Results 1 to 13 of 13
  1. #1
    BobL is offline Member: Blue and white apron brigade
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    7,182

    Default Heated 3D printer enclosures.

    A month or so back over in the woodies forums there was a discussion about setting up a heated 3D printer enclosure, so recently I decided to have a go at doing this to my enclosure. A heated enclosure is typically used to print certain materials without warping.

    My current enclosure looked like this.
    enc3.JPG
    My enclosure is based on 20mm square tube Al frame held together by 3D printed Qubeloc style connectors nd covered in a 1mm thick PC sheet. It was not necessarily intended to be heated but to keep the printer clean and reduce the effect of drafts from a nearby AC. I chose clear PC primarily for visibility and so external lighting could be used.

    When printing with the print nozzle at ~200ºC and the (heated) printer bed @ 50ºC the temperature inside the enclosure is typically 5-7º hotter than room temperature

    Normally to get to the desired heated enclosure temperatures of say 50ºC a thermally insulated enclosure and some form of heat like an incandescent globe or small heater, are required. It just so happened that I had replaced the stock heated printer bed with a PID controlled higher temperature bed but I did not disconnect the original bed from the printer and just leant it up against the inside of the enclosure and while printing, the temp of the original bed was set to zero BUT I could run that up to 100ºC to add extra heat. When I did this it only added a few degrees to the enclosure temperature which indicated the added heat was simply being lost through the very transparent PC walls.

    In the woodies discussion I suggested using a foil rescue/survival blanket as a reflector on walls to help retain heat. However , I didn't want to apply this to the PC since I would wish to be easily able to remove any heat retaining covering/materials. There was also a mention of using "core flute " as lightweight insulation for an enclosure so I decided to combine the two materials and purchase a sheet of 2400x1200 mm black core flute from Bunnings and a $5 foil blanket from Ebay which turned out to be aluminised PET. The first thing I noticed was the foil blanket was slightly transparent so this may in part explain the final outcome.

    I started by sticking the blanket to the core flute - a tricky job because the space blanket is very flimsy and static as all get out, and kept sticking in all but the right places. Then I cut the foil covered core flute forming a large "U" to wrap around the back and two sides, and made a shallow box like structure that popped over the top that helped hold the sides sides and back up against the PC .

    The front door has a foil covered core flute panel inset into the inside, held in place by two pieces of lightweight 12x20 mm Al angle along the verticals. This panel can be very easily popped in and out as required. If needed the whole thing can be removed/replaced in 30s.

    Not that exciting to see, but here it is - apart from the door, a black box covering a transparent box on 4 sides and VV on th front.
    The Al angle strip across the top holds the sides of the shallow core flute top box, up against the sides of tehthecore flute wrap around .
    Enclosureh.jpg
    Eventually, when I have the front door panel in place, I will have to cut a small viewing port in the core flute/foil to see what is going oil while printing.

    Perhaps more interesting is the inside showing the foil covering the core flute
    Enclosureopen.jpg

    Below is the view with the front door closed but the door's core flute panel removed.
    If enclosure heating is not required the visibility is good enough to see what's going on while printing without addition lighting which means I do not need to remove the whole foil covered core flute box for non-heated printing. I will eventually install a light because I will need to see what is going on when the door's core flute panel is in place and looking through the view port.
    Encfrontremov.jpg
    Here you can also see the original printers heated bed leaning up against the LHS wall of the enclosure.
    It's the gold foil coloured rectangle.

    I would have though that covering 3 sides and the top would significantly increase the enclosure temperatures but it only seems to be doing this by about 3 degrees under low temp printing conditions.

    I've done another heating test with the front door panel in place, the printer PID controlled bed @100ºC and the original heated bed leaning up against the inside wall of the container also at 100ºC, and here is a graph showing what is going on.

    Two temperatures are monitored, one is about 100 mm above the printers heated bed (red line), and one is right up at the top of the enclosure (orange line). It looks like it takes about 1/2 hour to get to ~24ºC above ambient. This would be close to useful in summer when the room temp will be around 23-24ºC is but I'm less likely to be doing any higher temperature printing in summer.

    I should add this is without the additional heat from the printer nozzle which would be at ~250ºC and without being too fussy about sealing up some of the remaining gaps etc. I will work on these and expect to be able to get close to a 50ºC inside the enclosure even during winter.


    Screen Shot 2022-08-13 at 6.55.43 am.png

  2. #2
    BobL is offline Member: Blue and white apron brigade
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    7,182

    Default

    Added an inside light, a supposedly 4800 lumen LED ring light (same as I have on my DP).
    To me it looks like about 2000 Lumens.
    Anyway the light puts out a fair bit of heat and adds about 6ºC to the enclosure temperature.
    However, long term it's probably not a good thing to leave it running as it's unlikely the associated light circuitry will like it too much.
    I can probably use it just to help get the enclosure up to temperature quicker and then turn it off.

    Enclosurelam.jpg

  3. #3
    Join Date
    Sep 2012
    Location
    York, North Yorkshire UK
    Posts
    6,439

    Default

    Hi Bob,

    I had a ring light fastened to the mill quill with double sided tape. After a while it fell off. The heat caused the tape to loose its stickiness. So it got replaced by an Ikea flexible clamp lamp. That gets quite hot as well.
    Best Regards:
    Baron J.

  4. #4
    BobL is offline Member: Blue and white apron brigade
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    7,182

    Default

    That ring light comes with 6, 8 x 2mm disc magnets on top.
    I put 3, 18 x 4mm magnets on top of the PC top of the enclosure, and the ring light sticks firm up against the top of the box, it's going nowhere.

    On my DP I used a hose clamp to hold a disc of galv sheet metal around the collar above the quill and and it just sticks to that.
    Ringlightinsitu.JPG

  5. #5
    Join Date
    Jul 2010
    Location
    Melbourne
    Posts
    9,088

    Default

    Given your graph doesn't show the rate of increase slowing down much yet, it looks like you'd get to 50C in another 20 minutes or so.
    How about you put a 2400W fan heater or perhaps a heat gun in there to pre heat it? then see if the setup you have can hold the temp?

  6. #6
    BobL is offline Member: Blue and white apron brigade
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    7,182

    Default

    Quote Originally Posted by Stustoys View Post
    Given your graph doesn't show the rate of increase slowing down much yet, it looks like you'd get to 50C in another 20 minutes or so.
    How about you put a 2400W fan heater or perhaps a heat gun in there to pre heat it? then see if the setup you have can hold the temp?
    I agree about the "not slowing down" and it would eventually get there.
    In some ways it's advisable to (initially) go slowly and it wouldn't worry me too much if it took an hour to get there.
    Not sure about using a high heat source to speed things up, somethings are liable to suck more heat and possibly melt/fry/let out smoke etc.
    I've moved the CO2 fire extinguisher from the front door to the study and installed a fire alarm.
    What I should really do is get all the electronics and the PID controller outside of the enclosure but this is going to requires a complete rebuild of the enclosure.
    Regular steppers may also do funny things when they get hot.

    Usual problems - I'd like to be testing all this but meanwhile I need to print a bunch of stuff and can't do both t the same time. I need another printer

  7. #7
    BobL is offline Member: Blue and white apron brigade
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    7,182

    Default

    I've done some more enclosure sealing and testing, and the enclosure now gets up to ~48ºC in about an hour.
    This was high enough to try out some ABS but after about 20 minute the smell was unbearable.
    As has been discussed venting is not an option as this drops the temperature so odour absorbers are what I'm about to tests.

    Fortunately our friends in the "grow your own" dimension have created a high demand for odour easting filters so have ordered a 2kg filter canister of activated charcoal through which I'm going to recycle the enclosure air. I have a suitable 100mm fan already installed but that's oriented for venting so some enclosure mods will be needed to make it suitable for internal air recycling.

  8. #8
    BobL is offline Member: Blue and white apron brigade
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    7,182

    Default

    I decided to see if I could improve on the heating time of 1hour to get 48ºC. After printing a few pieces of ABS, even although I could not smell anything, I did get a headache afterwards. This suggests something is escaping the filter and enclosure and my poor sense of smell can't pick it up. SWMBO, who has a much better sense of smell has been away so I haven't had a chance for a second opinion.

    All this is not surprising as the enclosure is poorly sealed which would lead to loss of heat and fumes. The enclosure is made from an extruded Al frame and 1mm thick polycabonate (PC) cladding held onto the frame by M3 screws every 150 or so mm along the extrusion. As the enclosure heats up the PC expands and up to mm sized gaps appear between the PC and the frame.

    So I took the whole thing apart and stuck the PC cladding to the frame using double sided foam tape (and then put the screws back). Between the access door frame and enclosure I used single sided foam tape. The enclosure is made in two halves for ease of installation of both the printer and enclosure. The extrusion contact between the the two enclosure halves was also sealed with single sided foam tape and clamped together using small G-claps.

    The whole thing sits on a large cork mat so is not sealed at the bottom but as the heat and fumes would tend to rise in the enclosure I figure sealing the enclosure is the more important. This would also give a chance for the filter/fan to grab the fumes before they escaped through cracks.

    sealedenc1.jpg

    As the whole thing warms up the foil lined black core flute also expands and slightly moves away from the PC cladding leaving gaps for radiant heat to leak out. This is especially a problem around the doorway. To hold the core flute close up against the PC I used pairs of small disc magnets (see green arrows in photo) and this works really well. I could have stuck it down with tape but as soon as I do that, I will need to take it apart

    The smaller access door at the bottom of the main door is for accessing the printer control panel but unfortunately it does not allow for viewing printing progress, It's currently not sealed but that is in progress.

    Sot I needed a viewport higher up that the lower access door to see what was going on while printing. If (like the rest of the enclosure) the core flute panel was on outside this would have been easy - just cut a U shape in the core flute and lift it up/down as required, but for ease of adding and removing the core flute from the door it's installed internally so I cut a U shaped port (see H in photo above) and jerry rigged up a mechanism so I could open and close.

    Here is the viewport partially opened
    The viewport core flute flap is lifted by yanking on a string (see green arrow)
    sealedenc2.jpg

    And here is the Heath Robinson mechanism that closes the viewport from the other side.
    sealedenc3.jpg
    Pulling on wire S opens the hatchway and the spring closes it.
    Its a pretty flimsy arrangement but I will use it till it breaks.
    Just after I finished I realised I could tape some magnets to the outer edge of the inside of the view port and just make up a small foil lined core flute panel that could be attached from the outside with magnets.

    All this took a whole day !! so does it work?
    A quick test showeds the enclosure heats up about 50% quicker than before (35 minutes to get to 48ºC) so it appears to be well worthwhile.
    I haven't tried it with fumes yet.
    SWMBO has returned so I will be able to use her nasal sensitivity as ell.

  9. #9
    BobL is offline Member: Blue and white apron brigade
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    7,182

    Default

    Have done some more testing, found some more leaks and sealed them up.

    Here is the latest temp/time curve.
    Time to 48º is now 22 minutes (albeit from a slightly higher starting temperature).
    Main Printer heater bed (up to 750W) is 105ºC, nozzle at 245ºC
    The heating process is helped by heat from a second heater bed (up to 280W, set at 80ºC, can go to 100ºC if needed) leaning up against the walls of the enclosure.
    Also have the 4000 lumen (100W continuous) LED lamp and fan (20W) on inside the enclosure


    Heating-curve3.jpg

    Problem now is the enclosure temperature going too high, I don't really want it going over 50ºC
    I walked away and left it for a few minutes and when I came back it had reached 54ºC and was showing no signs of slowing down
    Turned off second heater bed, lamp and nozzle at point A on graph
    Fan has to stay on to remove fumes, lamp can go off when not checking printing process.
    At B I turned the nozzle and and lamp back on, at C I tuned the lamp off. and it looks like it the temperature was stabilising.

    I'm pretty happy with this.

  10. #10
    BobL is offline Member: Blue and white apron brigade
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    7,182

    Default

    Printed a couple of ABS pieces yesterday while SWMBO was working in her adjacent sewing/craft room (separated from room with 3D printer by open double doorway). After about an hour of printing she had said nothing so I asked her to come into 3D printer room where she reckoned she could just faintly smell something plasticky but I wonder if it was her imagination - I could not smell anything. However, she said she could not smell it in her sewingt room where she was working. A couple of hours she reckoned she could just smell fumes in that room, SWMBO said it was very faint odour and certainly not offensive. Her sense of smell is so sensitive she gets ill from the fumes of just fuelling her car so I have to do it, and with regards to any smells I might emit she's not shy in telling me if there's a problem

    I was in the 3D printer room for about half the day while it was printing and no headaches for me last night so I reckon I might have got it under control. I might bring the VOC (Volatile Organic Compound) sensor up from the shed and do some measurements later today.

  11. #11
    BobL is offline Member: Blue and white apron brigade
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    7,182

    Default

    I've printed a few more things out of ABS with the heated enclosure and it's all working quite well.

    My son gave me some 30% Carbon Fibre (CF) filled PLA filament for my birthday so I thought I'd give it a crack in an unheated enclosure.

    Below is an example of the CF filled PLA on the left and standard PLA on the right, all printing parameters are the same.

    The advertising blurb indicated the CF filled filament would print with less warping but if anything yhere's a bit more warping on the CF print flange as shown by the red arrows.
    Other than the slight warp, dimensionally the two prints are identical which means I can print prototypes in std PLA and the print using the more expensive filament material

    The recessed flanges were printed with minimal support - might be less warping if more support was provided and the the enclosure is heated.
    COmparison.jpg

    This CF filled filament is a budget CF filled filament material but it still costs 70% more per kg than std PLA.
    This CF filled PLA prints are about 5% less dense than std PLA
    CF is also very abrasive so it needs HSS printer nozzles.

    Improvements in stiffness/strength etc over stD PLA are supposedly only in the 5-10% range so nothing special but I do like the way it looks.
    It's still early days for these composite filament materials.

  12. #12
    BobL is offline Member: Blue and white apron brigade
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    7,182

    Default

    Quote Originally Posted by BobL View Post
    . . . .
    Just after I finished I realised I could tape some magnets to the outer edge of the inside of the view port and just make up a small foil lined core flute panel that could be attached from the outside with magnets.
    The following shows use of pairs of magnets for closing the "view" and "access" ports on my heated 3D printer enclosure.

    The topmost arrow points to a magnet on the outside of the view port re. The magnet is held there by another magnet on the inside - sandwiched between them is the 1mm thin layer of polycarbonate (PC) cladding and 3mm of black core flute.
    To open the view port in the core flute
    - from the outside of the enclosure, slide the magnet pair above the view port are and pull down on a wire on the outside (connected to a core flute flap on the inside) to lift the core flute view port flap up,
    - to close the view port release wire and slide magnet pair down over the view port area.
    Magnets.jpg

    To get to the printer controls the bottom access way is used.
    - again slide magnet pair up.
    - open hinged hatchway in PC and push core flute flap up - use printer controls
    - to close flaps slide magnets down.

    ll this to minimise heat loss from the enclosure.

  13. #13
    BobL is offline Member: Blue and white apron brigade
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    7,182

    Default

    This may amuse someone for ~10s - if you are lucky.

    New way of opening and closing 3D printer enclosure viewing portal.
    Viewing Portal is hatchway cut in the core flute which is on the inside of the Polycarbonate enclosure.
    A magnet on the outside of the Polycarbonate enables the magnet on the inside (attached to that wire) to be dragged upwards to open and downwards to close the portal.

    https://youtube.com/shorts/GkDNyqVTFzM?feature=share

Similar Threads

  1. 3D printer
    By SurfinNev in forum GENERAL DISCUSSION / OFF TOPIC
    Replies: 3
    Last Post: 1st Sep 2021, 08:24 PM
  2. 3D printer
    By China in forum CNC Forum
    Replies: 8
    Last Post: 28th Feb 2017, 10:19 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •