tuna55
tuna55 MegaDork
7/13/20 12:49 p.m.

I print a bunch of ABS stuff at work for prototyping, valves and other things which would not be wise for me to discuss in great detail. We flow water. The valves are around 100 psi max for what I need to learn. The strength is there (well other than the threads) but they leak everywhere. I've tried this stuff: https://www.amazon.com/Smooth-XTC-3D-Performance-Print-Coating/dp/B00PFXK4JY and will try carnuba wax next, but it doesn't give me a lot of hope. The epoxy worked OK, but remains tacky and delaminates. I don't need them to be completely sealed at 100 psi, but I am losing 2-3 GPM from leaking at around 50 psi on the last one. What's the hot ticket here? Some of this needs to maintain a relatively tight geometry, so I can't cake something on because I am trying to coat the inside of a valve which may have more than one chamber.

Mr_Asa
Mr_Asa Dork
7/13/20 1:02 p.m.

Pretty much all filament types are hygroscopic.  As they absorb moisture they're going to change geometries.  Not only will this mess with the tolerances you have but it will cause whatever you seal them with to swell and contract which will cause the sealant to do all kinds of funky stuff.

Any possibility of doing a mold and casting the valves out of something that will resist moisture and will seal better?  May be worth looking into a professional source for the printing once you have the basics sorted out.

tuna55
tuna55 MegaDork
7/13/20 1:31 p.m.

In reply to Mr_Asa :

I'm not really worried about a gradual absorption of moisture, I am worried that I pressurize a closed cavity and instantly water springs from 45 different places and goes everywhere.

 

I only need the things to work for a few minutes to get some data.

RevRico
RevRico PowerDork
7/13/20 1:35 p.m.

ABS... Smooth it with acetone on a rag to seal all the layers together?

That's been recommended for home prints, not sure if you can get away with it in an industrial setting

RacetruckRon
RacetruckRon Dork
7/13/20 1:44 p.m.

ABS is kind of hates to adhere layer to layer from the get go.  Upping your nozzle temp can help with that.  Like RevRico said Acetone smoothing works pretty well at sealing the layers together either with a spray bottle and a rag or with an ultrasonic mist vapor bath.  That's worth a shot with ABS otherwise you might want to try out stronger, higher temp filaments like PC and Nylon.

tuna55
tuna55 MegaDork
7/13/20 1:50 p.m.

I may try the acetone, but to be clear, the printer is amazing. The prints look amazing. I have never operated a 3D printer that can do what this one can do. There are zero visual defects between layers, but turn on the pumps, and oh, look, there are 45 leak points.

bgkast (Forum Supporter)
bgkast (Forum Supporter) PowerDork
7/13/20 2:09 p.m.

I agree with trying acetone. Look up"vapor smoothing". It's typically done for looks, bit might help seal them up too.

akamcfly
akamcfly Dork
7/13/20 3:43 p.m.

What about post-print baking? AKA annealing. I've heard of it being done to improve strength of a printed part, but it might help with the layer adhesion? I haven't done any reading on it it as I just print silly stuff with PLA, but maybe it'll do something for your parts.

Honsch
Honsch New Reader
7/13/20 6:07 p.m.

Since you're printing with ABS, you could always try brushing a layer of ABS cement on. It's MEK, acetone, and ABS resin in a liquid available from your local hardware store.

tester (Forum Supporter)
tester (Forum Supporter) Reader
7/13/20 6:08 p.m.

In reply to tuna55 :

How about vapour deposition of epoxy and cure in a vacuum to pull the epoxy into the part? You might have to clean up the critical surfaces of the part. It would not be the first time that epoxy paint held pressure in something with porosity and thin walls. 

tuna55
tuna55 MegaDork
7/15/20 9:59 a.m.
Honsch said:

Since you're printing with ABS, you could always try brushing a layer of ABS cement on. It's MEK, acetone, and ABS resin in a liquid available from your local hardware store.

Will try this

tuna55
tuna55 MegaDork
7/15/20 10:00 a.m.
akamcfly said:

What about post-print baking? AKA annealing. I've heard of it being done to improve strength of a printed part, but it might help with the layer adhesion? I haven't done any reading on it it as I just print silly stuff with PLA, but maybe it'll do something for your parts.

Hmm, interesting, but it sounds risky. The chamber is already quite heated when it is printed.

tuna55
tuna55 MegaDork
7/15/20 10:00 a.m.
bgkast (Forum Supporter) said:

I agree with trying acetone. Look up"vapor smoothing". It's typically done for looks, bit might help seal them up too.

Will try this as well

tuna55
tuna55 MegaDork
7/15/20 10:01 a.m.
tester (Forum Supporter) said:

In reply to tuna55 :

How about vapour deposition of epoxy and cure in a vacuum to pull the epoxy into the part? You might have to clean up the critical surfaces of the part. It would not be the first time that epoxy paint held pressure in something with porosity and thin walls. 

How would I apply it? Also, I don't have a vacuum chamber at the ready, but it's super intriguing.

 

 

tester (Forum Supporter)
tester (Forum Supporter) Reader
7/15/20 5:00 p.m.

In reply to tuna55 :

Honestly, I think spraying with a paint gun would work. Heck dipping them might work. Then throw it in a vacuum to cure up. 

The other thing rattling around in the back of my head is static  charge painting. I am using the wrong terminology, but the idea was to improve the affinity of the coating.   It was used on plastic interior trim panels to improve paint coverage. 

Oh, last thought, electrolysis might work.  


 

 

wawazat
wawazat Dork
7/16/20 6:37 a.m.

Chrome plating works well on injection molded ABS.  Pre-plating process chemically  removes the butadiene segments (the B in ABS) on the surface which gives the plating a mechanical bond.  

Bonus points for shiny and Mad Max references.  

tuna55
tuna55 MegaDork
8/5/20 10:25 a.m.

I am trying a wide variety of things for testing later this week. I have one part epoxied, one part coated with carnuba wax, one run with acetone, and several with ABS glue. I will report back. So far here are my notes:

 

Wax: Super easy to apply. It looks thin and I don't think it will help very much.

Epoxy: Hard to apply thin enough, and the water seems to delaminate it.

Acetone: Very easy to brush on, leaves a dull white finish which is quite splotchy.

ABS glue: This stuff is ridicu-thick, and glops on. It makes spiderwebs with itself, and is dissolving the thing you're painting, so it's a chore to get a good surface where you need one.

RevRico
RevRico PowerDork
8/5/20 10:33 a.m.

In reply to tuna55 :

So I've been digging into part smoothing lately. What I'm seeing in regards to ABS and acetone for smoothing, not sure how it would apply to water proofing, is to pour the acetone into a vacuum chamber and suspend the part above it, either on a rack or hanging, and draw a vacuum for 15 minutes, so the fumes do the work. 

Not sure how feasible that is in your parts department, but this topic has become very relevant to my interests so I'll be following along and sharing anything useful I find.

 

tuna55
tuna55 MegaDork
8/7/20 10:19 p.m.

Here are the initial results after a few failed tests today. I was short on time, and didn't get the data that I needed, but I did gather some for you.

 

Carnuba wax: Was maybe 25% better than nothing. Water still spurted out random places in the print which appeared solid.

 

Acetone: It seemed brittle, though I did not test this in any meaningful way. Since I am bolting my test piece together, this is a bit tough to deal with. It did far better than Carnuba wax. It was mostly sealed, with a few pinholes poking out around 20 psi.

 

ABS glue: This stuff worked. I had to abort the test due to some other issues, but even at 55 psi, it seemed mostly solid. I had lots of other leaks, and I'll get those shortly, but for now this is the winner. The issue is that it makes a "thick paint on canvas" style texture to the finished product, which is not ideal. I cannot really sand this stuff flat either. 

tester (Forum Supporter)
tester (Forum Supporter) Reader
8/7/20 10:45 p.m.

In reply to tuna55 :

After sleeping on this a bit, I would take a look at your 3D print settings and see what options to improve the infill or overlap or some equivalent setting. It will probably slow dow the print process but yield a more solid part. A lot of print processes use a porous honeycomb like build inside "solid walls" to reduce material use and speed up printing. 
 

Good luck.

Mr_Asa
Mr_Asa Dork
8/7/20 11:07 p.m.

ABS glue is literally just acetone or other hot solvents (M.E.K. would be a good one) with ABS plastic and other fillers melted into it.  You can make it yourself with acetone and the supports from your prints.  You can make it as thick or thin as you want to get rid of that thick texture.

tuna55
tuna55 MegaDork
8/8/20 5:30 p.m.
tester (Forum Supporter) said:

In reply to tuna55 :

After sleeping on this a bit, I would take a look at your 3D print settings and see what options to improve the infill or overlap or some equivalent setting. It will probably slow dow the print process but yield a more solid part. A lot of print processes use a porous honeycomb like build inside "solid walls" to reduce material use and speed up printing. 
 

Good luck.

We have explored that to the best of our ability. We are using 100% fill. That was an early step we took. Needed for the pressure we were running. 

tuna55
tuna55 MegaDork
8/8/20 5:31 p.m.
Mr_Asa said:

ABS glue is literally just acetone or other hot solvents (M.E.K. would be a good one) with ABS plastic and other fillers melted into it.  You can make it yourself with acetone and the supports from your prints.  You can make it as thick or thin as you want to get rid of that thick texture.

Our supports are some other dissolvable material. If I can add acetone to the glue that might help though. 

Mr_Asa
Mr_Asa Dork
8/8/20 5:34 p.m.

In reply to tuna55 :

Look up the MSDS for the glue.  Something else might be the active ingredient in it.  Shouldn't be a huge issue to mix them, but if you want repeatability you need to know what they are.

Our Preferred Partners
wisRCfV6r3JyDrqWVSubXsc1twyFjC2MV19KzIIHe85yUAdYjOcZw8CFiMkZI624