Heat bed warping



  • Greetings all!

    I am not a metallurgist at all, but I do understand that metal, especially aluminum will warp a bit when heated and cooled.

    With that, I have a CoreXY with an aluminum 300mm x 300mm heated bed that is 3mm thick.
    I have a 3mm sheet of glass that rests on top that I use as my build surface.
    Recently I have been on a campaign to tweak my machine to squeeze out a bit more quality out of my prints to try and get the outside perimeter walls as seamless as possible. One thing I am noticing is that when printing a big model, that is several hours to print, the first few hours I am getting the outside finish that I desire, but then as the print progresses over the next couple of hours, I am seeing inconsistencies in the layers on the Z axis. Some thin and some bulging. I chocked it up at first to the filament diameter changing through the roll. But I now notice there is a gap that appears between the glass build plate and the heat bed on one corner, that I am positive is not there before I start printing, as I make sure all 4 corners are seated properly after re-installing after a cleaning it so to get an accurate measurement when I create a Mesh Grid Height Map.

    I know the terrain of the bed changes from when it's cold to when it's up to print temperature, but should it flex and change shape over the course of a print, even though the temp stays rock solid throughout it all?
    The gap that develops gets to almost a millimeter at times, and that is obviously what I am seeing on my prints.
    It's worth noting that I have the heat bed secured in all 4 corners only. I never have a need to heat the bed more than 50C.

    I am thinking about redesigning the build plate and doing away with the aluminum bed altogether.
    My idea is coming off some money and investing in one of those rubber heating mats and attach it straight to the glass.

    Does anyone have any thoughts on the bed warping and/or my new bed design? Or maybe another/better option?

    Thanks in advance.



  • @scottbg1 said in Heat bed warping:

    rubber heating mats and attach it straight to the glass.

    That's a bad idea. Don't do that. Glass is a bad thermal conductor, so that heating pad will be very hot in the areas where the heating wires are, and stay cool where they aren't. That can lead to some very bad hot spots and generally uneven heating. Aluminum is a very good thermal conductor and makes for a great heat spreader.

    To clarify, aluminum doesn't warp when it gets hot, it expands. When that expansion is constrained, it can warp, because if the even expansion has nowhere to go, it will bow out.

    3mm aluminum is pretty thin and won't be able to resist much bending force. So if you are planning on improving your print bed, I would suggest getting a thicker 6mm aluminum plate. Preferably a cast and milled MIC-6 plate so that it's very flat and rigid. Then mount it at 3 points instead of 4 so that it's easier to level. You can attach the silicone heater pad directly to the aluminum, and then the glass can sit on top if you wish.

    Finally, how do you have your bed heater setup currently? Is it PID tuned, or is it running in bang-bang mode? If it's in bang-bang mode it can lead to the PCB heater flexing as the current is applied and removed constantly. Doing a PID tune of the bed is highly recommended.

    Also, how long are you letting the bed sit at full temp before homing and running your mesh grid? It may not be fully heat saturated before you start.



  • Hi Phaedrux. I was hoping you'd chime in. 🙂

    Thanks for the info on the heating pad. I will look into finding a thicker aluminum plate.

    My bed and Hotend are PID tuned.

    Coincidentally, the bed was at priting temperature well over an hour before I began the print where I discovered the warping. I had installed a new nozzle in my hotend to prepare for the long print I mentioned. I started the heating process as I began breaking the hotend down and putting it back together, and then it was still hot while I re-calibrated my nozzle height and ran the Mesh Bed Calibration. Add in my girlfriend demanding my undivided attention for several minutes while she vented about an inept co-worker who is ruining her life, and it certainly was at temperature quite some time before printing.



  • @scottbg1 said in Heat bed warping:

    I am thinking about redesigning the build plate and doing away with the aluminum bed altogether.

    You may want to view this video from 3:45 timemark https://youtu.be/8C4jtj0OuE4 . It's a voron design but general good heatbed information.

    I think I have similar bed to yours. 300mmx300mm, 4 screws, thin and PCB like heater traces. and results are consistent throughout long prints (I use magnetic bed cover with 1mm PEI rather than glass).



  • @zapta I like the thermal fuses they use. The bolt hole makes it way more convenient than the ones you have to silicone to the plate.

    https://www.aliexpress.com/item/32820046480.html



  • Heaters on glass are a bad idea, as has been stated. Here's an example of what happens when you do that- it's Taz printer that came with a heater mounted on a glass bed plate. Notice the 30C temperature variation across the bed.

    alt text

    Here's the same heater mounted on a piece of 1/4" cast aluminum tooling plate:

    alt text

    There are a couple problems with screwing a thermal fuse to the plate. If the screw comes out, the thermal fuse will hang in the air by its wires and won't do any good. If the heater comes off the plate, and the TCO is bolted to the plate, it won't do any good. It's better to mount the TCO on the heater using silicone.

    Here's what happens when the adhesive starts to let go of the bed plate.

    alt text

    Keenovo's instructions for mounting the heaters say you should put a bead of silicone around the edge of the heater (which I didn't do, never having seen those instructions until recently). Alternatively, get a heater without adhesive and mount it using silicone (which I did when I replaced the burnt heater).

    Finally, I would not trust no-name, no-spec parts purchased via ali-express to protect against a fire. Quality parts with full specs and multiple safety certifications only cost about $1 each.



  • @mrehorstdmd said in Heat bed warping:

    There are a couple problems with screwing a thermal fuse to the plate. If the screw comes out, the thermal fuse will hang in the air by its wires and won't do any good. If the heater comes off the plate, and the TCO is bolted to the plate, it won't do any good. It's better to mount the TCO on the heater using silicone.

    Finally, I would not trust no-name, no-spec parts purchased via ali-express to protect against a fire. Quality parts with full specs and multiple safety certifications only cost about $1 each.

    Good points.


Log in to reply