Fire Hazard

Just wanted to give an update on the rebuild of my RepRapPro Tricolor printer that caught on fire and led to the start of this blog. Sorry it has taken so long for this but I have been traveling for a good part of the past two months. As I've said before, I really value the response / ideas / recommendations that have come out of this discussion and have incorporated several of them into my ongoing redesign. First off, since I could not print new parts, I improvised by making wooden parts to just get the printer up and running. Yes, they are very flammable but they are just temporary until I get the printer running reliably enough to make new plastic parts. On the safety side, I have fabricated a heat shield that is installed under the top frame and protects all of the filament drives from any potential flames. This was formed from a sheet of galvanized steel and is attached to the two top connecting rods with some wire end lug connectors bought at Home Depot. Secondly, I rewired the printer with all of the wires installed inside conduit made from 1/2" copper pipe (again purchased at the local hardware store). Most of the conduit "change-of-direction" fittings were modified with cutouts to ease the routing of the wires during installation. I then purchased two new metal hotends (E3Dv6). Now that the printer is 95% operational, I plan to convert the x-axis frame to two side-by-side rods as per the original Mendel design that will allow me to easily mount the new dual hot ends using a design found on Thingiverse (http://www.thingiverse.com/thing:364650). I still plan to install a thermal fuse in the vicinity of the hot end to sense an over-heating condition and shut off the power. I'll add new updates as the rebuild progresses. Photos attached.
Gonetotx

If you use a strip of that steel plate as a heat break from the thermal block you could in essence use a lower rated thermal fuse. All it would take is some experimentation on length of strip. Let's say you heat up the hot end to 200c and measure the steel strip at the edge. And it measures 100c and when you heat up at 240c it measures at 124c then you have the curve you want. Select proper thermal fuse for let's say 200c and attach in line with the hot end. Then fool the printer to heat up the hot end and see if it trips. An ir thermometer could come in handy. In theory you should be able to get a proper range and some much added safety. You could start using some of those resistors instead of cartridges as well.

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

Jaguarking,
You have read my mind. When I change the x-axis to a horizontal frame, the hot end support rods will allow a lower and flat hot end base. Once I get this assembled, I can then sketch out the thermal barrier you suggested that would fit between the thermal block and the fluted cooling section. My thinking is that this thermal shield should actually wrap around the sides of the hot end assembly to better isolate the plastic parts from any flames that might come from a burning part being printed. At this point, I can do some testing for the optimal point to locate the thermal fuse. I like your idea to locate the fuse at a point where the temperature is lower than normal operating temperature (240C for ABS). I'll have to do a search for a reasonable fuse temp range and then find a location on the assemble to match that temp. I'll post an update when I reach that point. The general safety guidance here is to install steel thermal shields under and around the sides of all exposed plastic parts above the hot plate and place as much of the wiring inside metal conduit to shield it from flames. In spite of adding features to prevent thermal runaway, I believe there will always be that potential and so there will always be the potential for the part being printed to catch on fire. Finally, my control board has the ability to add dual thermistor control on at least one of the hot ends to automatically shut the system down if the temperature difference exceeds a set value. This is a cheap, easy and effective way to prevent thermal runaway. Any thoughts or suggestions on this plan?

The only thing I do not like is the second thermal probe. Your adding two more wires, and two more connectors. Your increasing complexity by two fold, the chances of failure double as well.

One thing I am using at this point are the RC car battery connectors TX-60 style. They are inexpensive and can carry up to 60amps. With that being said I would find a way to fasten the connectors to the head and have the 4 wires, 2 heater and 2 thermistor securely fastened. This should reduce the point of failure and fatigue. Being that I use a bowden setup this gives me the ability to fasten the connectors to the bowden tube. Giving me very little flex. Another solution is to buy some very highly elastic silicone coated wires. Generally high end soldering irons have them. Also, try to build a metal holder for the hot ends, skip the plastic crap as it will warp and it is flammable.

Another point to take, I have seen richrap use a sheet of stainless to cover the thermal block. To me it seems like a safe way to proceed. The cartridges cannot produce enough heat to melt the stainless captor.

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

[XT-60]

---

Realizer- One who realizes dreams by making them a reality either by possibility or by completion. Also creating or renewing hopes of dreams.
"keep in mind, even the best printer can not print with the best filament if the user is the problem." -Ohmarinus

Hi,

am I glad I found this forum.

I am just about to build an i3.

It will actually be an xmas present for my son, we will build it together. He will do 90% of the work under my supervision.

As he will probably use it unsupervised safety is an important consideration.

I now realise that step 1 in the project is risk assessment and avoidance so here is my plan

1 build a cabinet, forget polycarb and tech materials, good old 4x2 and plasterboard will do nicely, the volatile components won't burn for too long, so a single layer of 12mm board will suffice.I will make the base from a double layer of 15mm chipboard. The lid will hinge up and over to make access easy. materials approx £30

2 in the top of the cab will be a mains powered kitchen heat detector. These things go off with high temp rather than fumes. £20. They have a signal wire that is intended to connect to other smoke alarms in the house, If this wire is connected via a 10w 5k resistor to the live wire to the heat detector, it will trip a RCD

3 I have a modern UK power distribution board that has a 30mA RCD, so my heat detector will trip the power to the whole house so this is the failsafe system in place. I figure the worse that can happen is damage to the printer.

4 as suggested mount a temp sensitive fuse above the hot end, wire it directly in series with the heater.

5 choose a heater with the correct resistance to warm the nozzle to the correct temp, but will not get much hotter ie it will reach an equilibrium temperature of not more than 230c for PLA., hopefully in this way the use of a fan can be avoided - I am not bothered that warm up could be slow.

6 make sure that thermistors are securely mounted and that the wiring is secured to avoid strain.

-----------------

The good thing about putting the printer in a cabinet, is it will keep the bed warm and will also allow me to increase the rigidity of the frame by putting in a top support. It will also allow me to install an extractor fan to get rid of fumes, initially I was thinking of mounting a cooker hood on top of the cab, but the vibration could be a problem, so I think I will use cooker hood ducting and put the fan on the outside wall.

Any other recommendations welcomed.

Pete

Edited 1 time(s). Last edit at 12/14/2014 08:19PM by repete.

repete:

You would not want a detector to cut all the power to the house. You might still end up with a fire, but in the dark with all the lighting turned off and trying to get out of the house with your wife and kids. I have a smoke detector next to my printer, which is wired into a relay that cuts the power to an extension lead, into which I have the printer plugged. Of course it does not help if the fire has well and truly taken hold.
Also the latest Repetier firmware has coding for stopping a thermal runaway. It should catch if a thermistor is pulled out and the hotend temp races away.
I would not personally print overnight or when I am out of the house.

I figured I would add to this. Recently I had a thermal probe come loose on my hot end(my own damn fault) and the hot end cooked itself for close to 2 hours using a 40w heater cartridge. No fire and no craziness, the abs I was using formed a solid charred block in the hot end and blocked the printer. The second thing is that I used a larger than normal heater block, still in aluminum but a much larger thermal mass.

The things I took away from this are.

1 - I got lucky
2 - a larger heater block prevented melt down and even saved the cartridge.
3 - the thermal plastic will stop flowing at higher temperatures.
4 - tighten the set screws properly.

Other thoughts?

I think we need to switch to brass thermal blocks, they may take longer to heat up as they have a denser mass but it should prevent the temperatures from spiking causing a natural temperature control. Also all metal hot ends are a godsend. No more PEEK or teflon for me. I even made a rebuild video of my hot end. [www.youtube.com]

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

Deleted.

Edited 2 time(s). Last edit at 06/15/2015 11:28PM by amaurer.

It seems like there are some firmware features that would help here too. For instance you could set a Max Printing PWM value that would kick in after the hot end reached temp and the print started. That way if the thermistor came loose the hot end wouldn't go to full power. Most firmware already detects a disconnected thermistor as a MINTEMP error so it doesn't seem like it would be that hard to also add a Heat Failure error that kicks in if PWM goes up and temp does not follow. With autotune for the PWM system, the firmware already has a pretty good idea of how the heaters should react. Firmware solutions can fail but it would be an easy cheap addition to our safety efforts. Hardware fail safe systems should probably be the first line of defense though.

PS. After posting I did a little investigation. It looks like Repetier is implementing detection of decoupling of the heater from the thermistor. I'm not sure if other firmware has followed suit or not.

Edited 1 time(s). Last edit at 09/04/2015 02:47PM by bryanandaimee.

Quote
Dust
Or you could just apply some common sense.

Its a heater (well actually its two) , you don’t ever leave a heater unattended!

some printed take 24hrs to complete....