Layer Adhesion Problems with ABS-PC and Polycarbonate

As new and ever more exciting printing filaments get released, I just had to try out Polycarbonate and ABS-PC for high temperature applications.
However, there is one problem: Neither ABS-PC nor pure Polycarbonate exhibits good layer adhesion for some reason in my tests.

My setup:
-E3D hotend v.4 for 1.75mm filament; thermistor values checked with thermocouple and multimeter to be mostly dead-on with occasionally 1-2°C positive deviation
-ABS-PC [http://reprapworld.com/?products_details&products_id=443&cPath=1590_1663]purchased from RepRapWorld[/url], judging from the label of the spool, 3Dthermoplastics.com is the manufacturer
-Polycarbonate bought from Ebay (sorry guys, I got the last two sample packs)
Both filaments come pretty much fresh out of their packs with desiccant.
-Skeinforge 50

As I tried Polycarbonate first and got frustrated after some tries, I switched to ABS-PC and subsequently tested it more.
When I couldn't get the layers to stick to each other at 260°C, I first suspected a miscalibrated thermistor, but after checking temperature settings with a multimeter + thermocouple, this doesn't seem to be the source of my troubles.
Next, I increased the temperature up to the current limit of my hot end, 300°C, without success. Altering layer height and extrusion width (tried 0.3 height + 0.4 width / 0.1 height + 0.3 width / 0.2 height + 0.4 width) also yielded very unstable objects.
To give the hot end more time to fuse the plastic together, I reduced the feed rate until I hit 10mm/sec. No change whatsoever in layer adhesion.
Both materials show the same symptoms - layers that can even be peeled apart by pure finger strength. Some prints already started flaking during the print, due to thermal shrinking stress. Although I temporarily suspected the pigments in the ABS-PC filament for causing this phenomenon, the pure Polycarbonate is uncolored, so this is not very likely.

Did anyone else have trouble printing with PC and/or ABS-PC? I know that PC is indeed possible to print; RichRap demonstrated it on his blog, so I am guessing that I might be doing something wrong.



As you can see, the test objects are pretty much pulled apart completely...I didn't finish most prints, though as I could see problems mid-print.



The layers all flaked apart like this after light prodding and pulling. Sometimes, I tried to separate layers with a knife. Cut through the object like brittle sandstone.



The layers themselves look pretty well fused together, though upon further testing, the filament tracks also fall apart easily.

Edited 3 time(s). Last edit at 11/20/2013 05:17PM by uGen.

I had similarly disappointing results with polycarbonate from the same ebay seller. Things printed just fine, but layer adhesion was poor. Shortly thereafter I got some Taulman 645, so I put the PC on the shelf for the time being.

I tested the dried filament with no cooling fan up to about 300C, with no substantial difference in adhesion. I was thinking about testing the PC again, but I have no real idea about what to tweak other than bump temperature even higher.

Have you tried printing in a heated chamber? You can use an oven bag or cardboard to build a temporary one.

---

WWW.ZATOPA.COM - Your Place for high quality 3D Printing Filament and accessories

Have you tried printing in a heated chamber? You can use an oven bag or cardboard to build a temporary one.

---

WWW.ZATOPA.COM - Your Place for high quality 3D Printing Filament and accessories

ive got some of the ultimachine pc, and some of the prints came out fantastic, the plastic has to be freshly dried for it to bond well, basically it needs 4 hours in a 250f oven or more, then you need to print with it right away, the moisture causes it to become foggy and bond very poorly.

@karlo: ah, what a bummer! Hope that drying the filament in my box with a more potent desiccant helps.

@jzatopa: yes, my printer is permanently installed inside a heat insulated box. The inside reaches temperatures around 40-45°C

@aduy: I didn't notice any fogging or sputtering like with wet taulman filament. Basically, I took it out of the bag and printed with it right away. Maybe the desiccant was not strong enough. The ABS-PC on the other hand came out quite matte. Don't know if that is because of the ABS or because of fogging being partially masked by pigments and ABS. I will try to dry the filament in my drying box...

40-45°C probably won't be sufficient for a heated chamber. Stratasys uses a 110°C chamber for PC-ABS and 135°C for PC.

---

[haveblue.org]

From how I understand the issue, layer adhesion should be affected mainly by the temperature of the nozzle as it welds together the previous with the current layer...the heated chamber is more of an aid against the object deforming. But thank you for your insight, in case I get stuff to work, I try to bump temperatures up a notch.

What layer heights are you guys printing at?

I noticed a substantial increase in inter-layer adhesion with nylon and PC when I went to .35-.5mm layers with my .6mm nozzle.

I have tried 0.3, 0.2 and 0.1mm layer heights with a 0.4mm nozzle. Good that you mention this - AFAIK, Taulman recommends printing the 645 nylon with layer height == nozzle width, so maybe other plastics behave similarly.

Maybe a 1mm nozzle?

---

_______________________________________
Waitaki 3D Printer

The chamber temperature will affect the bond strength just as much as the nozzle temperature. The bond strength of a weld will be proportional to how long the interface of the two materials stays above the melt point of the material. The interface temperature will initially be the average of the nozzle temperature and the temperature of the previous layer. The chamber keeps the previous layer hotter.

---

[www.hydraraptor.blogspot.com]

I was also thinking that bond strength is affected by how thoroughly the interface is molten. So I decreased the feed rate to 10mm/sec and layer height to 0.1mm in order to really melt everything together with the nozzle at 250°C alone. Didn't work. In fact, I couldn't notice much difference from my other test settings.

I've been printing extensively with trimmer line for the past 4-5 months (somewhere in the neighborhood of 25lbs of filament), and I've tried .25, .4, and .6mm nozzles.

The .25 nozzle at .08, .10, .15 layer heights all look fantastic, but they have the strength of PLA. Easy delam

.4 nozzle at .2, .25, .325 layer heights only produced strong prints at .325 layers.

.6 nozzle at .4, .45 and .5 layer heights all produce virtually indestructible parts.

I'm using a RostockMax with an E3D end. 1.65mm filament

Print temp 265C, bed 75C with glue stick

Speeds between 25-80 depending on the nozzle

Anecdotal evidence suggests that there's a significant correlation between layer height and width and layer adhesion.,

@ cope413

What Nylon trimmer line manufacturer are you using, and where are you sourcing it?
Are you drying it?

desert extrusion - .065" diameter

Little tough to find - I cleaned out everything they had on Amazon - but I'm sure it's out there.

Yes. drying is a must. The results are amazing. I love the stuff.

does anyone know the max temp you can print pc abs with? With pc i had good results printing 30mm/s at 280c when it was dried for 8 hours, when i sped it up I noticed the pc came out cloudy and not fully melted. keep in mind that i have a water cooled hot end so the transition zone is cold and only the nozzle is hot. an e3d hot end might be better for this stuff since it is all metal. with their new water cooled metal hot end you could actually heat the water cooling block to start transfering heat to the filament before it melts.

I also used a .6mm nozzle for my tests and they all came out quite well, even a 0.05mm layer height. although that one shattered layer for some reason.

oh i almost forgot, i got pc to stick to cold abs juice on blue painters tape, the warp was massive though, now that i have a hot air blower setup ill give it a try, i have to finish my super grip extruder first though.

Edited 1 time(s). Last edit at 11/22/2013 06:30PM by aduy.

Have printed PC from Ultimachine at 300C on Kapton+PC-Juice with a 130C bed. E3D Hotend obviously.

All the prints I have made are incredibly strong with absolutely no chance of delaminating by hand. The strongest items I've ever printed in fact, by a long way.

I've some of the RRW ABS-PC on the way, so will test with that too and see how I go.

---

I Design/Sell all-metal hotends. My company is called e3d-online - you can buy at [www.e3d-online.com]

I would guess that the max temp. depends a lot on the specific resin that's being used. Stratasys PC-ABS is SABIC Cycoloy C1110HF and is run through a 320°C nozzle. Their PC is Lexan HF1110, also run at 320°C.

---

[haveblue.org]

Yes, Matweb gives a range from 250-330°C as processing temperature for PC-ABS depending on the blend. However, the thermal decomposition temperature of ABS is around 260-275°C-ish, again depending on the blend. I tried to extrude at 250°C because I assumed thermal decomposition might have played a role in bad layer adhesion, but that was not the case after all.
I am drying the filament right now and will try to print something tomorrow. Another thing I might try is cleaning the filament. Who knows what will happen to layer adhesion if the outside of the filament is slightly oily? Just a wild guess because I didn't notice any oil.

uGen: Keep us posted. I suspect that there is something fishy with the Chinese PC filament. At least it was cheap...

The thing is that the PC-ABS I got from RepRapWorld is "USA branded" according to them. "USA branded" filaments cost more than their decent standard filament. I haven't tried their more expensive ABS filament, only the PC-ABS. Upon further investigation, it turns out that the manufacturer is 3Dthermoplastics, which seems to be in some way affiliated with inDimension³.

To be honest, the filament diameter was way off in some places - down to 1.5mm and maximum over 1.75mm. Some segments were rather elliptical...other than that, the color is one of the best I have seen in filament. Even after printing, the vivid blue stayed the same.

Anyway, I will try both the PC from China and the PC-ABS again tomorrow after thoroughly drying the filament for one week. I may also wipe it down with ethanol to remove any grease or oil that might be on the surface just to make sure.

Update: Drying the PC-ABS in my drying chamber didn't yield any effects at all. I didn't try to print the pure PC after seeing the PC-ABS fail again. Instead, I asked the instructor of our plastics workshop if he had any idea what could prevent the layers from adhering to each other. He offered me two possible scenarios:
1. The print temperature is too low - I was printing at a measured 300°C at times, so either the hot end temperature doesn't translate into actual filament temperature that easily or this stuff has to get even hotter. Keep in mind, Matweb states the range of processing temperatures of PC-ABS from as low as 250°C to as high as 330°C. So maybe I got something that requires an even higher temperature. The same goes with the pure PC.
2. There is something in the filament that gases out when getting hot and thus prevents it from sticking. I was thinking either a thin coat of oil or grease on the surface (removable by rubbing alcohol / isopropanol), so I will try that out the next time I print something. OR my drying chamber didn't get the filament dry enough. So I put some meters of filament in his curing oven. Let's hope that all the other stuff like resins etc. in there don't affect the filament further
Anyway, I hope to have some new results either Monday or Tuesday next week.

Actually, I found this out yesterday already, but didn't really have time to post it yet: We are one step closer to successfully printing PC-ABS and PC (other than that from Ultimachine, of which there are multiple accounts of success already). Currently, I am reading the book "Generative Fertigungsverfahren: Rapid Prototyping - Rapid Tooling - Rapid Manufacturing" (3rd edition) in which pretty much every additive manufacturing process is explained. The chapters about FDM turned out to be especially interesting and the key to hopefully solving the layer adhesion problem. Turns out that you need a lot of pressure from the nozzle and extruded filament to fuse the new layer to the older ones. Filament tracks have to be 3.5 to 5 times the layer height in width according to this source (which is contrary to most RepRap practices of extrusion width/layer height ratios from 1.34 to 3 or so).
After a first test with layer height of 0.1mm and 0.4mm extrusion width @300°C, the layers indeed stuck together better than before. Unfortunately, I was busy with other printing stuff that is more urgent, so consecutive tests with even wider extrusions will follow some time later. Stay tuned!

Quote
nophead
The chamber temperature will affect the bond strength just as much as the nozzle temperature. The bond strength of a weld will be proportional to how long the interface of the two materials stays above the melt point of the material. The interface temperature will initially be the average of the nozzle temperature and the temperature of the previous layer. The chamber keeps the previous layer hotter.

Crazy idea,

Just like we use a fan pointed at the nozzle on PLA while printing,

What would it take to have a miniature heat gun pointing at the nozzle when printing ABS or other high temp plastics to heat the previous layer?

my setup on my printer has a space heaters output running through a tube which is then pointed at the nozzle, however the output is stationary and the nozzle moves.

i wonder if you could use a high powered laser pointed at the spot right below the nozzle if that would allow it to heat the plastic enough.

I have bought this little device some time ago, but never really implemented it (intended for a heated build chamber). Lasers are a little bit dangerous in my opinion - reflecting off the heated bed's surface and whatnot.
The linked PTC heater is designed to work with fans blowing across it, so a ducted affair strapped to the carriage might be feasible. Sort of a miniature version of your space heater, aduy.

yeah i thought of that, but then i thought of the laser and liked it better, although i have almost no experience with high powered lasers.

the issue with the small duct is it would have to metal, and fairly flexible unless the head is stationary. they do make some of these though, and you would ideally need to insulate the outside of it. I measured the output of the hot air at 98c on low feeding from a 21c room.

now another way to do it without a heavy tube would be to have a second heater cartidge that heats a heat exchanger block, then a small blower fan with high pressure blows into it.

at the end of it you would use a nozzle that would control the width of the hot stream of air.

the second option seems a lot more feaseable because it uses already existing components to work.

If the hot air "only" reaches 98°C, you might not even need a metal duct. After watching this interesting video, I (sloppily) tested the mechanical properties at elevated temperatures of the taulman 645 nylon that I am trying out now. A test piece placed on a heated bed at 110°C (won't go higher) was a little bit more bendable than a cold one, but it returned to original shape after I stopped applying forces to it. Even with pliers, the nylon was rather difficult to bend after a certain degree (the test piece is the same I posted in my opening post; bent to about 45°).
Oh, by ducted, I actually meant a little duct to direct the heat of the PTC heater towards the print, much like the cooling fan ducts that are common nowadays. The PTC heater, by the way is rather small at 27x17x90mm, but I can also imagine using NiChrome wire coiled around some stainless steel screws like in a hair dryer. This would omit the heat exchanger block that I and many others here might not easily be able to machine.

Hey guys, I have good news! After numerous failed prints, I tried following settings:
0.1mm layer height
0.6mm extrusion width (I have read that commercial machines have extrusion widths of 3.5-5 times the layer height to ensure that sufficient pressure is being applied for optimum layer adhesion)
10mm/sec for perimeters and infill
300°C nozzle temperature
100°C bed temperature
Material: PC-ABS, dried thoroughly first in a curing oven, then in my desiccant chamber.
Basically, what I wanted was a hot and slow print to fuse the layers together as good as possible.
I also flattened my usual test object from 7.5mm to 2.5mm height so that the bed heat reaches all the way to the top of the object. In addition to that, I wiped down a length of filament with rubbing alcohol (ethanol).
These measures somehow either in concert or singularly seemed to have done the trick.

I will evaluate which factor exactly was the deciding one, but preliminary destruction tests showed that it is much more difficult to split the object along the layers.
As I don't have my camera with me in the workshop, I will post pictures and some more info when I get home.