Under Extruded Layer in middle of otherwise perfect print

[Wolverineairsoft 0]

Hey all, have run into a baffling issue that I haven't been able to track down. I'm printing ABS on a brand new UM2. What happens is my prints are going along perfectly and then will randomly massively under extrude for a layer or two and then return to normal operation. By then however the part is ruined and breaks apart. Here are some pictures. Don't mind the ugly bed...I've been using an acetone/abs glue to stick parts down.

Things I have tried that I can think of:

 

• loosening material on spool
  
• slowing print speed to 50mm/s
  
• atomic clean
  

 

I'm sure there are other things...I've been working on it for a while. It is not a problem with the gcode. Sometimes the same part will print fine.

Any suggestions would be greatly appreciated. Right now I can't find any suggestions to try other than that the Teflon isolator could be bad...but that doesn't seem reasonable considering this started after only about 5 hours of printing.

[gr5 2,270]

This could be underextrusion, or it could be sudden Z movement.

Does it usually happen at the same spot? Anyway there are 2 common causes of sudden Z movement. One is having too high of a current on the Z motor - some recent PCBs have the z stepper driver overheat and it shuts off for just a second which is long enough for the table to slip under it's own weight. Also if you use S3D the typical profile sets the Z current much too high. Cura doesn't modify Z current. You can lower the current on the front panel of the latest versions of marlin including tinker marlin. Default I believe is 1300ma, S3D likes 1400ma, UM recommends 1200ma, some people out in the wild use 1000ma. Hot air temperatures under the printer also can make this problem worse.

Also sometimes the bearings stick and then slip suddenly. You can test this by moving the bed up and down with your fingers with the power off. Lift from the back. Go past the bad spot many times - if you feel anything then your bearings should be removed form the bed and experimented with and possibly cleaned.

[Wolverineairsoft 0]

Ok thanks for the tips. Bed definitely is a bit gritty especially at the top. But doesn't bind.

Can I change the current during a print? Also would adding a pause and resume help it not overheating?

[Tomty 4]

I have just resolved this problem with my printer. Here is the link to my original post and my solution. https://ultimaker.com/en/community/17753-um-2-skipping-layers?page=1&sort=#reply-125974 . At the bottom I have described how I solved it. I found that the problem was due to the bearings sticking so after taking it apart and removing the bearings, I cleaned them through with WD-40. I have since printed many things without any issue and feel confident to print anything without the risk of this happening again.

Hope this solves it for you.

[Wolverineairsoft 0]

Awesome! I'm a little baffled cause I examined 3 failed prints more closely. Two of them had compression/mushrooming I the layers immediately before the failure as you would expect if the bed is sticking and then dropping. However the third did not. Pristine layers up to a sudden failure. Possible I have both a sticking and an overheating issue? I have reduced the z current to 1000 and so far the print has been good, but that doesn't necessarily prove anything. Will report back once I have more data.

[Wolverineairsoft 0]

Incidentally, I have not put anything on the z posts and they seem dry and free of grease.

[Wolverineairsoft 0]

That one came out perfect. Printing another taller part now to see if the change was sufficient or I need to do the bearings as well.

[Wolverineairsoft 0]

Next test was...interesting. First inch printed perfectly. .25mm 80m/s. After that the print was looking under extruded and not adhering well between layers. However it didn't ever completely skip a layer. thoughts? At 105 or 110% flow I think. 105 on the tray and 260 at the tip.

[gr5 2,270]

260? For abs or pla? I recommend never going over 240C for PLA. You may have a coating of carbon now inside your nozzle if you did PLA at 260C.

How fast are you printing and how thick are the layers?

[Wolverineairsoft 0]

This is ABS, not PLA so 260 should be fine. It seems I have both the overheating z-axis controller and a binding bearing...so that's fun. The one is fixed, the other still needs to be dealt with. I am fighting a different issue as well however.

Print is:

ABS

.25mm layer thickness

.4 mm tip

80mm/s

110 on the bed

260 on the tip

It was a thin wall part with all wall thicknesses right at 1.2mm

Throughout the entire part (175mm) I had extremely poor layer cohesion. It never actually seperated, but it looks week and has small gaps. my daughter dropped it on the floor from only about 18" and it broke in half. Very weak print. I tried slowing the printer speed to 85% part way through and it seemed to make no difference.

[Wolverineairsoft 0]

Actually should clarify printer basic speed is 80mm/s but inner shell was 50, outer shell was 60mm/s.

[Wolverineairsoft 0]

Linear bearings are binding on the shaft. Removed bearing from build plate and it binds all by itself. Can't say I'm impressed by this. :-/ pretty glaring qc problem. I suppose I shall be contacting Ultimaker about a replacement.

[Wolverineairsoft 0]

OK so after a week and a half of printing I am pretty sure this is sorted out. The first issue I had was the z-axis controller overheating. I changed the steady-state amperage to 1000. However I still had the issue of The z-axis bearings binding. I remove the rods and bearings and tested and found that it was just one of them binding. I cleaned the Bering thoroughly with WD-40. This took care of the sticking and after letting it dry I reapplied some lithium bearing grease to the linear bearing. I do not suggest allowing the Bering to run with no grease on it at all. Despite the fact that this bearing she's very limited travel it will only lead to premature wear on the bearing and the shaft. after reassembly I am seeing no problems with z-axis errors. Thanks for all that posted up ideas for me to look at.

[gr5 2,270]

Throughout the entire part (175mm) I had extremely poor layer cohesion. It never actually seperated, but it looks week and has small gaps.

Probably underextrusion. I know you slowed to 85% but next time you get underextrusion try slowing to 50% or even 30% for a few layers just to see.

Also with ABS you want very little fan. It's not needed anywhere except overhangs and bridging. The older slicers would turn the fan on just for overhangs and turn it off - it might go on and off 10 times on a given layer. I usually just do the minimum fan (around 30%) for the whole print.