Forgive my ignorance: How exactly did that fire start? I've literally never heard of anything even vaguely close happening in any shop I've been in. Burying the head/collision, "yes", but nothing like that. Just curious.
Erik
Printable View
Forgive my ignorance: How exactly did that fire start? I've literally never heard of anything even vaguely close happening in any shop I've been in. Burying the head/collision, "yes", but nothing like that. Just curious.
Erik
Erik, generally the fire starts because of friction and the heat it can generate under certain conditions and with certain materials being more risky. Heat from the tooling is normally thrown off (literally) via proper chip load: the correct speeds and feeds where the chips produced carry the heat away. When material, such as the fine dust generated by materials like MDF, gets packed into the "kerf" of the cut, the temperature of the tools can skyrocket in some cases and that, in-turn, ignites the dust. This can also happen when the workpiece isn't secured and it causes the tooling to remain in physical contact with the material in essentially one place and, well...it gets hot. Working "deep" with a down-cut also raises the risk because it packs the swarf in, rather than releases it from the cut line. These fires are pretty much always because the material gets ignited from heated tooling. Sometimes it's just a smolder that suddenly gets a mind of its own...that can be the nature of "kindling". Rarely is the machine itself bursting into flames.
It is hard to tell if the machine lost steps and then knocked the blank loose from its hold down or the hold down failed and that created the loss of steps. In any case an unpredictable, but not uncommon event occurred that caused a fire. Happens often. I have experienced 3 spoilboard fires. Each time I was there to shut down and extinguish the fire. Never fast enough tho. In one case the PVC manifold was destroyed.
CNC fires are like your wife getting pregnant. You really want to be there when it starts! :)
I get it now, Jim. The lack of or inefficiency of dust collection on that unit in the video is actually pretty shocking. It's more like one those "how to start a fire" bushcraft videos. I guess I am just not used to seeing such an accumulation of sawdust on the workpiece or spoilboard.
Erik
Dust collection helps, Erik, but it doesn't clean out impacted cut lines at all. It can only catch the chips that actually get up into the air. The issue isn't lose dust; it's impacted dust and material that stops the heat from being dissipated.
Jim, I get it but still find that crazy. You would never see that kind of chip buildup on a commercial router. There is so much wind and extraction happening that the buildup just never occurs in the first place. Wild stuff.
Erik
Erik, the first time your tool holder hits a part you are cutting due to any number of errors an operator can make you will see how easy this can happen. I did it years ago, and in very short order (seconds) the wood was black, and the tool holder was too hot to touch. A short time longer and I would have had a fire. The DC does squat to prevent this as the collet is spinning anywhere from 16,000-24,000rpm. The video in question is an example of a lot going wrong, but we all have those days. I have had 36hr cuts and left my machine unattended for periods of time. I am very nervous about it, but I only do it after the roughing pass. The one time I had a problem I broke the tip of a ball nosed bit. Should I do it, nope, but...
I suspect that dust collection is a double edged sword. It helps remove the chips and keep the cutter clean. But it also creates air movement which basically fan wood smoldering.
Watching that video it almost looks like the cutter is occasionally hitting something metal, like a hold down clamp or screw. If so that'll dull the cutter which will also create lots of heat.
Wouldn't a higher end CNC catch the loss of step count? I assumed only the cheaper hobbyist versions didn't have an encoder. With feedback the controller should know if there's any loss of steps and hopefully after a preset number abort.