Ok so my head is a little dizzy but the results of my completely un-scientific testing are back and hopefully my wife doesn't find out that I used one of her baking trays to melt plastic on
I tested a mix of support scraps / test prints / failed prints in 3 different brands of plain PLA and 1 brand of PETG and used a long necked BBQ lighter to apply heat.
The most interesting thing I found was the main factor that determined if a flame will continue to sustain itself or not is the orientation of the part vs the flame, not the actual material!
All 3 brands of PLA (1 cheap red, 1 cheap white and 1 high-end white) were very difficult to ignite from the top (flame facing down onto the part or from the side). It consistently took around 10 seconds of direct blue flame being applied before the PLA was able to sustain its own flame which then always burnt itself out in a second or two. For the flame to sustain itself it needs to use the melting plastic as fuel, with the flame on top of the part gravity pulls the melting plastic down and away from the flame, and the flame is not able to melt the next layer of plastic fast enough to sustain itself.
The same occurred with PETG except the PETG needed 15+ seconds to ignite and them almost always extinguished itself in less than 1 second.
This is not surprising as in the video above Angus was pointing the flame at the top of the printed part. I suspect the ABS part and PLA+ parts acted differently as the material was able to act as fuel for the flame before it was fully melted (unlike with PLA or PETG).
However I wanted to try and simulate an electrical fire inside a power board so I took the top of V1 of the power board (high-end plain white PLA) that I had printed a while ago and put the long neck of the lighter inside the print to light the flame from underneath.
What I found was that a flame was able to sustain itself after only 6-7 seconds of direct blue flame being applied and it would not extinguish itself. With the melting plastic dripping down through the flame it was able to fuel the flame on enough to melt the next layer up and out to the side, continuing the cycle and spreading the fire (and likely would have kept going until there was almost nothing left).
Curious what the PETG would do I propped the failed PETG print up and applied the blue flame from underneath. It also ignited a little faster (about 10 seconds each time i tried) and it also
kept on burning with no signs of slowing down.
So my conclusion is - as long as the flame is coming in from underneath a part so the melting plastic can drip into the flame and add fuel to the fire, both PLA and PETG with in fact sustain a flame after being ignited.
My other observations are:
- PLA was quite slow burning, the fire did not spread quickly (in my opinion at least
) but once the PETG got going the fire definitely sped up a little.
- Both PLA and PETG fueled fires could be extinguished with a very light breeze (much easier to blow out than birthday candles
). To try and get a repeatable test I set up a very weak 12V DC computer fan 50cm away from the fire. Each time I was able to get a good fire going, plugging in the weak fan was all that was needed to extinguish the fire within a couple of seconds. I would say putting an intake and an exhaust fan on the side of my electrical cabinet (on my to do list anyway
) would create a stronger draft going past all parts.
- Neither PLA or PETG really had any after-glow or were able to be easily re-ignited. Once the flame was extinguished it took almost as long as before to start another fire. If the part was left to cool for 10-15 seconds, it took approx 5 seconds of a constant flame for PLA to reignite and approx 8 seconds for PETG to reignite.
And probably the most important observation IMO (already stated above) - Both PLA and PETG took quite a long time to ignite. 6-7 seconds (PLA) and 10 seconds (PETG) of a constant blue flame is already a pretty decent fire so they would actually need a fire to have already well and truly started before they could add fuel to it.
It is because of that last observation that
I am still fairly comfortable using PLA or PETG (the latter is probably the preferred as it takes almost twice as long to ignite) for 3D printed electrical housings. I do not think they are as safe as UL-94V rated enclosures, but I am fairly confident they are no more dangerous than normal power bars or ABS project boxes. All of my power boards will also be run through a fused switch and then all the circuits in the house have an RCD. I believe that both of those safety measures would need to fail before an electrical fire is a possibility.
I do plan to place a smoke detector inside my electrical cabinet above my tank's electrical components and also add some intake and exhaust fans to create some airflow past all the components, but those are both things I already have been meaning to do for a while as they are just good things to have in a wooden cabinet that is packed with electrical equipment
