Air leaks through 3D Printed PLA

[vagulus]

I posted this on another forum. I hope no-one minds. No-one on the other forum had any answer. Perhaps someone here will.

My current prototyping project involves an air ram which I use as a pump. Its cylinder is drawn copper tube (150mm nominal) and its head and tail are 3D Printed (polylactic acid, PLA). Its working pressure is only 0.1Bar.

My problem is that the head and tail leak air right through the PLA.

My printer uses 0.8mm nozzles and the wall thicknesses involved are 4mm (with the current printer settings this means that the walls are solid PLA, there is no matrix infill involved).

I find it hard to believe my eyes, but it is true. Air escapes right through the printed PLA walls. It gets worse ... I have painted the pressurised surfaces of the head and tail with Rust-oleum sealant and the air still gets through!

Does anyone have a solution to this problem?
I'd love to hear it because I'm all out of ideas.

----------------------------------------
Artificial Intelligence is no match for Natural Stupidity

 

[MintJulep]

This perhaps.

https://www.instructables.com/3D-Printing-Make-Water-Tight-and-Air-Tight-Contain/

 

[vagulus]

Thanks MintJulep. That's certainly a whole lot more overprinting than I have considered. I'll give it a go next time I print a head.

----------------------------------------
Artificial Intelligence is no match for Natural Stupidity

 

[System Protection]

Have you tried chemical smoothing?
Not sure if it would work but may be worth a shot.


Or leave it hollow and fill with epoxy.
I believe there is an infill pattern that would allow an epoxy fill in between walls.

 

[System Protection]

A video from CNC Kitchen on YouTube also comes to mind.

He was attempting to print translucents as clear as possible.
I believe it resulted in very strong layer adhesion.
The technique may work here.
Can't remember exactly what parameters he used, maybe slower and hotter.

I'll see if I can find it.

 

[btrueblood]

Vacuum impregnation works.

https://www.godfreywing.com/blog/the-ultimate-guide-to-sealing-3d-printed-parts

I've used thin uv-cure resin to fill and seal some smaller parts, using a vacuum pump (brake bleeder pump) and a canning jar. Cactus Juice is a brand of heat-cured acrylic resin used to impregnate and stabilize woods for wood crafters, again uses a vacuum immersion method to insure infiltration of pores.

 

[vagulus]

It seems the general consensus is:
1. Printed PLA has microscopic gaps between extruded lines if material;
2. Gaps can be filled by excess extrusion and overrun;
3. Gaps can be filled by application of some impervious substance.

I thought Rust-oleum would do the job. Perhaps there are areas not coated with Rust-oleum that are permitting air to pass. I'll slap some more of it on next time I have the thing apart.

Thanks folks.

----------------------------------------
Artificial Intelligence is no match for Natural Stupidity

 

[btrueblood]

The problem with using a solvent type sealant (Rustoleum paint, or water-based glues) is that as the water or solvent evaporates, the remaining sealant may or may not bridge the gaps, and you are left with less porosity, but not zero porosity, in the part. You might also have the skin appear dry, while the inner layers are still soft and full of solvent. 2-part epoxies and uv-cure or heat-cure acrylics, and other (typ. 2-part) condensation-reaction fillers (silicones, some urethanes) on the other hand, don't shrink (or not as much as solvent-based stuff), and can be had in thin enough/low enough viscosity formulations to more readily impregnate small pores, and will more reliably cure all the way through the thickness of the part.

 

[kjoiner]

Can you increase the perimeter number? That will result in a thicker skin for the sliced part and may help reduce leakage.

 

[CWB1]

If you've never designed a pneumatic valve, air leaks through everything including decently thick metal. I'd start with the coatings mentioned above, then move on to thickening walls if possible.

 

[vagulus]

Very interesting the air leaks "though everything". Originally, I tried using varnish to seal the PLA. That was no help. I thought Rust-oleum should fix it but, "No!" Funny that, it plugs up anything else I have tried it on. There must be something that seals thoroughly. Question is, What?"

For the record, my current print settings involve four walls and five floors. That is to say that the perimeters are four thick and the horizontal surfaces five thick. That is to say, in most cases, 3mm of material for the air to permeate.

Why so heavy duty? I am prototyping mechanical devices not printing pretty ornaments.

----------------------------------------
Artificial Intelligence is no match for Natural Stupidity

 

[3DDave]

Seems promising:

https://blog.prusa3d.com/watertight-3d-printing-pt1-vases-cups-and-other-open-models_48949/

 

[vagulus]

That is interesting 3DDave. I concur with the things they say about changing layer numbers etc. Very interested in alternative materials. (I have had a reel of ASA for ages and not found anything to use it on. I'll give that a try sometime.)

----------------------------------------
Artificial Intelligence is no match for Natural Stupidity