Saturday, 16 January 2016

Interesting Filament Reading Test Print (model not yet available)
I came across this nicely put together series of articles on 3D print filaments, you might find as interesting as I did.

They write an easy to follow article on quite a few filaments together with some analysis.

For reference at the moment I typically use:
  Faberdashery and recently sometimes Makerbot PLA filaments
  0.2mm or 0.25mm layer heights
  15% infill
  Hexagonal form infill
  28 mm/s print speed
  Print temperatures ~200 C

There is a rather large caveat list in the article. Though some additional things not mentioned in their article:
- Faberdashery supply in meter, yes meter lengths!
- No smell test!! Faberdashery smells nice (apparently)!!!
- No colour consistency

- No colour palette evaluation
- No details on purging between filaments
- No mention of filament pre conditioning or storage methodology

The latter seems a little odd given the article breadth, so its surprising though it does state it's a high level summary. Still it would seem unlikely all materials arrived at exactly the same time. Industry practice for such report and testing would include a description of purging technique, storage, drying times, storage temperatures and so on. As plastic can absorb a good amount of moisture which will have some impact on performance. If you have had filament in storage for a while even with silica gel. Consider placing it somewhere nice and warm say ~ 60C for a good couple of hours before use. Professional injection moulding facilities use de-humidifiers and heaters to pre condition polymers prior to moulding. For the article it would have been good to understand supplied material moisture content, assuming the filaments were not pre-conditioned. We are left to assume parts moved from manufacturing to materials test lab in a controlled manner.

How do you measure moisture content, simple enough. Put a length of filament on a weigh scale and note its weight. Pop it in an oven for say 24 hours at 60C (with your silica gel). Allow to cool in a dry environment (the same oven with now dried silica gel for example) then weigh the filament again. Given filament cross section it should have dried out thoroughly by this time and probably a couple of hours would suffice. Do the maths to calculate % by volume absorption. Is moisture take up a risk, well the average humidity where I live is around 80%, though December 2015 nearer 99%.

My main takeaway from the article is keep print speed down on external walls, perhaps drop the infill percentage a touch, increase my print temperatures by around 10 to 15 C. The latter being something I have been doing more recently having read another research article on 3D print strength.

I will probably keep 0.2 and 0.25 mm Z height increments as these increments bridge the gap between print speed, strength and give 5 or 4 layers per mm. So print my engineering type design final layers without Z heights rounding errors, especially for some of my smaller prints. I will keep moving to 220 C print temperatures and adjust the infill to balance print time, filament use and print strength. This latter being a relatively small change from my more usual 0.15%. Though top surface fill quality needs to be taken into account as well. So this is where I intend to move to:
  0.2mm or 0.25mm layer heights (see above on rounding errors)
  10% infill
  Hexagonal form infill
  28 mm/s outer layer print speed
  Print temperatures ~220 C

For long enough in the UK you couldn't source Makerbot filament. It's still a major hassle to get it readily. More often than not, on 1 months backorder for basic colours! Hence much of my early experience was and remains with Faberdashery filament. Seems the research and advice given, led to selection of a good filament. I see the price of Faberdashery filament holding and reels becoming available.  I really did not like the eSun filament purchased from RepRapPro as part of an order with them. It was sort of OK when just out of the packet, however after storage for a week or so it did not print at all well.

Here are some of my previous prints, in Faberdashery filaments:
Test print Huxley #710
Touchololu for Huxley #710
RASPI case by Huxley #710

No comments:

Post a Comment