Ultimaker 2

Tuned by Martin Ling 2020/04/01

Filament comes on 1kg spools. PETG has a density of 1.27g/cm^3. So there's about 124 metres on a 2.85mm 1kg spool according to this calculator (assumes you're weighing a 248g empty spool plus material, so put in 1248g to get that number

Cura spits out 2m of 2.85mm filament for the 3dverkstan design. So we get something like 124/2 = 62 copies per spool. At 11 minutes per print that means 11h21m of print time per spool.

Material Configuration

PETG (Grey stuff)

Stored as CUSTOM5 on the printer

  • Nozzle is a 1mm ("1.0 Match Race - World's Fastest Nozzle" 3dsolex Matchless nozzle)
  • Nozzle temperature increased from 245 to 255C.
  • Buildplate temperature increased from 60 to 65C.
  • Fan speed reduced from 50% to 25%.
  • Material flow increased from 100% to 105%.
  • Print speed reduced from 50 to 45 mm/s.

Pushoff / Restart

  • Cooldown now starts two layers later and won't cool below 45C.
  • Z height increased from 0.5mm to 1.0mm during pushing operations to avoid dragging drips from the nozzle across the bed so much.

Pusher design notes

The idea is to just have a beam that sits about 1mm higher than the nozzle, so it's clear of the printing process but can be lowered after printing to push the part off with a bit of custom G-code appended to the end of the print code. Everything else about the part is just shaped as needed to allow it to be attached to the head - in that case by cable ties.

The key thing is getting the level of print bed adhesion right - it needs to be strong enough that the part doesn't shift of its own accord during printing, but weak enough that the carriage can break off the print without stalling its steppers. I am working on tuning this. We found it easier to achieve with PLA filament but I'm trying to make it work with PETG, which is preferred for these items.

Problems encountered & solutions

  1. Problem: The Ultimaker 2 carriage cannot push much on the X/Y axes. Any strong horizontal resistance will cause the drive belts to jump teeth. Solution: nudge the part in several different places to gradually free it from the bed before pushing off.

  2. Problem: The big flat area in the design was still stuck too hard. Solution: let the bed cool before trying to push off. This helps the part to "pop" off the bed.

  3. Problem: Waiting for the bed to cool takes ages, hurting the production rate. Solution: start cooling the bed in the middle of the print. This can be done by inserting an "M140 S" command to change the setpoint. I put this after the first 2mm / 5 layers. As soon as the print is complete, do the same to restore to the normal setpoint.

  4. Problem: bed adhesion.

Solution: A very simple solution to them getting stuck under the build plate is to add a sheet of paper like this https://www.youtube.com/watch?v=avlengYsJdw

  1. Problem: pusher retention: The pusher continuously slides up, and at some point it's too high to remove the print, and the nozzle drives straight into the print.

Solution: We need to figure out a better clamping system.

Retuning notes for different batch of grey PETG material (values given above are for this stuff)

This stuff likes to run a bit hotter, and I had to back off the speed and the fan cooling a bit to allow the heater to keep up with that. It will also pop off the bed without being cooled as far, so I had to tweak the cooldown to stop it popping prematurely at the rear hook ends.

Unfortunately, I haven't found a sweet spot yet where the big flat section can be popped off but the hook ends don't pop off the bed prematurely.

It might be necessary to add some wait time ("G4 ") to the g-code to let things cool down more before trying the push off.

I guess you could also try parking the head over the flat bit and running the fans hard to help cool it quicker.