Help with my print. Need advice how to fix. new user

[gman 0]

Ok, now have nice prints! thanks. Height also seems closer now to real one. Problem i have is im printing PLA, and the corners of some layers are curling up. therefore after some time, the next layer will go overboard and create lines on outside finish. How to solve this? PLA sticking well to the bed.

 

[illuminarti 18]

Try increasing the minimum layer time so that there's more time for one layer to cool before the next gets added. How much depends on your layer height - somewhere between 3 seconds for sub-0.1mm layers to 7s for 0.2mm layers.

 

[illuminarti 18]

What did you do in the end to get better prints?

 

[gr5 2,172]

and the corners of some layers are curling up

 

Is this happening on mildly overhanging corners? I mean - where this is happening - is each successive layer supposed to be farther from the center of the part? Or is it happening on vertical corners and shrinking corners as well?

 

[gman 0]

lubricated the z-axis(i think this is whats causing the inaccurate z movement), tweaked settings based on your suggestions. the z-axis was the crucial one.

Usually on mildly overhanging corners. but sometimes also happening on vertical corners. What is cause for both?

 

[gr5 2,172]

I haven't figure it out. The lifting is usually not a problem but it is scary. I think it's going to knock the part off the bed but it rarely does. Unless I forget to put isopropyl down. But it only does it on corners that lean out. Straight up or shrinking layers are always fine. I guess it's a shrinkage issue where higher layers are pulling back on the lower layers. Each layer pulls on the next. More fan should help, or print slower (give lower layer more time to cool before next layer gets placed on it. Or heated chamber might help. Or low expansion PLA might help. But usually I just try not to think about it and it works out okay. Maybe reduce the infill a little if possible.

 

[gman 0]

What is the effect of FAN DUCT? Does this improve the print? i see people using cones/donut shaped ducts, whats the concept behind it? is there a difference or is the stock fan enough?

 

[gr5 2,172]

Daid claims all those fancy ducts don't work very well because the stock UM fan isn't designed for restricted air flow so you get much less air flow than with no duct at all. Unless you get a different type of fan designed for this. Also adding any weight at all to your print head will increasing "ringing" lines although those fan ducts tend to be extremely light weight and not affect anything. They also have a tendency to melt/droop.

 

[gman 0]

how fast can you guys print? is 50mm/s considered slow?

 

[gman 0]

Z position Inches(calipers) millimeters Difference in mm

0 4.421 112.2934 3.5687

4 4.2805 108.7247 3.9497

8 4.125 104.775 4.064

12 3.965 100.711 4.064

16 3.805 96.647 4.0132

20 3.647 92.6338 3.81

24 3.497 88.8238 3.937

28 3.342 84.8868 4.0386

32 3.183 80.8482 3.8608

36 3.031 76.9874 3.8608

40 2.879 73.1266 3.9624

44 2.723 69.1642 3.8354

48 2.572 65.3288 4.064

52 2.412 61.2648 3.937

56 2.257 57.3278 4.0386

60 2.098 53.2892 3.8862

64 1.945 49.403 4.1402

68 1.782 45.2628 4.2926

72 1.613 40.9702 4.0005

76 1.4555 36.9697 2.0447

78 1.375 34.925

Average:

3.964405

E steps Z :

3413.333

After, lubricating Z-axis sometimes extruder will hit the print itselft and dislodge it. I dont know if this is because PLA curls up and becomes hard so extruder hits or my Z-axis steps is off.

I have measured as best i can with calipers and these are the values, can anyone tell me if my Z-steps is off?

 

[illuminarti 18]

I don't understand your 'E-steps Z' value.... I presume that's your steps per mm for the Z axis? If you have an Ultimaker, then, by definition, the steps per Z should be 533.3333. The way this is calculated is:

1) one rotation of z-screw is 3mm change in platform height

2) one rotation of motor is 200 full steps

3) motor is set to 1/8 microstepping

Therefore... the number of effective steps per revolution is 200 x 8 = 1600 to move 3mm.

So the number of steps per mm is 1600 steps / 3mm = 533.33.

PLA certainly can curl slightly, especially if its not very well stuck to the bed, or if you're over-extruding (due to a bad steps-per-E setting, or wrong filament diameter). You can get better bed adhesion by wiping the blue tape with isopropyl alcohol; but as the filament extrudes and the head moves, it does pull slightly on the print, so if they are too tall relative to the contact area with the bed, there's a good chance you'll eventually shake it loose. The secret there is to either re-orient the print on the bed, if you can, or print with a brim, so that the contact area becomes larger.

 

[gr5 2,172]

50mm/sec is a good speed for .2mm layer height. I don't like to go over 75mm for .2mm layer height. For .1mm layer height 100mm/sec is a good speed.

 

[gman 0]

oh sorry, miscalculated the z but i think no problem since my prints coming out great lately.

I however tried printing in with .1 layer height and got quite poor results. There are holes again...

This is due to?

 

[gr5 2,172]

.1 is definitely more difficult and I keep telling people that although it is possible to get better results you often get worse results.

Anyway, please post a pic.

 

[gman 0]

Here are some photos. Started out well...after a while the infills got so bad that it clumps up. had to pause and intervene. Then it repaired and went sooo well, got my hopes up then the head happened. Holes visible at the top.

here are my settings. they seem sound...

.1 layer height

print temp 220

speed 70

1.73diameter

.4 layer thickness

fill density 10%

with raft

infill speed 70

min layer time 5

 

[illuminarti 18]

You're printing just a single perimeter loop and almost no infill. Once you get to the flat part of the head, you're going to be printing mostly in thin air, and the layers are going to collapse into the print. If you look at the gcode view in Cura you'll probably be able to see that the perimeter loops are just hanging in mid air, rather than building on the layers underneath.

10% infil is so thin it probably isn't doing you much good - you could probably print it with zero infill - but you'll need to have 2 or 3 loops - 0.8 or 1.2mm shell thickness - to get decent results on the more horizontal surfaces like the top of the head.

The infill failure may be due to the layers not aligning with one another due to the way Cura currently positions infill. Another reason to simply avoid printing infill if you don't have to.

 

[gman 0]

sorry forgot to specify. it was printed with .8 shell thickness actually. .4 was the top/bottom layer thickness. in this case 10% enough?

 

[illuminarti 18]

it's still going to be a bit marginal I think. 10% is pretty much no infill - it leaves huge gaps to try and bridge. You really need to take a look at the gcode in something like Repetier Host, and see how the loops line up. If you can see gaps between the filament layers when you view it from the top, then it's not going to print well.