Why do the corners of my ABS object lift off the bed?



I print ABS on a LulzBot Taz 5 and frequently have issues with the corners of objects lifting off the bed.

My extruder is at 230 °C and the bed is at 90 °C for the first layer and 100 °C for the rest of the layers.

I have experimented with using ABS slurry (ABS + acetone) on the bed for increased adhesion, building a foam enclosure for the printer, and varying the fan speed. I have noticed the problem is more common the taller the parts are and the sharper the corner is.

Adding ABS slurry helped for smaller parts (less than an inch tall) but with my more recent larger parts the adhesion to the bed was so good that the corners of the part lifting actually peeled the PEI tape off of the bed.

I have tried using both a skirt and a brim with no change. The skirt stays on the bed, the brim gets pulled up with the corner.

Kevin Morse

Posted 2016-01-12T20:40:43.047

Reputation: 1 483

Doesn't TAZ 5 have PEI tape? – hroncok – 2016-01-12T21:09:27.377

Correct sorry. I was just using Kapton at work so I had that on my mind. – Kevin Morse – 2016-01-12T21:11:35.653

1What fill percentage are you using? Adjusting the fill percentage of the print should alter the rate of cooling and therefore the stresses generated by the temperature differences as you print. – VirtualMichael – 2016-01-12T21:40:17.690

50% I guess the same could be said about slowing the print down??? – Kevin Morse – 2016-01-12T21:50:14.360

150% is quite high. I often use 10-15% and the print is still very sturdy - even with a single shell. – VirtualMichael – 2016-01-12T22:19:54.617



There are many different approaches to solving this issue and most of the answers already are spot-on. However, the fundamental reason for the "warping" is incorrect and inconsistent temperature across the material.

If there is too much fluctuation in the temperature across the object in this heated state can result in warping. The reason you see this mostly on the build plate is because the temperature of the first few layers of molten plastic vary much more against the build plate than against higher layers. Note that you can see additional warping mid-print using ABS and this can be a result of a draft or sudden drop in ambient temperature.

So, to help solve your problem, here are some suggestions (sorry if there are duplicates):

  • Completely enclose/seal your machine's build area to reduce (or eliminate if possible) draft and prevent the natural heat of the machine from escaping.
  • Increase the temperature on your build plate. I almost exclusively use ABS on my printer and I keep my HBP at about 112C. However, I live in the NW of the US, so my climate is naturally cooler than say Florida.
  • Alternatively, try decreasing the nozzle temperature to a lower point within the ABS melting range. This will just shorten the gap between the inconsistencies in temperature across layers. It is typically better to print at lower temperatures if you can help it. Obviously there are differences in the filament, so you'll have to find that "sweet spot".
  • Ensure your build plate is flat and your tape doesn't have bubbles. Your BP being flat should be a no brainer, but if your kapton tape (or whatever you use) has bubbles, your freshly printed plastic may not be getting the same temperature from the BP as the rest of your part. This is a bad thing as mentioned earlier.
  • Use extra adhesion techniques such as "ABS Glue" (ABS w/ acetone) or even hairspray. This works about 80% of the time for me, but can make it a bit difficult when removing the parts off the BP.


Posted 2016-01-12T20:40:43.047

Reputation: 6 144


Most common problem with corners is to low temperature of heated bed. Set the heated bed temperature to 110°. If this will not help then try to set brim in your slicer.

The problem can be cause by other issues. You can find additional tips in following Troubleshooting Guides:


Posted 2016-01-12T20:40:43.047

Reputation: 1 881


Adding a manually added brim with a larger bulk at the corner extremities that you can cut off after printing should help. My larger prints come off the printer looking like tents with concrete weights tied to their corners.

The 'weights' are attached to the print by very short a 2-3 layer (depending on print size) brim-like strip that makes them easy to remove using side cutters. The 'weights' are usually enough to hold the corners of my target print on the plate in conjuction with a weak slurry.


Posted 2016-01-12T20:40:43.047

Reputation: 181


This is a common problem with ABS. You might prevent it by enclosing the printer inside a box/chamber - that will create a warmer environment and the extruded material will cool down more slowly, hence not creating such a tension.

Other option is to use PLA instead if possible, the problem is not so significant with PLA.


Posted 2016-01-12T20:40:43.047

Reputation: 1 076

and the problem still exists? – hroncok – 2016-01-13T19:46:24.497

I have enclosed the whole printer in a foam "oven". It gets quite warm inside. – Kevin Morse – 2016-01-12T21:10:12.340


The solution I prefer to prevent ABS withdrawal is using Blue tape on the bed, and then spread a thin layer of Polyvinyl acetate (Vinavil glue).

If your printer allows it, you can also print with the bed at 110°C.


Posted 2016-01-12T20:40:43.047

Reputation: 189

As mentioned in the question. I have used ABS slurry to keep the part attached to the Kapton tape on the bed and the pulling up was so strong that it peeled the tape off the bed.

I guess I'm looking more for reasons the part is pulling up in the first place. – Kevin Morse – 2016-01-12T21:06:03.723

1Blue tape is not Kapton tape. It is a masking tape with heat resistance. – WalterV – 2016-01-12T21:09:23.067


To answer the specific question of "why" the corners lift off, it seems it has to do with the thermal expansion (or shrinkage) and surface adhesion area.

The material is deposited hot, and even after it cools is still warm, and as it cools it apparently shrinks in size slightly. Each layer therefore produces a little bit of force, pulling upwards and towards the center of the layer as the layer shrinks. So each layer is actually in a state of slightly tensile stress. Over a few layers this may not cause a noticeable effect, but as more layers are added, the total force pulling on the bottom layer increases.

The reason this affects corners more than the bulk surface regions is that the total surface area holding the part to the bed is smaller at sharper corners, so less force is required there to overcome the bed-adhesion force, causing corners to unstick eventually.

Since this is an artifact of thermal expansion, attempting to maintain a uniform temperature across the print volume/printed part will probably help a lot (for example, by enclosing the printer, as others have said).

In addition to the prior suggestions, the following paper suggests that chamfering (rounding) the corners can help with this (if your design allows for that):

D. D. Hernandez, "Factors Affecting Dimensional Precision of Consumer 3D Printing", Intl. Jnl. Aviation, Aeronautics and Aerospace (2015)

"If the bottom layer does not appropriately adhere to the print bed, the cooling process and material shrinkage in layers above it will tend to pull at the smallest features, with the least surface area in contact with the bed, causing sections of the print to warp. Sharp corners at the bottom of the print pose a particular problem."

Side-thought: I wonder if varying the bed temperature during the print (for example, slowly decreasing it to room-temp as more layers are printed) would help, since a constant bed-temp would theoretically produce a vertical temp-gradient for a very tall part. Don't think I've seen anyone try that (presumably because Cura doesn't automatically insert those g-code commands for us). Wouldn't be too hard to insert custom bed-temp commands throughout the g-code file though, but would take longer to temp-stabilize between layers.


Posted 2016-01-12T20:40:43.047

Reputation: 169


In answer to the question "why do the corners lift when using ABS", the answer has to do with the coefficient of thermal expansion (or contraction).

ABS, compared to PLA, has a higher coefficient meaning that as the temperature of you part changes the more the part will deform. With 3d printing, the material extruded from the nozzle is quite a bit higher than the build chamber. As the layers print and the extruder moves up, the lowest layers are cooler than the layers above, shrinking as time goes on.

For angular parts this is a larger problem than for round parts. With angular parts (parts with sharp edges in X and Y axis) as they cool, the internal stresses induced by cooling increase and can't be resolved in just the xy axis and cannot move into the build plate. The part then lifts up to reduce these stresses.

A brim or raft, is a thin layer that ideally cools at a constant rate through out which will effectively increase the bonding of your part to the build plate. Abs to abs bonding being stronger than abs bonding to whatever your build plate material surface is. For most parts this is sufficient to keep good adhesion while your part builds and cools. Raises the temperature of the build plate also reduced the stress in the brim/raft by reducing the difference in temperature. This is why those are the go-to solutions for stopping abs from lifting.

If your part is large enough or the angles are quite sharp in XY, (think of a five pointed star for example). Then the brim raft won't necessarily be enough to prevent the warping. That's where "helper discs" "mouse ears" etc . Are helpful. These, when dispersed around the outer sharp corners, increase the surface area of the raft and also reduce the angles of the thermal contraction due to cooling.

Note, that while I'm talking about ABS specifically, this will apply to any and all materials if the part geometry, or material properties have the same issues. Nylon for example is also a pain to print for similar reasons.

The final thing to say is that all if this is for nothing if your build plate isn't both flat and level.


Posted 2016-01-12T20:40:43.047

Reputation: 1 745


I would recommend using a raft on your print. A raft is a few layers that are printed on the bed before your print starts.

The larger you make the raft the stronger the center of it will be. The edges may warp, but the inside where your print is will be fine. Using a raft that is mainly comprised of lines instead of solid surfaces would allow a smaller chance of warping as it could only warp with the lines.

enter image description here

Eric Johnson

Posted 2016-01-12T20:40:43.047

Reputation: 2 168


Another approach is to put mouse ears on all the corner parts, that you clip off the print afterward.

See Mouse ears defeat corner curling monster.


Posted 2016-01-12T20:40:43.047

Reputation: 621


I think you answered your question in your statement. Lulzbot and ABS. Lulz does not have an enclosure.

Try using PLA for an open air system. Or build an enclosure. Following you can add glue or hairspray.

But I promise you, with any open air printer, you will face this problem. I only use ABS on my FFCPs. Even with their passive heat chambers I plan to move to PLA for all my printers.


Posted 2016-01-12T20:40:43.047

Reputation: 2 913


This post is meant to share experience with products that make your products stick to the bed. This is not meant to be a promotion of the products! In other answers I've seen answers hinting to Elmer's glue sticks.

I have been printing directly onto the aluminium heatbed from day 2 of the i3 Prusa clone (Anet A8) printer and the glass beds of the Ultimaker 3 Extended and my HyperCube Evolution using a PVA based spray (3DLAC, but there are more products with similar effects, e.g. glue sticks). This sticks so well that PLA and PETG can only be removed after cooling down the bed completely. For ABS you could use DimaFix which can be used for ABS as it increases grip with increasing temperature where 3DLAC looses grip over 80 °C (according to theory). After trying DimaFix on high temperature beds for printing POM filament (very tough to get this to stick as this is bearing material!) I found that prints stick better with 3DLAC on glass.


Posted 2016-01-12T20:40:43.047

Reputation: 25 570