Wow. A ton of discussion on the forum about this topic.
Gotta admit that although you can find some great scary pics on the web and even YouTube (who is standing watching their laser catch fire and videoing it instead of putting it out? Sheesh) - you’re more likely to get stinky fumes of burning bodies (leather just stinks) or burning plastic that bother people nearby than you are to have a fire.
Don’t lase stuff you shouldn’t (some prohibited plastics/foams are pretty flammable) and stay with the machine in the event it stalls with the laser on burning a bigger and bigger hole into something. Manage your power & speed and you’ll be okay (and you’ll get other undesirable affects from too slow/too high power like staining, oversized cuts, etc way before it burns).
Honestly you’ll probably spill a soft drink in there before you ever have something catch fire. Maybe a thread on how to soak up spilled drinks is in order (yes, it has happened to me:grin:).
A 40W laser beam is powerful enough to burn through paper instantly, even before it is focused. It would have to be a long way below the focal point to lose its burning potential. I think at the focal length from the focal point it would be back to its original diameter and the focal length is 2".
@dan - What would be a very cool safety feature is if the overhead camera could detect a significant flame ignition on the material (ie. cardboard), sound an alarm, interrupt the cut and move out of harms way. I can’t imagine a flame burning under the head assembly would be a very healthy thing for the optics…
The GF friendly neighborhood lawyer might look unkindly at the implied liability of such a feature. A claimed flame detector feature seems like something that would invite lawsuits, but maybe I’m just feeling especially cynical today. I personally like the idea, even if it only beeps at the user to alert them to a potential problem…
Detecting a fire through the cameras would be VERY hard. The camera on the head only sees the area being cut and a LITTLE more. So any flame in that vision is not particularly indicative of a problem. Maybe your air assist is off, or maybe the material is just VERY flammable.
The camera on the lid has a large portion of the bed obscured by the gantry, and is looking at a surface which should be changing in appearance (due to the engrave job taking place).
So you would have to monitor the camera continuously, and run algorithms which specifically test for fire. I suppose you could do some iR and visual intensity checks, but the laser itself and the ablation action reflecting off of surfaces would cause that to a degree in normal operation. So you have to look for longterm and/or widespread effects. But fire is naturally unstable, so you cannot just look for prolonged changes in isolated areas.
Basically, for any “I am absolutely certain there is a fire now!” detection via camera… the fire has to already be far more advanced than you ever want it to be. AND it requires an on board processor to be doing the monitoring. You don’t want to be waiting on upload speeds and cloud processing time (nor do you want to be using that much bandwidth with every cut)
yeah I wouldn’t use the camera. if I were to solve the problem I’d put sensors on the edges of the head and use the fact that the fan solution should draw much of the smoke/vapors down through the crumb tray and trigger of an abnormally high temperature. but that still only covers a small area behind the cut/engrave. A slow growing flame would likely not trip.
MAYBE the laser could do a “controlled burn” around the part of the material that’s on fire! Isolate the material by a wide enough gap that it doesn’t burn out of control.
I’m joking… but I also think it could work… but it also probably wouldn’t.
I would be concerned about the head of the laser passing over the flame, or being that near the flame for that long. I would expect that it would end up damaging the head in some way?
I could see that make sense if “fire suppression” was part of the idea - ie they added a pressurized feed to a extinguisher that failed - but it’s not. You might as well call it Cruise Control; it doesn’t remove the operator from operating the equipment responsibly.
The purpose is to move the head away from the unsafe zone so that the operator has a clear unobstructed vector to the flame point, to toss a wet rag or extinguish the fire. The last thing I would hope someone wants to do is pop the lid open and slam the head aside in order to reach a fire.
Agreed. But putting in an option to “move head furthest away from last cutting location when lid is opened” (vs. “do not move head when lid is opened”) is not inconceivable.
The only way a laser can clear a lot of material from a large area is to burn it. But trying to burn too much material too close together is precisely how the fire started in the first place. So you would isolate the fire by starting a lot of other fires all around it. Then isolate those by starting more around them. Then isolate those…
@dan How does the glowforge hande ‘hard’ power cutoffs? i.e. Cutting a material that starts a fire and the onboard exhaust fan can’t keep up with the smoke. Trips a nearby smoke detector which then kills power to the glowforge.
Obviously a nearby user would extinguish the flame/smoke and remove said material that caused the problem. But would there be an issue for the internals if power was cut during a op?
i.e. Thermal shock on the tube? Since the GF uses a peltier /w an unknown amount of magic coolant (reservoir size) would the standing fluid in the closed loop at temp cause an issue?
I’d assume from past convos about homing, it would simply rehome and also assume if it detects if the coolant temp is higher than nominal actions will be taken?
(yes, it has happened to me:grin:).
Had a housemate in college who rebuilt motorcycles in his bedroom.