 60C and it should start spinning - open the door to make sure it's spinning.  This is a common issue that can cause underextrusion.

 

printcore

 

Try a different print core.  The test is to use the MOVE command to extrude a lot.  Do it at the max recommended temp (or lower).  So 235C.  (250C is just too hot - don't print that hot except for weird experiments).  In fact the 250C could be the problem - some of the PLA may have caramelized onto the inner wall of the nozle hole and be restricting it.  You can scrape that out with something metal like a .35mm hypodermic needle.

 

printcore test

 

Back to the test:  use the MOVE command and you should be able to extrude some filament and get it to the print bed in ab out 8 seconds.  Less than 10 seconds you probably should fix things but 10 seconds is fine.  20 seconds?  Serious underextrusion.

 

Try a different printcore - your printer should have come with 3 printcores.  It's okay to use a BB 0.4 with PLA.

 

Try several cold pulls.  Read up about them for non-ultimaker printers and also you can do them from the menu on the UM3 under maintenance I think.

 

Consider buying a new printcore - Ultimaker considers them expendables like filament.  Since you replace them less often than every 5 spools, they are "cheaper than filament".

[breakkid 0]

Thank you for the answer gr5.

 

Quote

To test the feeder, fight it.  Use the MOVE command in the UM3 menu to position the filament well above the print core so it is free to move in the bowden and away from the print head.  Then move forward and at the same time fight the filament with your hand by pulling it the other way.  It should pull with 10-18 pounds of force (>5kg).  If it's less than 10 then something is wrong with the feeder. 

 

Unfortunately I don't know how to measure the force.

 

Quote

Maybe you disassembled it and put it together wrong (a common problem).  Or maybe the tension isn't set to the middle setting.  Or maybe you printed CF filament and ruined the gear that pushes the filament through the feeder.

 

Actually I had to diassamble it, because there were a lot of filament particles in the feeder. But you are right, I dissasembled it wrong. I just unscrewed the four screws and cleaned everything I could saw. After seeing a video of fbrc8 (I believe she is also a member of the ultimaker community), I cleaned it properly (with the POM gears). Besides that I seemed that the tension screw was in the wrong position, but from the outside it was set in the middle. It seems right now to work better. At least for the moment.

 

I'm working with the ultimaker at my university. The professor could not tell me, if CF-filaments were used with this machine. He just could ensure me that those kind of filament was not used since he started working at the university.

 

Quote

Make sure the front fan (not the side fans!) is spinning.  Heat the hot end to above 60C and it should start spinning - open the door to make sure it's spinning.  This is a common issue that can cause underextrusion.

 

The fan is working fine.

 

Quote

printcore

Unfortunately here is no other "newer" printcore. But I tried printing with the other BB 0.4. I had there the same underextrusion. The first time I made a test print here I had already some underextrusion. Whatever the problem is, it was already existing before I started using the machine.

 

I tried hot and cold pulls. the tip of the filament was clean and nicely conical.

 

I tried the printcore test with 235C and I guess everything was fine. Filament extruded with the MOVE command without any problem.

Edited January 12 by breakkid

[Slashee_the_Cow 280]

1 hour ago, breakkid said:

Quote

To test the feeder, fight it.  Use the MOVE command in the UM3 menu to position the filament well above the print core so it is free to move in the bowden and away from the print head.  Then move forward and at the same time fight the filament with your hand by pulling it the other way.  It should pull with 10-18 pounds of force (>5kg).  If it's less than 10 then something is wrong with the feeder. 

 

Unfortunately I don't know how to measure the force.

Pick up something that weighs about that much and feel how much strength that to get it off the ground. Pull on the filament with that amount of strength.

 

1 hour ago, breakkid said:

Besides that I seemed that the tension screw was in the wrong position, but from the outside it was set in the middle. It seems right now to work better. At least for the moment.

This is a big yellow flag to me (I'm not an expert on UM printers so I'll leave the red flags to @gr5). Tension is very important when you're doing something that affects the flow rate as much as going from 0.3mm to 0.6mm lines (roughly double the flow rate, but I'm not 100% on the maths for that sort of thing). It would also explain why it goes down after you leave the Z seam: it usually has to do a travel move to get to the Z seam which gives the extruder a bit of time to build up the pressure in the nozzle, but if it can't feed fast enough, that pressure will go down to a baseline level as that filament gets used.

[gr5 2,178]

2 hours ago, breakkid said:

I tried the printcore test with 235C and I guess everything was fine. Filament extruded with the MOVE command without any problem.

Could you get the filament to travel from the nozzle to the print bed in under 12 seconds?

 

Put a kink in the filament or something just below the nozzle and then spin that dial.  There is a speed of rotating the dial that maximizes the speed.  Count the seconds.  If it takes 20 seconds then there is a problem.

 

If you can pass this test then I suspect nothing is wrong with the printer and I'll have a look at your project file.

[gr5 2,178]

You have "zig zag" chosen as your bottom layer pattern.  I tried "lines" and it made the bottom layer much more like the other layers.  More lines.  I'm wondering if that's the issue.

[gr5 2,178]

Does the underextrusion get better after a few layers?

[gr5 2,178]

In your first post you are worried about the area near the seam position but that just exacerbates the underextrusion.  I think the underextrusion has to get fixed first.  Maybe.

[gr5 2,178]

I ran the resulting gcode through the website https://www.gcodeanalyser.com/ and set filament diameter to 2.85 and it simulates how wide the traces are.

 

It also shows gaps but only for the first layer.  If you switch to "lines" I think it will fix that first layer.  Green image below is first layer.  Red is second layer.

 

 

 

[gr5 2,178]

Also you have "line width" set to 0.6.  You should expect underextrusion on a 0.4mm nozzle with lines at 0.6mm.  Typically.  The area of a 0.4mm nozzle is about half that of a 0.6mm nozzle so you are putting out double the volume of filament than normal.  You'd expect that to require double the nozzle pressure but because the filament has to slip into that 0.2mm crack on either side of the nozzle, the required pressure is probably closer to 3X.

 

Printing 0.6mm out of a 0.4mm nozzle works great when printing a single walled cup or vase.  But for solid infill expect some underextrusion.  So recommendation:

 

Change zigzag to lines for top/bottom layers.

Change line width from 0.6 to 0.4

 

Print it again, see what you think.  Don't worry about the slight increased underextrusion at layer change as it should go away.

[breakkid 0]

The underextrusion is back.

 

Quote

Could you get the filament to travel from the nozzle to the print bed in under 12 seconds?

 

Put a kink in the filament or something just below the nozzle and then spin that dial.  There is a speed of rotating the dial that maximizes the speed.  Count the seconds.  If it takes 20 seconds then there is a problem.

 

Okay, I tried this:

 

Filament is loaded in, i heated material 1 up to 235C, after it reached the temperature, I went to the menu MOVE and started to spin the wheel. The moment I started spinning i started a stop watch. After the filament was in the middle of the way down, 33 sec had passed. I stopped.

 

Quote

Does the underextrusion get better after a few layers?

 

No. The first layer looks good, after that the problem starts to appear. At layer 3 or 4 maybe it stays in the same quality.

 

Quote

You have "zig zag" chosen as your bottom layer pattern.  I tried "lines" and it made the bottom layer much more like the other layers.  More lines.  I'm wondering if that's the issue.

 

Unfortunately not. I've tried this already. The only thing it did was just printing over the holes and making them dissapear. That's not the purpose of the experiment. Parts with holes in it will deliver corrupt data.

[gr5 2,178]

To do the 12 second test, do not stop spinning the dial.  You have to spin it quite fast, but not so fast that it starts slowing down again.

[breakkid 0]

Quote

Change zigzag to lines for top/bottom layers.

Change line width from 0.6 to 0.4

Right now I#m printing with 0,4 mm line width and I'm still underextruding. With 0° line directions more as with 90°.

[gr5 2,178]

33 seconds for half way is horrible.  Really horrible.  That's the problem: the printer.

 

Try the "fight the extruder" test again.  It takes pretty much all my strength to fight the filament and have it slip.  If you can't get it to slip with all your strength it's probably not the feeder.

 

It could be the filament I suppose.  I'm a little wary of a new filament I've never tried.  But EVERY different type of PLA I've ever tried including HT PLA filled PLA and many others have all had about the same printing viscosity.

 

 

[gr5 2,178]

And like Slashee-the-cow says - try lifting a weight of some sort to gauge the force.  I think the UM3 is around 10 pounds maybe?

[gr5 2,178]

I'd go back to the AA 0.4.  The BB cores tend to (partially) clog easily as PVA caramelizes in just a few minutes (20?) of sitting at printing temps and it raises the BB core temp everytime you do active-leveling.  And there could be a tiny drop of PVA in there.

 

AA cores also tend to get a caramelized coating on the inside of the nozzle. I scrape that off (while nozzle is hot) when the 12 second test fails.  I use a hypodermic.  It's hard to get a nice sharp piece of metal that fits in a 0.4mm hole.

 

If you go into the maintenance menu, you can get it to tell you how many kilometers of filament you have printed so far.  I'm curious what is reported for each of the cores you have access to.  That UM3 came with two AA 0.4 cores and one BB 0.4.  Do you still have all 3 or did one get lost?

[breakkid 0]

Quote

33 seconds for half way is horrible.  Really horrible.  That's the problem: the printer.

 

It was cooling down while I was spinning constantly in regular speed.

 

Quote

"fight the extruder"

 

Sorry, but I don't have any feeling for this test. I tried it. I guess it feels right. I really don't know.

 

Quote

AA cores also tend to get a caramelized coating on the inside of the nozzle. I scrape that off (while nozzle is hot) when the 12 second test fails.  I use a hypodermic.  It's hard to get a nice sharp piece of metal that fits in a 0.4mm hole.

 

Unfortunately I don't have any needles here. This is not my printer. I'm at the university right now, because my printer at home is broken.

 

Quote

If you go into the maintenance menu, you can get it to tell you how many kilometers of filament you have printed so far.  I'm curious what is reported for each of the cores you have access to.  That UM3 came with two AA 0.4 cores and one BB 0.4.  Do you still have all 3 or did one get lost?

 

Printcore 1: About 246 meters. Hot since 15 days about. Printcore 2 (BB) was barely used. And there is also the same underextrusion.

The Professor know nothing about another printcore being around here.

[breakkid 0]

Quote

Try the "fight the extruder" test again.  It takes pretty much all my strength to fight the filament and have it slip.  If you can't get it to slip with all your strength it's probably not the feeder.

 

I was not able to let it slip. Let's say it that way.

 

Quote

It could be the filament I suppose.  I'm a little wary of a new filament I've never tried.  But EVERY different type of PLA I've ever tried including HT PLA filled PLA and many others have all had about the same printing viscosity.

 

I have the same issues with other colorfabb filaments. PLA/PHA and allPHA.

[gr5 2,178]

Can you make a 10 second video of the 12 second test?  I'm not sure I understand what you mean.  Let me explain again.

 

Get it extruding with MOVE.  Experiment with the fastest you can spin without it slowing down extrusion (if you spin too fast it slips and you get reduced extrusion).  Then put a kink in the filament so you can see it's movement.  For example pull the solid (just extruded) filament sideways at 90 degrees.  Maybe hold it like that at 90 degrees.  Or pinch off/break the filament.  Then maybe start timer and then start spinning the MOVE COMMAND at the correct spinning speed for the maximum extrusion speed possible (if you go too fast then extrusion slows down).  Stop timer when the kink gets to the build plate (which should be in the lowered position of course).  Should be under 12 seconds.  Typically more like 8 seconds.  60 seconds is a problem.

 

246 meters is brand new.  I've seen cores with 2000 meters (2km) still working just fine.  I've seen cores at 6km finally needing refurbishing.

[gr5 2,178]

50 minutes ago, breakkid said:

Unfortunately I don't have any needles here.

I'm not surprised.  It's a specialty tool.  Some people manage to find a metal wire brush and cut off a bristle.  Your core is so new, I'm not convinced it's the core yet.  But it's a common problem.  It could be a defective core.  Not likely but often we only hear from the unlucky people on this forum.  The core has some nylon inside the heat sink and if it isn't cut to the exact correct length it can cause different issues that result in underextrusion.

 

It seems unlikely both cores would have the same issue.  Someone, somewhere knows where the 3rd printcore is.  There's other items.  A few tools, some grease, a gluestick, some oil maybe.  I forget what else comes with the printer.  Someone knows where that stuff is.  The person who wrote the purchase requisition for the printer probably is the person who unboxed it and knows where that 3rd core is.

[gr5 2,178]

47 minutes ago, breakkid said:

PLA/PHA a

good stuff.  Probably less viscous than regular PLA by a small amount.  So it flows nicely.  Less underextrusion.  But your underextrusion seems a bit extreme.  Assuming you did the 12 second test correctly.

[breakkid 0]

Made a video. There is no way to upload it here...

 

So here is a wetransfer-link. It took the filament 32 seconds to get to the bottom of the beatbed.

 

https://we.tl/t-loI1dPRZHi

 

THere are two specimen I printed again. I changed to 100% flow. THere are still gaps. First one is printed with zig zag, second with lines.

 

Furthermore there is a pic from the feeder and the tension. Its set to middle, but I wonder if this knob is positioned correctly.

Edited January 12 by breakkid