My layman impression of the part is that it looks like a traditional piece. Is there anything about it that couldn't efficiently be done with traditional machining?
If not, is the next step to leverage the additive process to produce a unique design? Meshes, hallows, etc.?
Totally fair question. This part would be really hard to manufacture traditionally - there's a hollow in the middle of the part that couldn't be machined, so you'd have to machine it in two parts and then weld it.
That said, this design is very much NOT optimal, and there's a lot I can do to tune it for the process. I'm working on that more in the coming months; first I needed to 1) prove that the process could produce a functional part, and 2) get access to more advanced design software (topology optimization + lattice structure generation). Stay tuned!
This was my thought - it's an awesome read and 3D printing definitely has its place, but its strength is generally in one-off or very-short-run custom items rather than building mass produced parts.
Those are definitely good use cases, yup. There are a few things I can do to increase the value of this part (or similar ones), and in general the high end cycling market tends to require a lot of 1) lightweight structures, which additive is good for, and 2) customization.
It's also worth thinking about this in context: Even without offering any customization, the cost to buy my part in batches of ~40 actually works out pretty well (caveat: this is a decidedly high end bike part). With forged, machined, and composite parts, that's an extremely low purchase quantity; in most cases, you'd never come close to amortizing the tooling & setup time.
That said, I'm definitely interested in exploring other products, and new business models for how to sell them. This part is just the first step in that process :)
I look forward to seeing how you apply gongkai to manufacturing. Knowledge sharing has facilitated the explosive growth in web entrepreneurship; surely something similar is possible in other industries, like 3D printing metal.
Yeah, that's something I've spent a bunch of time thinking about. I've gotten a lot of emails from people around the industry saying they'd like to be able to share more process data, but I'm not sure what the best approach is. If anyone out there has any thoughts, I'd LOVE to collaborate.
Do you mean "why threads and not a nut insert?" If so, yeah - for sure, that would be easier. Partly I wanted the challenge, though. There are a lot of applications where nut inserts aren't optimal, and in general they'd be heavier too. In the coming months I'm going to focus on a part that gets glued to a post, which obviously will skirt the issue altogether.
I go into the pricing in a bit more detail in the older post (http://pencerw.com/feed/2015/3/15/3d-printing-titanium-and-t...) but the gist is that 1) the powder is expensive, and 2) the machines are really expensive, and 3) the people who know how to run them are rare and highly prized.
Essentially, reducing printing time (by doubling layer thickness) means that you don't need to amortize as much of the cost of the machine during the build.
tl;dr: It's totally possible to print titanium bicycle parts, but the business model needs a bit more work still.
For reference, this is the thread from my last blog post on the subject:
https://news.ycombinator.com/item?id=9216474
Cheers, Spencer Wright