Adventures into Volcano

Discussion in 'General' started by AndyVirus, May 16, 2017.

Tags:
  1. AndyVirus

    AndyVirus Well-Known Member

    Joined:
    Apr 10, 2016
    Messages:
    233
    Likes Received:
    50
    I hope this is ok, I am posting what i have done so far and will use this as a mental dumping ground as i progress.
    I have switched both my Direct hotends on my IDEX system to Titan Aero and Volcano heater-blocks and 0.4mm Hardened Nozzles.

    I am printing on a Clever3D PEI coated bed with either PLA or PETG mostly (I ran out of edge).

    My first print after calibrating esteps was a mess. Bridging failed and layer separation. Ok something is wrong.

    2nd print: Lowered my cooling fan settings and instantly a better result. Looks like i was cooling the extruded filament too quickly before it had a chance to bond properly.

    3rd print: increased speed and disabled "increase fan speed for layers below 45.0 sec"... A bit better again but still the walls are terrible (zits, blobs, bleh)

    4th print: Changed extruder width from Auto to Manual and set 0.38mm (auto was setting it to 0.48mm), also changed retraction speed from 70mmps to 38mmps...Marked improvement! Looks like the extrusion width is the key.

    5th print: Reduced extruder width again, this time to 0.35mm... Worse. Looks like at layer change the start of the extrusion is too much (blob), will have to play with retraction? Current retraction is at 1.5mm. I wonder if I need less or more with a Volcano.

    6th print: Increased extruder width to 0.39mm, increased retraction to 1.80mm. Increased cooling fan strategy to cool more as while the surfaces were nice the shape of the part was warped.... Huge improvement but still on layer change the blob occurs.

    7th print: 0.38mm width, 2.0mm retraction. just nasty but less strings.

    8th print: 0.40 width, 2.2mm retraction (hope thats ok, im assuming it is due to volcano being longer). Printed but came off the bed. Walls looked better and almost no blobs. I am sure i have quite a good base now, i think my cooling fan is warping the prints now. Will now concentrate on that but 0.4mm nozel with 0.4mm with and 2.2mm retraction seems to be a good place to start with direct volcano it seems.

    I also tried printing PETG with the left Nozzel (all the above was orange PLA) and it printed ok but shed loads of curling. I think this is due to my cooling fan interfering with the air flow from the Aero fan... ill deal with that later. One thing i did notice was the cheap Real Filament PETG is quite a bit stronger printed with Volcano. I kind of expected that but nice to confirm it. Compared to EDGE, Real Filament PETG is fragile, but for £17 per KG, it makes a good prototyping filament.

    (Below Print 2 - 7 from left to right)
    IMG_1526.JPG ]
     
    #1 AndyVirus, May 16, 2017
    Last edited: May 16, 2017
    Alex Stevenson likes this.
  2. AndyVirus

    AndyVirus Well-Known Member

    Joined:
    Apr 10, 2016
    Messages:
    233
    Likes Received:
    50
    OK so trying PETG again on the left side. With 0.40 extrusion width, 2.2mm retraction and retraction speed set to 38mmps i was able to re-print my left hand side fan shroud but it had issues on overhangs greater than 45 degrees. this i think is due to the fan shroud as i modified it to be thiner which i think has messed up the air flow. now I have the correct hight for the air flow, i am trying overhang tests and bridging tests to get the cooling fan settings perfect before further tweaking. PETG is oozey as hell but less expansion. Also printing at 240c for PETG as at lower temps it is not flowing as well and looks dull, almost matte when cooled. 240c is glossy and flows well (and at rates of 10+mmps extrusion speed). 0.40mm extrusion width was the lowest i could go if i wanted it to stick to the PEI bed.
     
    #2 AndyVirus, May 17, 2017
    Last edited: May 17, 2017
  3. AndyVirus

    AndyVirus Well-Known Member

    Joined:
    Apr 10, 2016
    Messages:
    233
    Likes Received:
    50
    Ok so re-printed my fanshroud in petg. The left is the Volcano print ant the right is from a v6. The left was printed at 250c while the right was at 220c hence the difference in shine.

    The ledt printed in avout half the time of the right at 0.2mm layers.

    Not sure what some of the imperfections are about though. Thoughts?

    IMG_1529.JPG
     
  4. AndyVirus

    AndyVirus Well-Known Member

    Joined:
    Apr 10, 2016
    Messages:
    233
    Likes Received:
    50
    OK, so I have found something counter intuitive. I have found that cooling PETG actually has negative affects on print quality (strings and blobs) and reliability (clogging, layer adhesion).

    The greater the cooling fan speed the worse the print gets. I am printing 100% infill.

    So if i print with no fan, all good.
    Raise the fan speed to 30% and all good still but now on retraction I get strings between spots.
    Anything over that and the extrusion can and often does fail to adhere to the last layer and bridging is an utter no go (or very poor).

    It occured to me that every time i pick PETG of my nozel I was waiting for it to cool a bit as it comes off in clumps easier and a bit like rubber (until it has cooled). This got me thinking. If I cool an extrusion then retract, is that not the same thing? I think it is.

    The difference between the 2 above other than left is volcano and the right is from a v6 is that the fan shroud nozel at the back had melted so no air flow! I will do some more tests but i think the key to stringless PETG and maybe even EDGE prints is to limit cooling to an absolute minimum. Not 100% convinced myself yet but this is what it looks like.
     
  5. dc42

    dc42 Well-Known Member

    Joined:
    Aug 16, 2016
    Messages:
    526
    Likes Received:
    221
    Not specific to Volcano, but continuous cooling after the first layer is usually only needed when printing PLA. Filaments with a higher extrusion temperature usually need cooling only when bridging, if at all, and excessive cooling harms adhesion between layers.

    You should not normally print plastics with 100% infill except for a few bottom and top layers.
     
    AndyVirus likes this.
  6. AndyVirus

    AndyVirus Well-Known Member

    Joined:
    Apr 10, 2016
    Messages:
    233
    Likes Received:
    50
    Thanks DC42. I have been printing EDGE mostly with mixed results with cooling. Very forgiving filament as most my prints were either good or acceptable. I am now trying the cheaper PET based filaments for a bit while i figure out Volcano and slicer settings.

    So far I am now getting better results... Picture of a wiper for volcano on my bigbox in black PETG. Still 100% infill but i am back to 15% now. PETG is very opaque so was using 100% to avoid seeing the infill pattern.

    IMG_1534.JPG

    Not perfect but alot better. This was at 120mm per second in S3D though due to its size the speed is reduced to allow for cooling before the next layer is layed down. "Reduce speed for layers below 15 seconds" down to "35%". According to S3D most of it was to be printed at 60 - 80mmps. Its 30mm high by 40mm long by 20mm deep. Will try pressure reduction on direction change once i am happy with the walls.
     
  7. AndyVirus

    AndyVirus Well-Known Member

    Joined:
    Apr 10, 2016
    Messages:
    233
    Likes Received:
    50
    Printed lots in Colourfab_HT with .4mm nozzel at layer hight of 0.32mm with great results. The think that has been bugging me is the first few layers always look like crap. I think i have figured out this is backlash. One of my antibacklash couplers had lost its spring/fell down so not pulling on the lead screw. I will have to confrim if this fixes the first few layers issue being squished more than i expect but im pretty sure thats what it is. HT v Generic PETG... worlds apart in quality.
     
  8. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,561
    Likes Received:
    483
    Why would only the first few layers be crap? The most likely reason I can think of for that is that nozzle isn't at exactly the right height. So you're printing 0.32mm layer height, but your nozzle isn't at 0.32. If it were at 0.25, the result would be that the first layer would be squashed. Might have a benefit of connecting well to the printbed, but causes "elephants foot" and leads to the infill being squashed together, giving ridges on the surface. These ridges then affect the next layer; it's not laying on a flat surface, but a mountain range. After a couple of layers, this tends to fade away. Or maybe you're printing at 0.4 height, in which case, it's more likely that you see space between the infill lines. The second layer is then laid over a set of small chasms and again you get ironed out after a few layers.

    In S3D you can control the First Layer Height. Maybe you're oversquashing it? Personally, I find that laying that first layer at 100% height (so for you that would be having the nozzle at 0.32) is the best way.

    How you control that is going to a combination of machine control and calibration. I expect the IR sensor to get me started at the right height everytime, but I don't feel it does this to perfection. So, for me, every time the print starts, I watch it like a hawk as it prints my first calibration square - a 10mm square, 1 layer thick model - and this gives me enough time to make adjustments to the Z height via Z Babystepping before the print starts in full. It also means I don't tend to bother with a skirt (who's job is the same really - to get the juices in full flow (unless you're skirting for extra grip that is).
     
  9. AndyVirus

    AndyVirus Well-Known Member

    Joined:
    Apr 10, 2016
    Messages:
    233
    Likes Received:
    50
    I thought that too, but tweaked and tweaked but no change. Actually its not the first few layers, its the first few layers after the bottom solid layers that look terrible then a few after that they are fine again. I thought it was infill overlap causing the issue so i reduced it and added extra perimeters but no difference. take a look at the above image. I thouht it was a speed thing but slow or fast it happens... onlu on the first layers after solid infil layers though. Very odd.

    I am thinking it might be the coupler not pulling down on the lead screw bed nut causing a little play in the thread... I just built a corexy printer with a single lead screw in the center back and noticed the play in thread and nut on direction change in Z. Possibly barking up the wrong tree on the bigbox but then noticed that one coupler was not correct. Need to test.
     
  10. Old_Tafr

    Old_Tafr Well-Known Member

    Joined:
    Mar 7, 2016
    Messages:
    731
    Likes Received:
    93
    E3D went to extreme lengths to get the Z rods on the BB correct, including having a few microns removed from the rods so that the bearings didn't stick/drag, the info is I think on one of the updates on the BB Kickstarter website. If I can find the link I will post it here.
     
  11. AndyVirus

    AndyVirus Well-Known Member

    Joined:
    Apr 10, 2016
    Messages:
    233
    Likes Received:
    50
    OK still struggling. I think my issues stem from pressure on direction change on on infill.

    So on solid infils (bottom layers) the first layer comes put perfect or there abouts but the second layer goes instantly to s***. See picture.
    Basically I am over extruding at the edges on layer 2 and 3 (3 solid layers).
    I am using S3D and have tried the "Coast and End" option but that does not affect infill as far as i can see. I can see it coasts at the end of perimeters but not infill.

    Any ideas? Is this a pressure thing or am i barking up the wrong tree? Layer hight of first layer is perfect (2.4mm) 100% first layer.
    I have also set pressure advance coefficient on DuetWifi to M572 D0 S0.05 (tried 0.1 to 0.5) and that seems to make no difference either. Really struggling with this one.
     

    Attached Files:

  12. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,561
    Likes Received:
    483
    What do you mean by 2.4mm?
    What nozzle are you using? The single outline looks thick, but the indivudal lines look very close together.
    What layer height is that?
    You seem to have some peeling there even on layer 1? Is that material/heated bed temp/soak length/mesh bed levelling imperfection?
     
  13. AndyVirus

    AndyVirus Well-Known Member

    Joined:
    Apr 10, 2016
    Messages:
    233
    Likes Received:
    50
    sorry meant 0.2400 layer hight, 0.4mm hardended steel volcano nozel, yes the lines do seem thick, hight of layer 1 seem perfect, no peeling bed flat or as flatter than i have ever been able to get it, get the same with petg or matx, bed temp 130 to get matx to stick, not sure what soak lenth is.
    turned layer 2 down to the same soeed as layer 1 and its just as bad.
    What i am trying to do now is get the extrusion width to match s3d. the hight is spot on for layer 1 but width i am guessing not so much. i want to tell s3d what width it is mot have it attempt to make it that width... not sure how to achieve this. will try moving the hight again but then ill be off on hight an good on width... same issue or new one... any ideas how to tell s3d to stop pushing out so much material when you have the hight perfect? esteps spot on, exactly 100mm being extruded freely to calibrate, on the above i have extrusion multiplier of 0.80 so with out lowering that even more, any ideas how to get tell s3d what the width is so that there is no overlap?
     
    #13 AndyVirus, Sep 17, 2017
    Last edited: Sep 17, 2017
  14. AndyVirus

    AndyVirus Well-Known Member

    Joined:
    Apr 10, 2016
    Messages:
    233
    Likes Received:
    50
    OK, so i have the hight and width perfect by playing with extrusion multiplier. So hight and width of first later looks fine. Mesured with digital calipers seems spot on too.

    I am now back to thinking that it is a z motion thing where on layer 2 it does not move down quite enough and causes over extrusion. I dont know how this has become a thing but it is what it seems. After layer 5 or so it goes back to being perfect.

    IMG_1824 first layer
    IMG_1826 second layer, over extrude instantly (wtf)
    IMG_1828 All fine again after a few layers
    IMG_1831 Almost done, nice perimeters
    IMG_1833 Top layers, no over extrusion (may slightly under extrusion which i can deal with)
    IMG_1835 Done.

    It does seem that Z does not move as much for the first 2-4 layers then is ok. I am possibly wrong but i am getting to the point of taking an axe to this

    Update:

    40mm cube
    X=39.70
    Y=39.64
    Z=39.00
     

    Attached Files:

    #14 AndyVirus, Sep 17, 2017
    Last edited: Sep 17, 2017
  15. AndyVirus

    AndyVirus Well-Known Member

    Joined:
    Apr 10, 2016
    Messages:
    233
    Likes Received:
    50
    OK Twigged it. Only took 2 days solid of testing everything! Grrrr.

    So i was right, the first 4 layers were not moving! Layer 4 then jumped to Z1 and from there each increment moves as expected.
    So i though this must be a software limit of some sort...

    Trawling though my Config.g file I found that minima for Z movement was for some damn reason set to 0.98. No idea why. So set that to 0, re measured the distance between the nozzle and the bed using a feeler gauge, adjusted the z probe offsets accordingly and bobs your uncle, no more over extruding on layer 2 and 3!

    Back to less annoying issues...
     
    Spoon Unit likes this.
  16. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,561
    Likes Received:
    483
    Ha. That's got to be such a relief.
     
    AndyVirus likes this.
  17. AndyVirus

    AndyVirus Well-Known Member

    Joined:
    Apr 10, 2016
    Messages:
    233
    Likes Received:
    50
    OK, so this weekend been playing with Print speeds. So If i tell S3D to print at 200mmps (or anything above 150mmps really) it does not do so and that will be down to the limits imposed in the firmware. So I raised the speed limit i had from 9000mm per Min to 12000mm per min then 15000mm per min. Much faster but now when printing a 40x40 cube i have some weirdness on the 2 back corners of the cube (in the photo).
    I am trying to understand why this is occurring. The 2 hypotheses I can come up with is:
    1) Over extruding in those corners due to that is where the infill wiggle also makes contact with the perimeter
    2) As i am printing way faster now, the wall in that corner is cooling faster than id did when printing slower and warping making the next layer print in air and so forth.

    I have gone from a 14 second layer time to a 7 second layer time.

    The front 2 corners are pretty much identical from first print to last and the walls front and sides (except the deformed corners) are pretty much the same quality, the only oddness is the rear 2 corners, one worse than the other (which is also the layer start corner).

    Also played with Jerk and acceleration settings in order to get better layer speeds, jerk i changed marginally from 480 to 600 on x and y, acceleration changed from 400 to 1000 then 2000 (cube 1 2 and 3 respectively (left to right)).

    So acceleration seems to have caused this issue on the back 2 corners only, front 2, marginal difference in quality, maybe a tiny bit more ringing but will deal with that later as it does not bother me at this point as the exercise is to see how fast i can go with out loosing too much quality.

    The weird thing is on the last 2 cubes (right 2) there is an element in both of it correcting itself about half way on the right rear corner as you look at the photo and top left on second from last cube ... if it was as simple as starting points then why would it not be consistent?

    Z dimensional accuracy is very close, no more than 0.1mm between first and last print.

    Material is MatX so ASA of some sort, all printed with no cooling fan at all.
     

    Attached Files:

  18. MTJC

    MTJC Member

    Joined:
    Jun 21, 2018
    Messages:
    9
    Likes Received:
    0
    Why would you print the volcano version 30C higher than the v6 version?
     
  19. AndyVirus

    AndyVirus Well-Known Member

    Joined:
    Apr 10, 2016
    Messages:
    233
    Likes Received:
    50
    I was testing temps, speeds, extrusion rates etc. I got better results at higher temps specifically with petg. I have not visited this subject in a while (kids), but i think i extensively covered my methods above that statement. From memory it was probably something to do with my x an y movement speeds (as i was seeing how fast i could print also).
     

Share This Page