Inter-layer adhesion issue

[dramsey]

For lack of a better term. Anyway, after a restful three days with no printing, this morning I dragged the printer back out, double checked the Z clearance and bed level, and started testing.

Basics: 205 degree extruder, 60 degree bed; 3600mm/min print speed, 0.20mm layer height. Extrusion multiplier 0.9, extrusion width set to manual at 0.40mm.

This is a key fob with three solid initial layers. The first layer prints beautifully and the bottom of the print is very smooth, with demarcations between the extruded lines invisible for most of the area.

The second layer prints successfully, but has a visually odd "ripple" or "ringing" pattern. It looks like the X-Y equivalent of photos I've seen of "Z ringing".

The third layer doesn't print well at all. The visual ringing is very pronounced and on longer runs the filament pulls up when the head reverses its travel. It looks like this:



There are a couple of interesting things here: one, the problem only occurs on longer runs. Two, the problem only occurs in one direction.

If we look really closely, it almost looks as if the print head is "bouncing" across the second layer:



This is the cause of the "ringing" pattern. I would guess there's virtually no contact between the top layer and the layer below it in the space between the red arrows, which would explain the "peeling" problem, and why I can, if I let the print run to completion, dig a thumbnail into the edge of the fob, and easily peel it in half lengthwise into an upper layer and a lower layer.

I have run test prints with Z offsets varying from 0.00 to -0.20, and noted only minor changes in this behavior, although as the offset approaches 0, the extrusion lines become very evident. -0.15mm seems to be the sweet spot, at least as far as bottom layer quality goes. I have run tests printing in many different locations on the bed, center to corners. Again, I saw only very minor variations in the result.

And although nobody's ever mentioned print quality issues with the NEMA dampers, I removed them and printed again. Aside from a dramatic noise increase, there was no difference. Damn! Really thought that might be it...resonances in the rubber or something.

So that's where I am now. At this point I'm positing some sort of subtle mechanical issue-- slop in the extruder mounting or something, although it all feels solid. If I feel perky I may trying printing with a 0.15mm

I await informed suggestions as to how to address this. Or, heck, any suggestions!

[Slipshine]

I know this is not going to go over well.

When you do your side wall calculation on the configuration cube have it set to .5mm but adjust the multiplier so it actually measures .55mm.

I have been doing this since the latest release and it works flawless for me.

[Jules]

Post the Factory file so we can take a look at it. (File > Save Factory File) then zip it up before you upload it here.

[dramsey]

And here you go:

[Jules]

Okay, here's what I found in your file:

You were only printing with a 90% first layer height, which means that the first layer was either being under-extruded, or that the Z-Offset of -0.15 mm is bringing the bed too close for a 100% first layer. (It would squish out the sides and make a little lip.)

If you're going to use Z-Offsets to adjust the gap, you leave the first layer height at (or near) 100%. (I tend to use 99%, because of a bug in S3D.) You can't use both to adjust the gap, because you'll constantly be making adjustments back and forth between the layer height and the Z-Offset. Of the two methods, i find it easier to just leave the first layer height at around 100% and adjust the Z-Offset only, by a few .01 mms, plus or minus.

You also had too low of a speed setting on both your retraction rate, and for the X/Y Axis travel movements. That was what was causing the tiny threads that you had so much trouble with.

You had chosen a Rectilinear Infill pattern. That particular pattern only prints every other layer. I switched that to a Fast Honeycomb, which provides better support for the solid layer above it. (You can change that if you don't like the honeycomb effect.)

So what I did in the file below, I changed all the applicable speed settings, infill settings, and massaged your temps a little. I also changed the 1st layer height to 99%. When you next print this, you might get a slight lip or over-extrusion with your first layer, because I did not change the Z-Offset. (No way of knowing how much to change it until you print the first couple of layers one more time.)

Only If you see a small lip on the next print, you can take the Z-Offset number down to -0.13 mm and try again. If there's still a lip after that print, take it down to -0.11 mm.

If there is no lip, and you are still getting the wavy top surface that doesn't stick to the next layer, change the Z-Offset number to -0.18 mm and try again.

If it works okay for you, remember to save the Profile and export a copy of the profile to your desktop so you can go back to it if you start fiddling with the settings again down the road. (It's too easy to over-write these things.)

single_fob_revised.zip

(34.63 KiB) Downloaded 509 times

[dramsey]

Jules, I really appreciate all your help. I think the major problem was the 90% first layer height. I don't recall setting that at any time, but moving it to 99% sure seems to have (mostly) fixed the layer adhesion problem. Sadly, though, finished prints still look like hammered s**t:



The lower fob is one of the hundred or so printed back in the Days When Printing Worked. The upper fob is one printed from your new file. Although it's not evident in the photo, the lower fob has a flat matte finish, while the upper fob is kinda shiny. Also, at about 2:30, you can see where the upper solid layers-- the base the "Ford" logo is resting on-- didn't adhere well to the sides.

I'd noticed this latter problem before on this particular model with honeycomb infill, so I switched back to my standard rectlinear infill I've been using successfully. Yeah, that didn't work AT ALL.



I have reached the end of my endurance, I'm afraid. I will contact Makergear directly to see if I can pay them to figure out what I've f**ked up if I ship the printer back to them. If not, I'll post it for sale here, at a really great price.

[Jules]

Do one more thing before you quit....change the Z-Offset to -0.18 and try one more time.

You're almost there.

The lower one is not shiny because you were printing it too hot and damaging the plastic. PLA if printed at the correct temp is slightly shiny. (PETG even more so.)

If you want to make it matte again, increase the extruder temp back up to 215+.

[dramsey]

Do one more thing before you quit....change the Z-Offset to -0.18 and try one more time.

I had a small lip on the first print (the one in the photo), so I raised the Z offset to -0.13, which is what I used to try and print the rectilinear fill thing. I think dropping to -0.18 would potentially be bad, but I could try.

You're almost there.

It's an asymptotic progression. I can approach perfection, but never get it.

The lower one is not shiny because you were printing it too hot and damaging the plastic. PLA if printed at the correct temp is slightly shiny. (PETG even more so.) If you want to make it matte again, increase the extruder temp back up to 215+.

Actually, all the good fobs were printed at 205. Really. I swear. I've been using 205/60 since about the third day I had the printer.

Wife suggested I try printing something else. Currently, Utah Teapot in freshly-opened Makergear red filament. And, sigh, looks like the same problems.

EDIT: I just had a thought: when I removed the hot end to unscrew the nozzle for cleaning, I removed the thermistor. Not sure why I did, but I did. Perhaps I managed to somehow damage it, or there was a problem when I put it back in the hot end, and it's not reporting the correct temperature...

UPDATE: Utah Teapot print failed, mainly due to the fact that the printhead keeps hitting the print. After the model is a couple of mm high, you can hear it hit almost every time it does a fast traverse. It knocked the little knob completely off the teapot lid. Yes, I've printed this exact STL before several times with no problem.

[Jules]

Okay, final thing - go back to -0.14 mm for your Z-Offset. (Now it's too close.)

It's unlikely that you damaged the thermistor by removing it, they are designed to come out. Don't know what to tell you - the tiny threads are not print killers, (you can just brush those off with an old toothbrush), but those unconnected threads that I see in right curve of the top image shouldn't be happening. You should not be seeing those on the teapot.

Hold on - i wonder if the fact that the fob is sunk below the bed isn't having an effect on S3D - it can get weird with layer attachments sometimes.

Let me fix that on the file and see if it changes anything on the print. Don't do anything, and don't change the filament again - every time you change something you have to fine-tune your settings. Leave the filament on that you were using for the fobs and stand by.....


Well, that was fun....hate messing with mesh files.

Ford key fob(8).zip

(238.3 KiB) Downloaded 499 times

Okay, I just updated that file - I figured it out - your infill was too low for that honeycomb pattern (you were right about that.) Version 8 above has the fast wiggle with a 50% infill. That's going to take care of the threads not catching on that curved honeycomb surface.

If version 8 doesn't work....I give up. (Man....have I got a headache.)

[Jules]

Just in case you grabbed the file before i switched it out....you need to be testing version 8. (I fixed the infill problem.)