How does a heated bed stop ABS from warping

As above, what is the mechanism by which it stops warping? Does it keep the ABS more pliable or simply improve the adhesion?

Mike

To a first order, it doesn't. The thing that stops abs from warping is a higher ambient temperature so that the plastic doesn't shrink so much after cooling through glass temperature, hence heated enclosures. However, for a given amount of warp, increased bed adhesion (however you get it) reduces the chances of the part coming off the bed. And for parts that aren't too high, warm air from the bed may help maintain a higher temperature close to the print - provided there isn't much in the way of drafts and the bed isn't whizzing around in Y.

My view of it is that as the first layer goes down, it is tied to the bed and can't shrink. As subsequent layers go down further from the first layer, they do shrink. The bed heated to the glass transition temp. of the plastic allows the lower layers to relax and form an orderly transition from the unshrunk layers closer to the bed to the shrunken layers further from the bed. If the heated bed were only about adhesion, then the minute you popped the part off the bed, the bottom would still warp up. That said, you still need pretty good adhesion because there are still internal stresses trying to pull the lower layers up as they relax into the geometry dictated by the upper layers.

Edited 2 time(s). Last edit at 01/27/2017 10:10PM by LoboCNC.

That is much as I had imagined it. Having some clues that an ABS print can adhere well even to cold build stage, it seems that if I were designing a new printer a vacuum build stage and a heated enclosure would be more important than a heated build stage. A vacuum build stage without a heater is easy.

Mike

Quote
LoboCNC
The bed heated to the glass transition temp. of the plastic allows the lower layers to relax and form an orderly transition from the unshrunk layers closer to the bed to the shrunken layers further from the bed.

That seems like a very reasonable idea. Given that abs isn't a very good thermal conductor I wasn't sure what sort of distances the temperature would fall off over, so I printed a quick tower (20x10x50mm) and taped on some probes (badly - it was harder than I expected). I set the bed at 110C which in the current config with the polyester sheet you sent me gives a surface temperature of about 90C. I also ran the part cooling fans at a low level, with the print head up at the top of the printer. My thinking was that I wanted to model the movement of the bed during printing, which I assume disturbs the air around the print - but it's arguable if running the fans is a good model of that or not. I set them to a level where the bed could still maintain the set temperature comfortably.

Here's the print wired up for measurement:



Here's the status screen showing the bed holding temp at 85% of full power:



And here's the temperature readout, with T1 closest to the bed and T4 nearest the top:



To check the channel to channel variation and get the ambient temp I took a measurement with the probes hanging in free air



So, it looks like you are right, a heated bed will give you a region of gradually reducing temperatures that extends for several centimeters, and it seems reasonable that such a transition zone would help spread the stress caused by shrinkage. There are a couple of factors that might influence the measurement - the use of the fans may have been a bit heavy handed, and printers with non-moving beds probably have a bigger transition zone than this experiment suggests. On the other hand, all the kapton tape I used to hold the probes in place would have acted as insulation and made the transition zone bigger than it would otherwise have been. The probes are reasonably reproducible channel to channel but are sensitive to how well they contact the surface, so given the bad tape job I'd probably put a plus/minus 5C error on those readings.

Nicely done experiment! While you've got your setup going, what about doing a sample print with holes in it at various heights and then insert the thermistors into the holes. This would give you the internal temperature of the print vs. the surface temperature of this experiment. (And also make it easier to keep the thermistors in place

I could imagine there might be a case to test abs onto a cold bed in a heated chamber. But I'd suspect the chamber would have to be 80 deg C like a Stratasys. What can abs stick to at 80 deg C.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
LoboCNC
what about doing a sample print with holes in it at various heights and then insert the thermistors into the holes.

Yes, I should have done that in the first place, thought I could cut corners I'll give it a go...

...ok, that was quite interesting. With the probes inside the model I was able to insert each one during the print. I waited for a couple of layers above each hole before putting the probe in, and the highest temp I saw was 130 something C, so the initial cool down from the extrusion temperature (245C in this case) happens very quickly. But the subsequent cool down takes a long time, when the print finished the 2nd to 4th probes were still in a gradient getting hotter towards the top. I left it to sit for a while (no fans this time) and it got to the following gradient after about 1/2 hour when things were more normal (i.e. T1 is nearest the bed)



The holes are at 5mm intervals, starting at +5 through to 20mm of Z, with 25mm total print height. The depth of the part is 20mm and the holes penetrate 19mm, so the measurement is on the inside of the two perimeters. (stl on [www.thingiverse.com] )



I set the bed a little hotter during this print after realising the surface was only at 90C in the previous test, but then forgot to measure what the new surface temperature came out to. *oops*



Edited 2 time(s). Last edit at 01/29/2017 08:34AM by JamesK.

Quote
DjDemonD
I could imagine there might be a case to test abs onto a cold bed in a heated chamber. But I'd suspect the chamber would have to be 80 deg C like a Stratasys. What can abs stick to at 80 deg C.

One possibility is Kapton tape upside down - held to a vacuum plate with its adhesive face upward.

As "sticky side up" materials like the plastic film from document laminating pouches comes close to holding an ABS print at room temperature I wonder if 80°C would be needed, perhaps 60°C or even lower??

Mike

It would be an interesting test, I'd have to figure out how to heat my chamber without the heated bed being the heat source.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

I will try at ambient with upside down Kapton tape next week, possibly also with my only version of a heated chamber - consisting of a big plastic bag over the whole of my Delta printer I can get up to nearly 40°C in summer. I expect even the hot will fail; but if it is a near thing then it may be worth making a new Delta designed as an enclosed model.

Mike