[AR] Re: Hyperganic uses AI to design a 3D-printed rocket engine
- From: David McMillan <skyefire@xxxxxxxxxxxx>
- To: arocket@xxxxxxxxxxxxx
- Date: Mon, 30 Mar 2020 15:19:54 -0400
On 3/30/2020 1:34 PM, Joe Bowen wrote:
I'd be curious to see where it goes, I think the production cost of 3d
printing is still holding back this design method, but maybe rockets
are a better place for this innovation than cars (even race cars)
Well, a regeneratively-cooled combustion chamber is probably a top
contender for 3D printing -- no moving parts, but a very complex
interior geometry, relatively low-rate production. The trick is getting
the print to match the CAD, and being confident that the printed article
is structurally sound. Those are both solvable problems (look at the
F-14 central wing box), but it's still a struggle to scale up.
For cars... I can't say much, being at the intersection of multiple
NDAs, but there are multiple startups /and/ "classic" automotive
manufacturers aggressively pursuing new production-line automation
paradigms that look a lot more like 3D printing, but in structural
assembly, with production lines that are as parallel as they are serial,
and /much/ more flexible. Basically, sacrificing serial "line speed" in
favor of parallelization and the capability to switch between products
with amazing speed.
Not ready for prime time yet, but it's no longer pie-in-the-sky
either. The big trick will be finding the ideal tradeoff between the
production technology and the product design to get maximum leverage out
of the new production paradigm. Something I anticipate will probably
take a full generation before it all shakes out, similar to the switch
from steam power to electrical for industrial production.
Other related posts: