3D printed 400mm EDF project

[SkyersJet]

I am working on a project to 3D print a large EDF (electric ducted fan). Currently there are no specific applications but I have VTOL in mind. That's why it is large for better disc loading and the pitch is low for better static performance. However changing the blades to a higher pitch for forward flight is a simple task.

I use 3D printing (FDM PETG) in order to lower costs because an EDF this large would cost many thousands to machine out of aluminium or make out of carbon fibre. An aim of this project is simply to show it's feasibility because it is popularly thought that a 3D printed EDF this large would simply explode.

It has a gearbox in order to further reduce costs by enabling the use of a cheap, off-the-shelf inrunner instead of an expensive, custom wound low Kv inrunner or a bulky and heavy outrunner. Originally the gearbox was also 3D printed but I had to abandon that and now it is machined out of steel.

Here are videos of the performance so far (all thrust is recorded in kilograms).

[youtube]GGIXOSQOuCQ[/YouTube]

[youtube]Bv7MOF2637Y[/YouTube]

[youtube]3X5cXA_lClI[/YouTube]

 

[charlesmcnall]

Well I'll reply back. This is awesome. great job.
Could you put into the text the performance you have seen so far?
It's harder for me to tease out the values on the gauges at work with no headphones.

 

[SlowCo]

Nice project :thumb:

 

[Jordan325ic]

400mm! I know nothing about turbines but they do look good. Great work, what printer are you using and how many parts are in that build as shown?

 

[SkyersJet]

Well I'll reply back. This is awesome. great job.
Could you put into the text the performance you have seen so far?
It's harder for me to tease out the values on the gauges at work with no headphones.

Hi, thanks for your interest! All of the data is summarised at the end of my videos. In the first video (thrust test 6) I recorded 20kg thrust at 7712W. In the second video (thrust test 8 ) I recorded 18.7kg @ 7471W. NB: Different scales were used for both tests + I think that the batteries were sagging by the time I switched scales so I don't know which result is more valid.

 

[SkyersJet]

400mm! I know nothing about turbines but they do look good. Great work, what printer are you using and how many parts are in that build as shown?

Hi thank you! I use a Creality CR-20 printer to print the parts in PETG. There are 62 parts in the build.

 

[Jordan325ic]

400mm! I know nothing about turbines but they do look good. Great work, what printer are you using and how many parts are in that build as shown?

Hi thank you! I use a Creality CR-20 printer to print the parts in PETG. There are 62 parts in the build.

How do you like that printer?
I assume the impeller blades are all printed as one piece, so the limiting factor is the 220x220 build plate?
What methos did you use to bond the petg parts together?

 

[SkyersJet]

400mm! I know nothing about turbines but they do look good. Great work, what printer are you using and how many parts are in that build as shown?

Hi thank you! I use a Creality CR-20 printer to print the parts in PETG. There are 62 parts in the build.

How do you like that printer?
I assume the impeller blades are all printed as one piece, so the limiting factor is the 220x220 build plate?
What methos did you use to bond the petg parts together?

I love it. I don't know why it wasn't more popular than the Ender 3. It is not much more expensive while being far easier to assemble and is an all round better printer. Yes the print bed size is the only limiting factor. Most of the parts are screwed together except the intake ring which is glued together to save weight.

 

[spinningmagnets]

Here is a pic of a "segmented" stator, which allows each stator-tooth to be wound individually before the stator is assembled, which can increase copper-fill, and the speed at which robotic assemblers can wind the coils.

I really like how you made a much larger assembly from segmented parts. This kind of thinking allows smaller and more affordable 3D printers to be used for a much broader scope of products. I also don't like the tunnel-vision some builders who are new to 3D printing, and they want to make 100% of a product out of plastic in one single print. I am a fan of combining multiple materials, such as laser-cut / water-jetted metal plates as reinforcement (aluminum, stainless steel, etc). Several 3D parts can be wrapped in fiberglass/carbon fiber cloth. An assembly os smaller parts can be sized so that the completed assembly can fit inside a section of large-diameter PVC pipe, etc...

Beautiful work.

 

[SkyersJet]

Here is a pic of a "segmented" stator, which allows each stator-tooth to be wound individually before the stator is assembled, which can increase copper-fill, and the speed at which robotic assemblers can wind the coils.

I really like how you made a much larger assembly from segmented parts. This kind of thinking allows smaller and more affordable 3D printers to be used for a much broader scope of products. I also don't like the tunnel-vision some builders who are new to 3D printing, and they want to make 100% of a product out of plastic in one single print. I am a fan of combining multiple materials, such as laser-cut / water-jetted metal plates as reinforcement (aluminum, stainless steel, etc). Several 3D parts can be wrapped in fiberglass/carbon fiber cloth. An assembly os smaller parts can be sized so that the completed assembly can fit inside a section of large-diameter PVC pipe, etc...

Beautiful work.

Hi, thank you. Yes it's faster to print this way as well in addition to not needing expensive large volume printers.

 

[SkyersJet]

It needs some more work before I can put it on an EV.
[youtube]mNwQ2N1NJN4[/YouTube]

 

[fechter]

Too bad it exploded. But great that you got it on video! I'm very impressed anyway. Now you just need stronger material. Knowing at what point it blows up is pretty important if you want to fly with it. You also need some safety factor for bugs, hail, small birds, etc.

What if you printed a mold and used that to mold some carbon fiber?

 

[nicobie]

People are starting to use 3D printers to make molds for pouring aluminum using a lost plastic process. However it would probably be a nightmare balancing the finished product.

 

[Grantmac]

Continuous fiber filament printing I believe can produce parts with almost the strength of aluminum now. Prices are coming way down too.

 

[SkyersJet]

I think 3D printed blades can work so long as I stop them from hitting the duct wall.

 

[nicobie]

Continuous fiber filament printing

Now that sounds interesting. I didn't know that they did continuous fiber. All I knew about was the infused fiber filament.

It would take a hell of a slicer to do the g code for continuous fiber to get the fiber located in the proper place for maximum strength.

 

[stan.distortion]

Continuous fiber filament printing

Now that sounds interesting. I didn't know that they did continuous fiber. All I knew about was the infused fiber filament.

It would take a hell of a slicer to do the g code for continuous fiber to get the fiber located in the proper place for maximum strength.

Next wave of 3d printer tech but its big benefits might be out of the reach to hobbyists on anything but a really small scale, being able to weave thousands of strands simultaneously in a tightly controlled environment would beat any alternative process for strength to weight. Yep, one hell of a slicer!

 

[Grantmac]

https://markforged.com/3d-printers/mark-two?utm_source=gads&utm_medium=cpc&utm_campaign=12003716395&utm_content=117251557578&utm_term=%2Bmarkforged%20%2Bmark%20%2Btwo%20%2Bprice&gclid=CjwKCAjwgISIBhBfEiwALE19SQcPfBFXS9USNRAe0pPjJwoxXuFqYS0cl3agrQmoVUSjdljrcRE6BxoClNQQAvD_BwE

It's almost hobby level now.

 

[SkyersJet]

https://markforged.com/3d-printers/mark-two?utm_source=gads&utm_medium=cpc&utm_campaign=12003716395&utm_content=117251557578&utm_term=%2Bmarkforged%20%2Bmark%20%2Btwo%20%2Bprice&gclid=CjwKCAjwgISIBhBfEiwALE19SQcPfBFXS9USNRAe0pPjJwoxXuFqYS0cl3agrQmoVUSjdljrcRE6BxoClNQQAvD_BwE

It's almost hobby level now.

Don't those things cost £10k?

 

[Grantmac]

https://markforged.com/3d-printers/mark-two?utm_source=gads&utm_medium=cpc&utm_campaign=12003716395&utm_content=117251557578&utm_term=%2Bmarkforged%20%2Bmark%20%2Btwo%20%2Bprice&gclid=CjwKCAjwgISIBhBfEiwALE19SQcPfBFXS9USNRAe0pPjJwoxXuFqYS0cl3agrQmoVUSjdljrcRE6BxoClNQQAvD_BwE

It's almost hobby level now.

Don't those things cost £10k?

Yes. But compare that with getting the tools to machine 6061 or to produce carbon fiber parts in the traditional way, costs are similar or a bit less.

 

[stan.distortion]

Slicer software alone (or weaver software?) could justifiably cost 100 times that, it would be possible to design in far more than simple strength. How an object deforms for example, build a crumple zone into perfectly straight and even tube or modify its heat pathways, all sorts beyond aligning with stresses. $10k to get hands on experience is a bargain, not an easy reach for most but well within the reach of schools and collages.

 

[Greendog]

Well that is testing to destruction, looked beautiful just before it rapidly self disassembled, :wink: sounded like there was a bit of interference between the duct and the prop, ? Suspect one of the blades came off,
Maybe could reinforce blades by laying on some carbon mat and epoxy or an aluminum rod /spar fiberglassed on to make a strengthening rib, or a cavity for a piece of alloy to go in, balancing very important too, 10g on a car wheel equals approx 10t at 840rpm , maybe a car wheel balancer is sensitive enough could use to balance prop if you made adaptor to fit it to balancer this would be good as gives dynamic balance back to front as well as up down , but old fashioned balance method good, spin then note which part facing down when stops, add weight to top until stops at different spot every time , are you balancing now ? NASA has some interesting blade designs for some of there landers maybe some strengthening ideas you could borrow

 

[SkyersJet]

I will make more attempt to balance the next one yes. I will probably do it by sanding.

 

[stan.distortion]

Older (dynamic) wheel balancers where fairly simple, not much more than a flexible mounting, a switch/trigger and a strobe light that would light up a degree wheel on the shaft. And a motor to drive it of course but it probably wouldn't be too hard to print something up to mount the whole unit flexibly.

 

[Dui ni shuo de dui]

Awesome project
Do you have some pictures of what the gearbox assembly looks like?