Sometimes supports are very difficult to remove (physically) when I print with ABS. The image below, from Thingiverse - MOF-5 unit cell, is after significant effort to remove the yellow ABS supports from a black ABS model.

Yellow ABS supports removed from black ABS model

I've learned about polyvinylalcohol (PVA) as a 3D-printing filament which is soluble in water. It seems like a great option for dual-extruder printers, where you can print the model with ABS or PLA and the supports with PVA, then throw the whole thing in water and let the supports dissolve.

I've considered other options since PVA is ~4x more expensive than ABS.

Doing a PLA print with ABS supports, followed by dissolution of ABS with acetone, is my best idea currently.

Has anyone found success with another option?

  • $\begingroup$ PLA is not completely immune to acetone. It's unlikely you'll have any success trying to dissolve ABS supports away from a PLA model. $\endgroup$ Commented May 21, 2017 at 8:30
  • $\begingroup$ PLA supports come off a lot easier than ABS supports. Have you tried printing the object with just PLA? $\endgroup$
    – bill_080
    Commented May 21, 2017 at 12:57
  • $\begingroup$ Don't bother with ABS supports for PLA models… the two don't adhere well, and the acetone bath will bleach the PLA as well as staining it with whatever pigment is in your ABS support. I tried a few times and it never ended satisfactorily. See my answer below for a recommendation/alternative to PVA. $\endgroup$ Commented May 22, 2017 at 16:09
  • $\begingroup$ @khaverim Have you fixed this issue? i'm seeing that your supports are too heavy, try to use a less population like 50% or 40% of the infill support, also separation will help to avoid to get adhesio to the walls. $\endgroup$ Commented Aug 9, 2018 at 16:56

5 Answers 5


I've had great success printing with HIPS (high-impact polystyrene) as a support for both PLA and ABS. Most sites recommend it for use with ABS because the materials melt at similar temperatures and work best with heated beds, but I've had good luck using it as a support material with PLA on a bed at 60°C. It doesn't stick as well to PLA as it does to ABS, so supports tend to peel away very readily. The downside is that, if you need the support to anchor your print at all, it doesn’t really stick well enough to accomplish this task. For that, you must pair HIPS with ABS.

When you print with ABS or have complicated interwoven support structures, HIPS can be dissolved with D-limonene, a citrus based cleaner sold under various names like Citrisolv (others exist), or with dipentene (a mixture of L and D-limonene that doesn't smell as pleasant).

Regarding cost: I've found HIPS to be slightly more expensive than PLA/ABS, but only 1.5x the cost, not 4x like PVA. Additionally, it isn't hydroscopic in the same way as PVA so it lasts longer out of the package. Since you're using it as support, you also tend to use far less filament than you do for the main print (sparse support structures as opposed to solid print structures).

Water-soluble alternatives: There are a few proprietary blends of polymers sold by the big commercial printer manufacturers (3DSystems, Stratasys) that only work in their machines… these are generally soluble in basic solutions (water + sodium hydroxide or sodium carbonate). These are usually very expensive and you'd have to rewind the filament on a spool, as they come in cartridges made for specific printers. You'd also have to experiment with the right build conditions and solution blends to remove the material afterward. Airwolf has a support material called Hydrofill that purports to be soluble in plain water… I'm not sure how this is different from standard PVA, though I assume it is different. Hopefully more companies will work on developing water-soluble options to help us keep the 3D printing world full of renewable, less-environmentally-harmful options for filaments (both print and support).


Ultimaker now has a material called Ultimaker Breakaway. After using it for a few models, it works remarkably well, allowing me, for the first time, to print nice rounded surfaces on the bases of my prints. It really does just break away from the surface, much like HIPS but without the lack of adhesion problems between HIPS and PLA.


I haven't tried this myself, but the recommended way I've seen for dealing with the high cost of PVA is to print the support structure in ABS, PLA, or something else cheap, and print only the interface area (the top of the support where it meets the model) in PVA.


I believe this is on-topic because it describes my experience using PVA in a Prusa i3m3-MMU2 machine.

My experience has not been good. I started with a new spool, and dried it between attempts.

The problems may be specific to the Prusa MMU2 printing process, but some are intrinsic in PVA, or at least the PVA filament I've been trying, PrimaSelect PVA+ (link is for reference only. I have no skin in this game.) Aspects that may carry forward to non-MMU2 machines include:

  1. filament is very stringy,
  2. filament is friable, and crumbles into the drive gear,
  3. filament does not adhere to a PEI bed. I had to use Aqua Net hair spray.
  4. filament does not stick well to itself. Stacks of support will have internal defects, which may affect the desired print.
  5. filament is very hygroscopic. With the MANY manual interventions needed with the MMU2, it may help to bake the filament in the middle of a multi-day print.

Perhaps unique to the Prusa i3m3-MMU2 machine, these lead to some specific problems:

  1. the filament may be smaller diameter than 1.75, and it is sometimes not detected by the filament-presence detectors. This not only leads to failures, but it may also lead to large blobs of filament extruded into the purge towers, which leads to "crashes" and compounds the next problem. This may also be because I have too much tension between the drive gears -- something to check.
  2. the purge towers are weak where layers of the PVA overlay each other. Purge towers may break during printing, which ruins the print.
  3. the MMU2 filament drive gears are frequently fouled by PVA filament bits. I haven't found, though, that cleaning them reduces feed problems.
  4. the MMU2 system feeds filament through PTFE tubing. Over time, perhaps with humidity, the friction of the filament passing through the tubes increases, which makes the feed problems worse.

This is not a negative review of the filament itself, or of the Prusa i3m3-MMU2. I may be doing something wrong, and would love to know better. I am quite happy with the MMU2, and have tuned it so that it works very well with PLA filament. This is the result of launching into soluble support, which is a feature I strongly hope to use reliably.

I am still exploring alternatives.

  • $\begingroup$ I'm using 3DMaker PVA, and I am seeing some, not all, of the problems you have with PrimaSelect PVA+. The 3DMaker PVA is pretty stringy too, but it is also very sticky, adheres well to itself as well as the build surface (I have a powder-coated PEI steel sheet on my MK3S+MMU2S). It isn't too hydroscopic -- it takes about 8-12 hours to dissolve in still water. The filament is also flexible and doesn't get ground up by the Bondtech gears -- you may have too much idler screw tension. I have also replaced all my tubes with 3mm inner diameter to reduce friction and reduce the chance of grinding. $\endgroup$ Commented Jun 6, 2020 at 22:50
  • $\begingroup$ I will add that I don't see how PVA could be paired well with ABS, which is what the OP was writing about. My PVA has a max temperature of 230C, while ABS typically prints above 250C, I would think you'd get a lot of cooked PVA fouling the nozzle during filament switching. Has anyone tried this? $\endgroup$ Commented Jun 6, 2020 at 22:55

Not sure if frame challenge is on-topic here, but have you tried using different support options in your slicer? In Cura, I've found the experimental support tree mode both reduces the material usage a lot and gives supports that are easier to remove, at least for geometric (vs "organic", where it's sometimes good and sometimes bad) models. Turning off walls for normal (non-tree) support (infill and roofing only) and/or using a thinner line width than your nozzle can also help with removal but whether it provides sufficient support depends a lot on your model's geometry.


While I haven't used PVA yet, think of it this way, it may be 4x the cost but you use significantly less material for support structures even if you have a lot of support.

  • 2
    $\begingroup$ As this stands, it does not provide an answer to the question "How to make the supports best". Please extend on what you want to say. Mind, this is not a forum. $\endgroup$
    – Trish
    Commented Aug 7, 2018 at 9:02
  • 1
    $\begingroup$ Hi Brian, and welcome to SE.3D Printing! :-) Thank you for editing and fixing the missing word in your answer. As Trish states, Stack Exchange is a quality Q&A site and not a forum. Please take a look at the tour and How to write a good answer. Good answers are based on sound experience and in-depth research. If you could expand upon your answer, with more details, examples, etc. then that would be great. Take a lot at other highly voted answers to get an idea of the usual standard. Thx :-) $\endgroup$
    – Greenonline
    Commented Aug 7, 2018 at 9:27
  • $\begingroup$ @brianh Normally the infill support are 60% less than the infill part, however this percertage is strong, but an infill support of 40% or less can help a lot and easy removable, but only we will save the 60% of that 400%. $\endgroup$ Commented Aug 9, 2018 at 17:02

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .