Printing with polycarbonate. I have been having some luck with polycarbonate filament.

gplus
(Gianmario Scotti (Mario)) #1

Printing with polycarbonate.

I have been having some luck with polycarbonate filament. I will share my findings and hope to also receive some helpful suggestions from the community.

First of all, why polycarbonate? Well, it’s one of the toughest polymers - it has the highest impact strength of all extrudable polymers, and a tensile strength equaling nylon.

  • it has the highest glass transition temperature and the highest heat deflection temperature of all extrudable polymers.
    The latter property is the reason I need to print with polycarbonate.

However, the challenge with 3D printing it is also related to its great properties: the hotend needs to be at 270C - you may be able to print at 250C, but the inter-layer adhesion is very poor and it defeats the purpose of choosing polycarbonate. Then there is the fact that polycarbonate goes through considerable volume change during phase transition. This will cause warping similar to ABS printing, except ABS will “stay put” with sufficient adhesion to the printing bed. Polycarbonate is far less docile than ABS: good adhesion is achieved on BuildTAK, but during warping, polycarbonate will rip BuildTAK right off the glass on which it is glued!
Somewhat good adhesion can be achieved with clean glass with a thin layer of dry hairspray. But I could not find any way to avoid warping and partial detachment at corners and other such nucleation sites.
At least with BuildTAK you can print pieces with small footprint (below about 2 cm2). You can print, according to my experience, considerable volume, and even objects with much larger cross section, using this small footprint without detachment.

Before you ask: the bed must be heated to 100C. 120C is better for adhesion, but you are more likely to have the BuildTAK ripped off the glass substrate by the polycarbonate piece, at this high temperature.

(ThantiK) #2

In order to get rid of warping, you’ve GOT to have a heated build chamber with PC, that’s why so few people print with it. This also starts presenting its own problems unless your printer has been built from the ground up with a heated build chamber in mind. If you’ve got motors, etc inside of the chamber at 100C+ then you can damage things. If you’ve managed to design the printer in such a way that it can keep the heat isolated to the areas you want, without screwing up the places you don’t (you’ll likely need a water-cooled hot end, even) then you’re fine. But most people don’t go to these extremes, or believe they need to for any reason.

(Gianmario Scotti (Mario)) #3

@ThantiK That’s a really valuable package of concrete advices. Thank you!

(Ryan Carlyle) #4

I print a fair bit of PC on my R2x. It has watercooling and a very hot chamber. I use UHU gluestick for adhesion, that works pretty well.

(Gianmario Scotti (Mario)) #5

@Ryan_Carlyle can you snap a pic of the watercooled part(s), maybe post with a bit of description??

(Ryan Carlyle) #6

Sure, here’s a google groups thread from a while back: https://groups.google.com/d/msg/3dp-ideas/eoDkSKxYEGk/-fLn4PBuFSkJ

(Ryan Carlyle) #7

That’s watercooling the hot end, and indirectly cooling the extruder stepper via the cooling bar. I also have a peltier cooler on my Y stepper (which is mounted with an ABS injection-molded bracket). The X stepper is mounted on an aluminum carriage with thermal grease so the carriage mass acts as additional heat sinking. The chamber runs about 60C with the HBP at 120C. I would prefer to go to ~120C chamber temp for minimal PC war[omg (based on glass point and creep properties) but I don’t have the guts to run the printer anywhere near that hot!

(Ryan Carlyle) #8

…And yes, PC is a bad-ass material. Horribly difficult to print though. Interestingly, Esun is seriously working on a lower glass point, blended version that is much lower-warp. It performs much closer to ABS than virgin PC though. (In heat resistance and mechanical properties.)

(Gianmario Scotti (Mario)) #9

@Ryan_Carlyle OK, excellent stuff! So basically you are protecting all the stepper motors from overheating.

(Ryan Carlyle) #10

Protecting the few remaining plastic components in my bot from being destroyed by stepper heat, and protecting steppers from overheating.

ABS has reasonable structural strength up to about 75-80C. But stepper motors typically run ~40-50C above ambient temp, depending on the current pumped through them.

(Gianmario Scotti (Mario)) #11

@Ryan_Carlyle the best polymer for those plastic parts is actually… polycarbonate.

(Ryan Carlyle) #12

Agreed :slight_smile: I have that one Y-stepper bracket printed in PC already. Just haven’t bothered swapping it in yet.

(Whosawhatsis) #13

@Ryan_Carlyle have you printed with PCABS at all? I’ve used the Protopasta stuff, and I really like it. It works particularly will when printing diagonally, which avoids most of the warping issues, and it handles overhangs and support release amazingly well.

(Ryan Carlyle) #14

@Whosa_whatsis Have not. When I was gung-ho about buying exotics, all the PC-ABS was too expensive to be worth messing with. If you like it, I’ll check it out.

Have you checked the glass point? What I’m finding is that every single one of the “low warp” filaments people are releasing actually contain glass-point depressing additives. (For example, Esun’s PC goes limp at 120C vs 150C for virgin PC.) Which makes them a lot less interesting.

(Gianmario Scotti (Mario)) #15

@Ryan_Carlyle that is correct. Tg for all those “pro” and “advanced” filaments is pretty low. They may have good mechanical properties at room temp, but at least for me, that’s mostly irrelevant. By the way, heat deformation temperature is somewhat related but usually a more useful parameter. Sadly, its measurement is not very well standardized.

(Gianmario Scotti (Mario)) #16

@Ryan_Carlyle and @Whosa_whatsis the thermal properties of PC-ABS are dismal:
http://usglobalimages.stratasys.com/en/Materials/FDM/PC%20ABS/pc_abs_spec_sheet.pdf
The 455 kPa (66 psi) heat deflection temperature is only 110C! That’s rubbish.

(Ryan Carlyle) #17

@Gianmario_Scotti_Mar I don’t particularly like HDT for this application because the timescale of measurement is too short. It doesn’t capture creep behavior. For example, most nylons creep at 70C or less but have much higher measured HDTs. Vicat-50 is a little better, but not much. But you’re absolutely right, HDT or Vicat is often useful than glass point for load-bearing applications.

(Whosawhatsis) #18

@Ryan_Carlyle The Protopasta stuff isn’t particularly low-warp, just easier to work with than regular PC. Not sure of the glass transition, but it’s definitely higher than regular ABS. I don’t have a heated chamber (or even an enclosure), but I had the best results printing it at 280C onto a sheet of PC heated to 130C (would have tried higher, but that’s the hottest I could get my platform heater to go), using a raft and printing objects tilted up on a corner so that they were building entirely on top of supports. I’d be interested to see what you can do with it in a heated chamber.

(Ryan Carlyle) #19

@Gianmario_Scotti_Mar Dunno if I’d say that’s “dismal” since it’s #2 behind PC, but sure :slight_smile:

What I really want is filled PC composites. GFR-PC or CFR-PC would be amazing.

(Gianmario Scotti (Mario)) #20

@Ryan_Carlyle perhaps you are right that for your application HDT may not be better than Tg, I am not sure. The problem with Tg from a materials science point of view is that it doesn’t always give a good idea of the behavior of the material at a certain temperature. For instance, polypropylene, and especially polyethylene (depending on molecular mass) has a Tg that is below zero - sometimes well below zero C - and yet they behave well at room temperature, or even higher. This is related to the crystalline structure of the polymer (some are only partially amorphous) and the problems that exist with measuring Tg.