ABS Heated Bed Warping

I have been printing some quite thin pieces (4 layers thick) and the corners curl up about 0.5mm during the print. I have my heated bed at 85deg C (measured with IR thermometer)

I have tried adding a brim but it does not really do anything. Any other suggestions?

Try 110 deg on glass with Kapton tape cleaned with acetone.

I have a solid 8" piece of kaptop. I should have 110deg measured at the bed? I am also extruding at 220deg.

I'll share my settings:

J-Head extruding 3mm ABS @ 240 deg
Heated bed with 3mm ceramic glass @ 110deg
Kapton tape on glass - cleaned prior to heating with acetone to remove grease etc.

Ensure you're not cooling the print with a fan or draft. If your print's lifting at 0.5mm, It's just not sticking at all and sounds like what would happen if you printed on a cold bed.

yea I am having problems keeping be bed over 90deg while actually printing. I am going to put some insulation under it tomorrow and try the print again. Also will try clean my kapton off really well.

I will let you know how it goes, thanks!

@uMinded

Have you tried UHU Glue stick on glass? Right now I am printing 3mm ABS at 220C with a J Head onto glass with UHU glue stick applied to it. However, I am using about 125C bed temp for the first layer and 110C for the rest of the print (though this method would probably work with lower temps too). I get virtually no warp and the parts pop off easily once the bed cools to about 60C. Originally I was using an 8" square thick Kapton sheet on 3mm glass but I found that ABS would lift the Kapton right off the glass as it cooled. Also, the parts would stick so well to Kapton that I would end up destroying the Kapton tape trying to get the finished part off.

There are several advantages to using UHU Glue stick on glass. First of all, it is really easy to apply. When the glass is at ambient temperature apply at least 2 even layers of UHU glue stick to the surface. Try to make sure that the layers are spread on evenly as this will ensure that the entire part gets enough adhesion (the only time I get warping is when I miss a spot on the glass). Heat the bed as hot as it will go for the first layer and try to maintain the temperature at 110C for the rest of the print. Once the part is done let the bed cool down slowly. DO NOT put it in front of an AC or in the freezer. The first time I tried this I cooled the glass bed in front of an AC and shards of glass broke off with the part. Thermal expansion/contraction of glass is negligible while that of plastic is relatively significant. While the part cools, it contracts (shrinks) and tries to pull the glass with it. If it cools too quickly, the force of the contracting plastic is large enough to rip shards of glass out of your sheet. If you let the part cool slowly, the force of contraction will not be enough to damage the glass and the part will just pop off the bed. I find that by about 60C, the parts have almost lifted off the bed and can VERY easily be removed. The reason parts are so difficult to remove from Kapton is because the Kapton actually tends to contract with the part as it cools, so the adhesion is maintained. The disadvantage of using glue stick is that you don't get the clear glassy bottom that is characteristic of parts printed on Kapton. Once the parts are removed, the glue on the bottom can be rubbed off with a wet cloth (UHU glue sticks are water soluble!) to reveal a flat (but not glossy) surface.

I have also carefully engineered a heated build chamber that consists of a clear plastic garbage bag wrapped around my prusa I2 This is probably the easiest way to significantly improve your ABS prints as all you need is a garbage bag. With my MK2 heat bed at 110C, the "chamber" temperature is over 40C measured at the top of the bag (it would be much higher in the print volume). This really helps combat warping and also blocks all drafts that can negatively affect the consistency of your perimeters.

Even if you can't get your bed temperature above 100C, the glue-stick-on-glass trick along with the "garbage chamber" should really improve your ABS prints.

Happy Printing!

Eric

Thanks for the massive reply!

I was planning on getting a bunch of clear storage totes this weekend any ways so I will look for a 16" cube one and that way I would have easy top access.

I am printing on borosilicate glass (pyrex) and I find it does not transfer the heat nearly as well as tempered glass but it is nearly indestructible even when rapidly cooled. When you guys say 125deg and 110deg are you talking the actual measured temperature on the glass surface or what you have the bed set at in software? I have a MK2 heated bed and with my 12V/30A PSU and directly connected I can just barely reach 90deg on the glass (110deg printer setting)

I will try the glue stick method tonight after I insulate the bottom of the bed.

125C and 110C are the approximate temps for the surface of the glass. I have my thermistor rather inaccurately placed underneath my MK2 bed. Some time ago I tried to determine the temp difference between my thermistor readings and the glass surface temp and I measured about 10C difference with a thermistor taped to the surface of the glass when the bed was stationary and my "garbage chamber" was in effect. So the thermistor readings for the first layer and subsequent layers are 135C and 120C respectively.

As for your heating issues,make sure you have a stable PSU. 12V 30A is super overkill so it may be that your PSU has low consistency. I had some problems with cheap Chinese PSUs that couldn't get above 110C (thermistor reading) and would drop to 80C during printing! I recently replaced it with a quality ATX power supply and now I can get to about 135C (thermistor reading) bed temp within my "garbage chamber", though this takes ages . However, I now place a square piece of cardboard over my bed while it is heating and it can easily reach 135C! Cardboard is a great insulator and is the easiest way to speed up bed heating.

Also, I used to use 4mm borosilicate glass but I did not get better results. Moreover, I found that while borosilicate glass does not crack as easily under thermal stress, it does seem to bend much more easily than float glass The bed of my Prusa is very bent and the borosilicate glass seemed to conform to the shape of my bed. By the time I scrapped it I could only print in an area of about 50mm x 50mm! On the whole, I think 2mm to 3mm float glass is the way to go.

Since my prusa MK2 heat bed is about 3mm above my MDF top print plate I don't lose much heat (MDF is a great insulator). However, I would recommend that you insulate the bottom of your bed the same way Nophead does. He basically takes a sheet of cardboard and covers it is aluminium foil and places it below the MK2 bed. The aluminium foil really helps with heat loss through radiation as it reflects the radiated heat back up at the bed. This is why wrapping food in aluminium foil is such an effective way to keep it warm!

Let us know how things work out.

Eric

Voltage drop and thin wires can also be a problem. Ensure that your wires driving your heatbed aren't unnecessarily long and are of adequate gague/thickness.

You could also drive the heatbed with a relay and not directly from the driver board. Some driver boards can't provide adequate current for high heat.

Yeah, that's definitely something to consider. Also, not all Prusa PCB heat beds have the same resistance. It can vary a little which can lead to some heat beds not getting enough power and thus not getting hot enough.

Eric

i am using a piece of cork board under my heater, seems to be working well so far.

UPDATE:

I have insulated under my heated bed and used RP Iron Man's tip on a piece of cardboard on top while heating and I can get to 110 bed temp in <5 min and sustain it all print now.

I also rubbed a glue stick on the kapton tape and the print stuck really well. After the bed cooled to 50deg the part popped off by itself.

Now I just need to get extrusion temperatures and things right. I printed a pocket tactics figure at 0.1mm and overhangs/bridges droop to much and the spot layer changes happens is deformed. I am thinking that means that 220deg is to hot for my ABS. It's bedtime now but I will print a calibration cube tomorrow and fine tune temperatures.

Anybody here use Cura? Under "Packing Density" is says 1.00 for PLA and 0.85 for ABS. What is this setting for?



As you can see the base and the top portion printed pretty clean but the middle is an inverted cone shape and the layers kind of "spill over"

Edited 1 time(s). Last edit at 07/18/2013 12:18AM by uMinded.

Have you calibrated your Esteps yet?

I calibrated my E steps for PLA at 0.25mm layers. If the two filaments are the same relative diameter the layer thickness should not matter. Thats what I was wondering about "Packing Density" as if its like Slic3rs "Extrusion Multiplier" that makes me think that ABS changes shape when heated much more so than PLA.

If my E steps will change between the two filaments then I think it might be a good idea to change to volumetric extrusion instead. That way I just need to know brand x and color y need 0.93x extrusion multipliers.

Good to see that the glue stick trick is working for you. I have never used Cura so I can't help you out with that, but looking at your print, your problem may be that you are printing to object too quickly without allowing it to cool sufficiently between layers. You'll notice that if you try to print small ABS objects without some sort of cooling you will end up with a deformed object. The glass transition temperature of ABS is about 105C while that of PLA is about 60C, which is why ABS warps so much more than PLA. Even if ABS prints do not lift off the print plate, they can still deform if the previous layer is not given sufficient time to cool before the next one is put down. This is why I like to fill my print area with parts to get the best result from ABS prints. The downside is that you have to wait hours for the print to finish...

If you plan on cooling your ABS prints, there are a few thing to be aware of. If you cool the part too quickly, you'll get bad layer adhesion and the finished part will easily crack along the layers. Also, having a fan pointed at your part really kills your heat bed temperature, though using a fan duct would help prevent this from happening.

Eric

OK, some progress:

Clogged nozzles suck...

I have done several test prints and have found that reducing the extrusion multiplier by 15% makes the layers stay in the places their meant to be. I did not know that ABS swells so much when hot. Now I have a strange problem of things imploding somehow...



That is suppose to be a 20x20x10mm calibration cube at 20% fill. You can see that the middle layers have shrank a considerable amount but it happens quite a while after that layer is printed. You can also see it happens closer to the heated bed. Another issue is that with 0.1mm layers the top did not print every well. The thread was so thin it broke over the bridge. Any idea how to deal with that?

I am going to try lowering the bed temperature to 105deg and putting a 70mm computer case fan blowing on the print to see if that makes a difference.

you know i don't think i have ever seen that one before.

I have had a similar issue.. though not as bad.



I have done two things to overcome it. I redesigned my hotend to be more pointy, in my first design, the heater block was only around 2mm from the print, I raised that to ~7mm and added isotation around the hotend.

Also I lowered the bed temperature after the first layer from 110 to 95 degrees, this really made a difference.

The theory behind these two changes is that ABS needs to cool down before the next layer is laid down, nut too much, but just the right ammount. But when the print is sandwiched between the heated bed and the hotend, this doesnn't happen, until the gab is big enough, therfore the print is deformed due to too much heat, until the gab gets big enough.

I'm printing in a heated build chamber, with temperatures around 40 degrees, and I have a fan blowing around my hotend, to get a bit of circulation around the print, the fan is running at the slowest possible speed just before it stalls. But if you don't have a build chamber, a fan shoundn't be needed for ABS, I'm not even sure that I really need it, but it works for me.

I would also suggest starting out with bigger layer heights, maybe 0.25mm until you have the basics sorted out and is able to print dimensionally stable objects.

I had similar issues with the heated bed, warping and curling.
I overcame them all by changing the heated bed to mains powered and using a DC-DC solidstate relay connected to the Printrboard bed thermistor. The solidstate relay switched the mains power to the new heater pad and heats up very quickly, just 30 seconds to reach bed temperature.
The thermistor is built-in to the heated pad and the pad has adhesive to attach it to the underside of my aluminium build bed that has a 3mm glass top.
This removes the need for a heavy duty power supply and doesn't overload the Printrboard either. Should work with all other boards too.

I changed over to standard float glass 3mm thick with no tape. For PLA I use PVA glue 30% mixed with metho 70% and wipe over all the glass surface lightly. If too thick the part is hard to remove, just let it cool down and if it hasn't popped off, soak in water for a few minutes. This doesn't harm the PLA at all, I tested by soaking a print for well over a week without any bad effects.
For ABS I use ABS juice very lightly applied.

That happens to me when I print small ABS parts too quickly. Due to the high temperature of the heat bed and the speed at which you are building up layers, approximately the first 5mm of the print is at the same temperature and cools at the same time. Even if the part does not lift off the print plate, this leads to a warped section of the print about 5mm from the bottom.

Lowering your bed temp will help. Try lowering it as much as you can so long as the part does not start lifting off the surface. Having a large computer fan blowing indiscriminately at the print could be detrimental to your bed temp readings and you'll probably get warping as your bed temp will drop super fast. If you try this, make sure that the fan is not on during the first layer as this would make adhesion to the print surface very difficult. It would be best if you used a small fan with a fan duct directing air close to the nozzle. Excessive cooling will also lead to poor layer adhesion and the part will crack easily along its layers.

Also, what nozzle orifice diameter does your hot end have? It can be very difficult to print 0.1mm layers if your nozzle orifice is larger than 0.4mm.


Eric

Try another brand of ABS.

I think I have it nailed!


Bigger Picture

I got that print by lowering the bed temperature to 100deg, adding a 10 loop brim, a 20 second minimum time spent on each layer, 85% extrusion multiplier, and having a fan blowing on the print the entire time.

I am going to pick up a clear storage tote when they go on sale and put the whole printer inside. Have a false bottom for all the electronics and then I can move the printer around.

Thanks for all the advice!

Edited 1 time(s). Last edit at 07/20/2013 02:33PM by uMinded.

Hi Minded,

It's a silly observation, but are you 100% sure that your ABS isn't PLA?

good thought, try the burn test.

Definitely ABS, melts in acetone... kind of... It melts but does not mix into a slurry...

It is from a new Canadian supplier and not being sold commercially yet so maybe the formula is a bit funky...

probably a different butadiene percentage content?

[en.wikipedia.org]

I used to used some cheap Chinese ABS that produced similarly poor prints and did not fully dissolve in acetone. Now I am using some real ABS from Eckertech and I am getting much better print quality. The quality of your plastic makes a huge difference.

Eric

Congrats on your successes. I'm sure you'll find other ways to keep moving forward. On that topic, here's a suggestion: If you need to set your extrusion multiplier to 85%, your esteps value is definitely not calibrated. For me, anyway, getting the esteps right was the prereq to good bridging.

I found that on anything smaller than an inch squared required 85% extrusion multiplier but anything larger printed at 85% will have gaps between infill and perimeters. Large objects print great at 100% so I figure my E steps are correct and its probably a similar issue to my heat contraction.