Printable Linear Air Knife for CPAP Proof of Concept

This is a proof of concept inspired by this video by This Old Tony.

To summarize, compressed air exits a several thou wide vent and is laminarized in the process. On it's own this produces a sheet of air that can travel quite a distance before dispersing. By emitting the sheet of air along a surface, the coanda effect causes it to stick to the surface and can be deflected around a radiused corner. The radius then enables the sheet to entrain adjacent air into the stream, multiplying the amount of air propelled.

This is a linear blade, Tony discusses one to start with and then builds a circular blade. Because the linear blade can be built like this a lot of the precision requirement of the slot can be achieved with shim stock. I believe that the circular blade would need much higher precision than I can achieve, and I suspect any kind of 3D printing would need precision followup machining. I would love to be wrong!

This proof of concept is comprised of a chamber block and cap plate bolted together with a shim gasket

This is designed to use a 22mm cpap hose and cpap cpap, but you can just blow into it (might pop your ears). I connected it to my resmed airsense 10 and in mask check mode I could feel the blade distinctly from several inches away, and the blade was significantly deflected by the coanda effect, but not the whole 90 degrees.

It yearns for more pressure and better surface quality, but it does work! If the performance can be improved I think air blades have a place in 3D printers as they can target in cooling air from a substantial distance. This may allow for tighter margins on sequential prints, less risk to ducting from high temp hotends, and more precise cooling of sensitive filaments to maximize layer adhesion and overhang performance
 

Parts

• 40~g of PLA
• 5x M3x8 bolts
• 5x M3 heatset inserts
• Shim “stock”

You will need to cut a [ shaped gasket out of thin material. I tried 0.02mm tinfoil and a 0.10~mm stickynote, the tinfoil was too thin, the stickynote works.
 

Printing

This is not an optimized design, not any part of it, it takes silly long to print and doesn't need to be 36~g of filament. Print the chamber standing up so the radius is not stairstepped – or smooth the stairs yourself after printing, I presume that a stepped radius just won't work but printed face down on a smooth build plate might give great results.

I filed the chamber block and cap plate smooth-ish with a needle file in the area where they form the slot. I tried ironing the cap plate, but it warped and I had to mount it textured-PEI side down or the warp closed the slot. If you had the right sanding stuff to flatten and smooth this area properly I think you would get a better result.