Vapor Treating 3D printed ABS parts to make them waterproof


The problem with FDM (Fused Deposition Modeling) 3D printing (most open source printers, and Stratasys) is that with the layer upon layer printing, you get a thinly-stratified surface finish which makes the parts structually weaker, and enables liquids to ingress into the part.

I came across this method on the rep-rap blog. It gives FDM 3D printed ABS parts a nice annealed surface finish which is almost injection-molding like. I plan on getting a 3D printer soon so I can test making parts with this method, and pressure testing them to see if they can be truly water-proof. In my experience with FDM printers (Ultimaker, Replicator), printing with 100% infill with a bit of tweaking with the extrusion rate can make fully infilled solid plastic parts with no bubbles (as far as I could see).

Has anyone tried this method with 3D printed parts for underwater use? I imagine this could be really useful for making waterproof liquid compensated enclosures and the like.


I like this idea! I printed the main chassis for mine (see attached photos). My main concern is that, unless we can do the whole thing in one vapor "jar," a lot of sanding will be necessary to make the parts fit together. (I had to file the corners off all of the tabs on mine as is.) On the other hand, the Hysol might be more than enough to complete the seal.

I'm working from a high school with quite possibly a hundred hot plates lying about in the various labs. I'll try to get the parts together and test this, probably by Friday.


I have tried this technique on some printed parts with limited success. Below are some of my experiences with the method:

1) Attaching to the geometry to hold it in the vapor was difficult and sometimes messed with the surface finish at the location of the attachment-melt into the part

2) I found that the part tolerances went out the window very quickly. It was hard to gauge amount of immersion time in vapor from part to part. It seems like when the part started to get smooth it happened really rapidly.

3) I did manage to completely remove all the ridges from the FDM process but then mating parts would not fit together

4) I was not comfortable with doing this, from a safety standpoint, so I bailed.

Another method, a little safer, might be to use sealants. I came across a .pdf document where sealants were tested against each other. Click here to see .pdf file.


I printed the ping ball holders on an FDM3000 printer that I own. I didn"t realize that they wouldl absorb water. I could try to print something else?



Nice paper! Wow the sealants enable the FDM parts to be fluid pressure proof up to 40 psi. of the sealants are readily available, like the Minwax Oil Base Polyurethane. Liquid comp FDM parts seem like a plausible reality


I just purchased a vacuum chamber for making silicon molds. Looks like I might have to do some experimentation with infiltrating FDM parts!


Could you try soaking one for an hour, then drying off the surface and checking for mass gain? I used a Dimension industrial printer and will be trying that same experiment. If someone with a MakerBot or similar DIY-oriented printer could try this as well, we can generate some quick data as to whether printed parts absorb a significant amount or not.


exactly what we use for our rov systems works well


DO NOT DO.. Vapor method indoors very dangerous , outside only we use pressure cookers work very well, you can pick up cheap steamers cookers from thrift stores


Thanks for the heads up. I am going to use my school's laboratory fume hood just to be on the even safer side.


I would but don't have any way of weighing it. Do you know of a scale that is accurate and not to expensive?


I looked up kitchen scales on amazon and found tthis scale: 0.1 gram resolution, 1kg scale

We use this 2kg x 1g resolution scale in the OpenROV lab, and it works fine for most things


Cool good to know! Do you guys have any issues with the FDM plastic at high pressures? I'm wondering if there are any microbubbles of trapped air that result in structural weaknesses.


Russell, you use ping pong balls? Or the same scale?


Still haven't tried the vapor coating, but I did try soaking an untreated part overnight and measuring mass change. In my sample size of one, there was approximately a 7% gain in mass.


A couple days ago I 3D printed the entire electronics tube end caps as a single unit. Unfortunately, the roughness of edges resulting from having to scrape off all the support material in the grooves meant that the o-rings didn't provide enough of a seal. Consequently, I returned to this thread to consider the process, but in some outside research discovered another method that may be just as effective (and certainly much simpler and safer):