Hacker News Reader

Comment by samwillis | Hacker News Reader

Somewhat related thread [0] from last month about the "tipless" Sharrow MX-1 prop [1] which achieves efficiency gains of between 9-15% over comparable 3-blade propeller designs. They started out working on an aircraft propeller but pivoted to a watercraft prop.

0: https://news.ycombinator.com/item?id=33949895

1: https://www.mby.com/gear/sharrow-mx-1-tipless-propeller-1101...

View on HN

I immediately questioned cavitation and that link delivered. That's INCREDIBLE. The only issue is the price of course. That looks like a perfect candidate for 3D printing and then machining the final part vs billet.

> That looks like a perfect candidate for 3D printing and then machining the final part vs billet.

Now you need two expensive machines, not just one. Billet is not that expensive. I've seen people suggest this before and I looked on their site. They show a multi axis CNC machine making the parts. You would need exactly the same machine to post machine the 3D printed part, but now you have a second expensive and slow process to manage (the metal 3D printing). With billet a huge bar of material shows up, you slice each piece in a few minutes, mount it in your CNC and go. 3D printing feed powders have their own expensive processes they need to go through, so they probably cost much more than billet, reducing any potential cost savings on the material side. And the second expensive machine and process adds a lot of new expenses. Plus, you have to validate the product all over again instead of just selling them.

I bet the high price is mostly due to patents and lack of competition. It is probably not inherently super expensive to make, beyond the need for a high end multi axis CNC machine.

That makes it a great candidate for near-net casting. That way you save the 80% of tre billet you'd be turning into chips, and your fancy CNC gets more parts per shift.

Came here to say the same, and want to add: And longer tool lifetime as well.

Alternately, you can do the rough machining on a cheaper 3-axis machine (even outsourcing it).

That’s a good point. Running two CNC machines is a lot easier than running one machine and dealing with a foundry. I mean I’ve only ever dealt with one foundry and they were flaky, and obviously there must be reliable foundries out there, but I don’t like relying on outside shops any more than necessary.

Ultimately I suspect if they wanted to make these in significant volume, die casting or loss wax casting then machining would be correct route forward. 3D printing metal is expensive and the parts are likely to not be as strong. Most 3D printing (sintering) processes for metal result in a slightly porous part that needs sealing, those small gaps produce imperfections.

https://en.m.wikipedia.org/wiki/Die_casting

https://en.m.wikipedia.org/wiki/Lost-wax_casting

Take a look at turbine blades in jet engines, they are incredibly complex with tiny air holes. They are made using a loss wax then machining process.

> I immediately questioned cavitation and that link delivered.

of course, [super]cavitation can do incredible things for efficiency through water as well. maybe even more-so!

https://en.wikipedia.org/wiki/VA-111_Shkval

Reply on news.ycombinator.com