splits and gaps in large print

I thought I was getting some settings dialed in and the Taz was doing pretty good on most little prints. We needed to make a mold core so drew it up and sent it to the printer. The print took in excess of 48 hours to complete, and in general the machine functioned fine. But, as the piece got taller, splits and gaps began to appear on about 1 cm increments. In addition, they became worse as the print progressed. I’m assuming it might be a cooling issue or maybe a z axis indexing problem, or hopefully just a setting I should change. I’ve attached a picture and would attach the gcode file, but the page crashes when I do that. Please advise.

There’s a pretty long discussion with ways to minimize this here:


The problem is uneven cooling of the ABS which causes it to shrink at different rates. Several of us have been dealing with this issue. The thread above has lots of good information, but the quick summary is:

  1. Use an enclosure around the build area or whole TAZ to keep in the heat
  2. Print a tall skirt around the part
  3. Look at using a material other than ABS that isn’t as prone to warping as it cools (HIPS or PLA were suggested)
  • Adam

OK, I actually have to make a larger section that mates with the conical one above. I built an enclosure, added a surveillance camera, and a thermocouple iserver. This print is supposed to take 17 hrs, my guess is quite a bit longer. I also reduced the infill to .5, and a few other adjustments. Looks like the temp in the enclosure will settle in around 100 deg F. I’ll know more tonight. Fortunately, I can check both the camera and the thermo server remotely, just in case there’s a wreck.

Actually, the temperature inside the enclosure seems to have settled in at somewhere beetween 105 F and 110 F. I have a thermocouple on the left side of the enclosure that reads the higher temp and one on the left that reads lower. The TC that’s on the left is above the power supply so I figure it might be getting some residual heat from there. The other TC is pretty close to the bed. The print is about 16% complete and so far looks pretty good. No splits or gaps yet,? So the question is… is there a max temp that is recommended within the enclosure? I noticed where Lulzbot keeps the temp around 90 F to 100 F, but is 110 F too hot? with some experimentation It seems I can control the temp by opening or closing the plastic drape, but it seems that it might be detrimental to have too much convective flow in the chamber. Then again, I have the power supply in the enclosure and the fan in it seems to circulate the air pretty well. Maybe having the PS in there is a bad idea?

Power supply and controller in the enclodure = bad idea. It may not cook it the first time, but it definitly will eventually.

The 110 temperature for ABS shouldn’t hurt anything. white ABS might start to discolor a bit at above 112F or so, but if anything structurally it will help. that being said, keeping it at 90 probably takes less energy and should do the trick too.

I pulled the power supply out of the enclosure. It was definitely running on the warm side. The temperature of the enclosure has settled down closer to 100 F without the additional heat source. The splits and cracks are still developing, though. Creating plastic parts like this is a critical function for this machine, so I need to keep working on a solution to the cracking/splitting/delamination issue. It seems obvious that the size and thickness of this particular shape would suggest using a larger nozzle. Would a larger nozzle help with the splitting? I would assume it would lessen the print time.

Larger nozzle should help. I’d also try increasing the temperature of your nozzle slightly to get a better layer bond, keeping in mind that the closer you get to the PEEK isolator melting point the more likely it will melt, but a couple degrees might be what you need.

F, F, D, S, F!!

I ran out of filament before it was complete. I paused the print and put a new spool on, thinking it just ran out and maybe the print would continue. Then instead of clicking the “Resume” button, I did a major brain frt and clicked “Restart”. :astonished:

Oh well, the print had quite a few cracks and splits, and I’m going to try HIPS, a larger nozzle, and Oriases’ config file which should be much faster and use less filament. Man, this 3D printing has a pretty steep learning curve, but at least I’m learning… I hope. :confused: Thanks for all the help.