SoliForum - 3D Printing Community — automated build platform idea

Re: automated build platform idea

2013-10-25T20:26:29Z

Making it work on the entire bed would be a design struggle for the reasons Hazer mentioned. Yeah, you can crank enough power to get a delta across the device, but you need to move that heat someplace else in a time frame that is relevant. This becomes even more of an issue if your parts require an enclosure to run within acceptable tolerances (this is for production after all - or what would be the point).

You already are comfortable moving the parts on glass, so why not build a cassette system that has multiple, quick change, smaller glass pieces on the bed. The machine can stay running for consistency, and you just need to be there to pick up a finished piece and drop in a replacement glass whilst the other part is running.

Downside: You have to show up every few hours. Which is good for quality control and keeps you out of trouble.

Re: automated build platform idea

2013-10-25T07:27:41Z

Yeah! So this could conceivably work. But I like my bed at 86, 87 c sometimes, and the cheap kind won't like that over the long run (low service life at 70-80C and unrated performance at 80-90C, quick failures in 90-100C-from my experience in other applications).

My desire to try testing peltier modules was limited because I just have a magnetic glass bed and quickly pop it off and throw it in the freezer, and heat up my next glass pane. :-).

There are some people who used one for a chocolate printer, but that's a unique need.

Re: automated build platform idea

2013-10-25T01:56:58Z

elmoret wrote:

Hazer wrote:
Except a peltier cooler does not have enough power to cool an entire bed.
Huh?
http://www.allelectronics.com/make-a-st … LER/1.html
(you can get them cheaper than that link, but that link has a datasheet)
80w (80J/s) heat pumped @0dT (actually in this application for purposes of the performance chart we have negative dT)
Aluminum = .91kJ/kg-K
SD3 bed = 125 grams
if 80C above ambient, that's 9.3kJ stored energy. It would take just under two minutes to move that much energy - plus you'd be losing heat to atmosphere.
A peltier is actually perfect for the application thermodynamically, as it can be reversed electrically to output more energy in the form of heat than it draws electrically, and because it can as shown above, cool the bed. The only downside is physical in nature - they don't make 200mm x 200mm peltiers.
Reality check: If a 80w heated bed can get it to 100C in a few minutes, a 80w heat pump can do the same (in a vacuum) or better (in air).
Engineering.

Ah.

You have it all figured out.

I leave it to you then.

But dont take my word for it. Google practical uses.

Re: automated build platform idea

2013-10-25T01:15:54Z

Hazer wrote:

Except a peltier cooler does not have enough power to cool an entire bed.

Huh?

http://www.allelectronics.com/make-a-st … LER/1.html

(you can get them cheaper than that link, but that link has a datasheet)

80w (80J/s) heat pumped @0dT (actually in this application for purposes of the performance chart we have negative dT)
Aluminum = .91kJ/kg-K
SD3 bed = 125 grams

if 80C above ambient, that's 9.3kJ stored energy. It would take just under two minutes to move that much energy - plus you'd be losing heat to atmosphere.

A peltier is actually perfect for the application thermodynamically, as it can be reversed electrically to output more energy in the form of heat than it draws electrically, and because it can as shown above, cool the bed. The only downside is physical in nature - they don't make 200mm x 200mm peltiers.

Reality check: If a 80w heated bed can get it to 100C in a few minutes, a 80w heat pump can do the same (in a vacuum) or better (in air).

Engineering.

Re: automated build platform idea

2013-10-25T00:17:11Z

Except a peltier cooler does not have enough power to cool an entire bed. Also, if you stick that under the bad, the standard heater would not work. And they dont really cool anything unless you control the opposite side. All they really do is create a temperature difference. If you control the temperature of one side of the cooler with something that has more power (like a watercooling system that can bleed the heat away faster than the cold side can heat up). What I am saying is, you would need either large fans+heatsink for the cold side to get cold. Or a watercooling system. And even then, you will also have condensation problems. There is a good reason these never really were useful in PC cooling systems.

Now, if you reverse voltage on the cooler, you can create heat. So, if you installed a bunch of these you could heat the bed with them, and then reverse the voltage to make them cool down. But again, you would need something else to cool the opposite side.

But dont take my word for it. Google practical uses.

Re: automated build platform idea

2013-10-24T20:48:34Z

When I've printed on the past on kapton over aluminium bed, the parts still stuck after cooling. On glass, however, it definitely works how you describe.

automated build platform idea

2013-10-24T17:22:45Z

I saw a discussion on hackaday about this and thought I would see what this community thought about the feasibility of this idea: installing a thermoelectric cooler (aka a TEC or Peltier cooler) underneath the build platform. When the print is done a Gcode command turns on the cooler, quickly chilling the build platform and part. Would the difference in the platform and parts thermal coefficient cause the part to break free? If so then then gcode could be used to push the part off the platform.

The thermal electric cooler could even be used as a supplemental platform heater by reversing the voltage applied to it.

Thoughts?