100% Warp-Free MakerBot 3D Printing

I love MakerBot and I love Bre, and I think he and his team have done an amazing service to the world.

Recently, I purchased five MakerBots (actually Thing-O-Matics, factory-assembled and tested), which was one for each our current TechShop locations (Menlo Park CA, Raleigh Durham NC, San Francisco CA, San Jose CA, and Detroit MI).  I also picked up a Replicator for myself.  Research, you know.  ;)

After becoming very frustrated with ABS plastic objects warping when the bottom corners would pull loose from the build platform, I started searching for a new solution.  No matter what I tried, including using raft versus no raft, increasing the temperature of the HBP (heated build platform), cleaning the Kapton tape with acetone, putting shields around the openings in the MakerBot's housing to keep the heat in, increasing and decreasing the nozzle temperature, nothing would prevent the object or even the raft from popping loose and lifting up on the corners.  This happened with all six of the MakerBot variants that I had access to, including Replicator.

Previous solutions that have been created by others and adopted as supposed solutions included the HBP (heated build platform) which comes standard on all late-model Thing-O-Matics and Replicators, Kapton tape on the build platform, using a glass sheet, creating a vacuum table, using painters tape and other tapes, and countless other ideas.  But none of these seem to work.

So I thought of my own solution.

My solution has proven to be cheap, fast, simple, and 100% dependable.

Would you like to know what it is?  Then read on to the next step!

(I'll try to take better pictures the next time I print a nice large object.  Hopefully you can see enough from these photos to be able to follow these steps.)

Step 1: Put ABS Scraps and Acetone in a Jar

The first step is to make a thin slurry of acetone and ABS plastic scraps (raft and support, and rejected objects you've printed).  We all generate lots of scraps on our MakerBots.  You can probably make a similar slurry with PLA and acetone, but I haven't tried it yet.  Plus, PLA doesn't seem to have as bad of a problem with lifting and curing and warping as ABS does.  I personally find PLA to be too brittle for my projects, since I usually make structural objects.  Let the ABS scraps and acetone sit in the covered jar for several hours, and the ABS pieces should completely dissolve.  You want to have a very thin watery solution.

Comments

[olawlor]

Jim, the acetone + ABS trick definitely works well on my Printrbot+ to glue prints down to the heated print bed's Kapton tape. This slurry is more commonly called "ABS juice".

But even with the juice holding the base flat, I find the warping forces on very large prints (6 inches wide or more) can actually crack the print between layers when I'm printing in my cold garage. I use an insulated box to keep the build area at about 70C, which seems to be about the only way I can get crack-free prints.

[icache]

olawlor - rather than a box over the printer, you might try something Nophead pioneered. Print a wall around your object. I use a 1mm thick, 25mm high wall 5mm from the object. This holds the heat in and prevents drafts. For tall object, if you need a taller wall, you'd need to either go with a thicker wall or add a beam half way down each wall, otherwise gravity caused the 1mm wall to bulge after 25mm high.

[icache]

First of all, if Kapton tape requires doctoring or "sanding", it's probably the wrong tape to use. I use blue painters tape with the bed set at 115C. At that temp painter's tape is slightly tacky and the biggest problem will be getting some of it off your object, not adhering to it. To get it off the object a hot soapy water soak and a nylon pan scrubber works. Creating 2mm rounded outside edges with Fredo6's Rounded Cornors (Sketchup Plugin) creates a way to get a Painter's 6in1 Tool under the object to pop it off, usually cleanly.

But, Bigal21 has a point about stresses in the flow directions. However, there is a built-in software solution - honeycomb infill pattern (Slic3r has it). When printing a cell only the cell walls that have not already been created by adjacent cells are printed. That eliminates linear tensions between rows. Honeycomb is strong laterally, that's why bees use it, and a 20% fill is stronger than a 40% rectilinear. But, more important, honeycomb creates the anti-draft effect used by Nophead, but does it internally. Each layer of a honeycomb cell is printed on top of the lower layer. Notice that that is not true of the rectilinear hash pattern where two lines on layer C are spaced between the corresponding two lines of layer A & E. Honeycomb therefore constructs convection columns which allow heat to rise but protects the area from the effects of draft caused by fans, etc. I use a direct drive filament driver with a fan - going to honeycomb eliminated all warping in objects smaller than about 100mmX100mm.

For larger objects I also add a wall around the object (Nophead) - 1mm thick, 5mm away from object, and 25mm high. That creates a "hot box" effect that prevents the effects of draft and convection cooling. (Such a wall will collapse due to gravity if taller than 25mm, but below that is where warping would happen.) Combining that with honeycomb I've eliminated warping out to 175mmX175mm objects.

[dylanpil]

i have discovered something pretty cool with rafts which i still haven't seen in any post for 3d printers. it kinda does what this article achieves but in a cleaner manner.
here goes.
1. in your raft, make 0 base layers (no layer at all for base!)
2. ensure interference layer is set to 1 layer ONLY.
3. reduce feed rate to 0.5 or less and flow to 0.234 or less.
4. lower interface layer thickness over layer thickness to around 0.2 (for approx 0.3mm carve height)

what you achieve is a very thin interference layer ( 1 piece) that connects to the heated platform. this platform is a brilliant raft as the print speed is very fast and finishes very quickly. the pieces stick to it so nicely, and it sticks to the heated bed so much better than any raft i have made.
your print then only has a fraction extra on it that can be sanded off in a couple of secounds. the piece also adheres better than any other raft i have made. less material is also used.
PLEASE try but you may need to fine tune your measurements

[missmt]

Thank you. I use this around the edges of my prints, because until I upgrade further, my heated print bed lifts in the middle by about .08mm. I find an extremely thin smear applied when the bed is cold lasts 5-6 prints. I tried applying hot, but had instant acne bubble where it boiled!!!

[Bigal21]

I have identified the real reason why the edges of the models are warping.   Just because healing/cooling has a definite cause-and-effect on the warping does not mean that it is the ultimate source of the frustration. All of these solutions are simply masking the root cause! 

To prove it, try printing a circle (or a gear) without any filling (just a thin circumference).  I can guarantee there will be ZERO warping if the nozzle builds up the Z thickness by traveling around the perimeter, versus filling up a bed of plastic with rows/columns (or a honeycomb).

What happens is that the nozzle goes back and forth, squeezing out plastic the entire time.   When the nozzle returns, it yanks back the plastic which was recently laid.   It is just gooey enough to allow a bias, pulling the freshly secreted plastic in the last direction that the nozzle traveled.

If the nozzle goes back and forth, why doesn’t the yanking cancel out?  Because when the nozzle returns it yanks back BOTH strokes.  The “coming stroke” is yanked back less than the “leaving stroke” (due to cooling and other factors), but a bias exists nonetheless.  The bias creates a tension towards the center. 

To exacerbate the “warping effect”
---------------------------------
*  Let cool air come in through the windows of your MakerBot.  By cooling the plastic off quicker, the freshly laid plastic is less “moldable”.  Hot and gooey plastic will not contain any tension as it finally cools.

* Put less space between the nozzle and the build plate when calibrating.  With the nozzle closer to the freshly laid plastic, it digs in deeper and yanks back more plastic.   Part of the reason that the warping is so bad is that once curling starts, it puts less space between the bed and the nozzle.  That in turn causes more curling in a sort-of feedback loop.

* Print faster.  The reason why “print slower” is on the list of solutions is that it gives the freshly laid plastic some time to solidify, resisting the “yanking effect”.  This is in contradiction with letting “cool air” come in.  That is why neither tips offer a great solution!  There is a balance between cooling sufficiently to completely solidify (impervious to yanking) versus cooling enough to allow yanking (with tension).

* Print with old dirty Kapton tape.  The reason why the “acetone / ABS slurry trick” helps is because it anchors the bottom layer.  With a sticky first layer, it is able to combat the tension a little better.  Again, this is just a method to fight the symptoms without eliminating the cause. 

Bad Side Effects
-------------------------
The number one way to clog your nozzle is to let it get too close to the bed (or the last layer printed).  The stepper’s teeth strip out the plastic filament when the nozzle is “blocked off”.  So, if your model starts warping, it often has the unwanted side effect of digging a pit into the filament which will ruin the print and then require maintenance on the machine.

Possible Solution
-----------------------------------
Modify G-Code so that the extruder lays down plastic in one direction.  When the nozzle returns it is necessary to lower the Z-axis a tiny bit so that it will safely clear the return path.   Unfortunately this adds to the build time.  It is just one solution, so hopefully someone out there will think of a better method now that the source of the problem has been identified.

[ruben.gutierrez]

This did work for me... a little too well, as you warned it might. I think I used too much of the slurry. Thank you, Jim.

[mortinus]

Thank you sir! You have cured a headache that has been plaguing me for far too long.
Old record: 2 Successfuls (Let's say perfect), 15+ Fails
New: 100% Success Rate, ZERO peel/curl!
I have to spread the word...this has been such a horrific problem.

[SlimJim]

Do you have to remove the dried slurry before the next print?

[TechShopJim]

No...you can just wipe more slurry onto the bed with a paper towel and print the next object.

Martin told me that a certain type of aerosol hairspray works really well, but I forget what kind...I will try to find it in my notes and post it.

[lime3D]

When will the new Makerbot be running at the Detroit location?

[lime3D]

Nope. There has been one siting there for over a month, with a sign on it that says it is not in service.

[TechShopJim]

I'll check on this and let you know...thanks!

[TechShopJim]

I think it already is...I know we got a bunch of Replicators for all our locations.

Thanks!

[clgonsal]

Possible typo: where you said "nothing would make the object or even the raft pop loose and lift up on the corners", don't you mean nothing would stop them from popping loose, etc.?

[TechShopJim]

Hi Clgonsal...

Fixed! Good catch...thanks!

[Fredrick Hagemeister]

This has been a pretty good solution although many people find that adding additional fan cooling by the extruder hot end helps. Anyhow, I have a lot of different colors of ABS so I bought a bunch of the 8oz. nail polish remover bottles (I make sure to get the colorless, 100% acetone variety).

[datenkrieger]

this works fine with my printrbot with heated glass bed too, even without kapton tape. Small prints started to come off the heat bed when about 5mm high after i had installed a fan for cooling. Now they stay in place. Also the first layer sticks perfectly now. This helped me a lot, thx

[Enlightx]

im a repraper so never used a replicator but every heated build bed has the same problem.

the solution i posted on the rerap boards is very simple and silly cheap

just get some PVA glue water it down to a milky type solution.

brush onto the heated bed it will dry in around 30 secs (with the bed on)

print and nothing will lift. then leave to cool and it will crack off easy.

to remove PVA just use some glass cleaner and will wipe clean.

[bfk]

I'm going to try this too. PVA glue is cheap and easy to work with. Thanks for the tip.

[bfk]

Thank you, thank you, thank you. Makerbot sent me over to your instructable, so you know they're interested as well.

I'm in the middle of an 8 hour print at the moment, but as soon as this job's finished, your slurry will have been sitting in a jar on my bench, waiting to be painted onto the platform.

Kudos.

Did I remember to thank you?

[PlanetaryGear]

Having MUCH better results with this than anything else. If your rafts or prints are curling this is the way to go! It seems that the PLA will stick to abs on the platform pretty good too, at least I’ve been printing with both and keeping a film of ABS on the platform and the PLA is performing just fine too.

The only problem I”ve had is with really big prints. Still struggling with the mackerbot remote controlled car, the front print I did last night stuck so well to the kapton via this ABS glue that it actually lifted the kapton tape off the platform! Thats dedication ;) Probably happened when the AC kicked in and blew cold air at it, next step is some draft dodgers for the sides...

But this is definitely the way to go!

[Edgar]

Nice!
Put an intro to this on my Blog:
http://faz-voce-mesmo.blogspot.pt/2012/07/papel-metal-e-instructables.html

[Edgar]

This the second very good Instructable I see today, congrats!
You can always make an object with a small slot or two, where you can insert some kind ot tool, that will apply some leverage on the print...

[danielroca]

Very good idea I'll definitely try it out in my next print.

[Light_Lab]

PLA dissolves in chlorinated solvents like chloroform. Slower drying but should still do the job. When working with any organic solvent use good ventilation and avoid breathing the fumes.

[rusty0101]

Not sure if this would help or not, but you might try 'painting' with the q-tip, through a screen onto the bed so that instead of having large surfaces of the abs to adhere to, you have lots of small 'lands' with more space between to allow for easier separation later. If you have several build platforms, you could set up a simple silk-screen process to apply the dilute abs with, and have a consistent build platform as well.

[megaduty]

Here is an idea for easier removal from the print bed; turn an air duster can upside down and lightly spray it on the print bed/printed object. The rapid temp change sometimes makes the item "pop" off the print bed with minimum effort.

[jensa]

Good idea for those that have an aluminium bed. I don't think it would be a good idea for my glass-bed though ;-)

[d-briggs]

GREAT!! The "KISS" principle still works, even with new technology! Kudos for the simple idea!

[smessud]

Thanks for the tip.
I will try it on my (old) cupcake.

It would be nice to have a flexible and easily removable surface to build on.
Kapton with quick release and without glue would be nice.

[Lectric Wizard]

Thanks for the tip ! Nice to see something useful, instead of the latest object to glue your USB drive into ... I'm just about to get into 3D printing so any info is great. CHEERS!