Hacker News Reader

Comment by aguyfromnb | Hacker News Reader

>They designed and built a custom welding machine _and_ a custom X-Ray machine for building tank domes in four weeks?

Can any engineer speak to whether this is actually progress, or even necessary? It seems like every time something goes wrong, Elon "invents" a way to do it better.

I'm old enough to remember rescue subs, battery swaps, brain implants that will cure autism, factories spitting out factories, solar shingles, truck carriers, one-hour body shop service, tunnels at a fraction of the cost, the ability to identify squeaks in your car with your phone, robots assembling cars so fast they need a strobe light to be seen...

View on HN

I can't speak to how common that sort of custom tooling is in the industry, but as to whether it was necessary the article already explains that pretty clearly:

> The current process for building a pressure dome takes about a week; 1 or 2 days to tack up and fit steel sheets, 4 days to weld the sheets together, and 1 to 2 days for X-ray inspections and repairs.

> [...]

> The knuckle seamer looks something like a giant zipper that articulates over the front and back of a dome, like a taco shell around its filling. [...] Then, in about 10 minutes, an automated torch will trace down the length of the curve, providing a precise weld. Following this, the dome is rotated to bring the next seam into view.

> [...]

> With this new X-ray machine, SpaceX hopes to compress a process that can take a day down to a few hours.

So the new machines reduce the amount of time needed to weld and inspect a pressure dome from several days to several hours. I'm certainly no expert, but that seems like a pretty big deal to me.

I'm not in that area, but I struggle to find an analogue of a large pressurized tank on Earth where mass is the dominant cost factor.

You might have equally big tanks on Earth, under equal pressures, but it will be more cost-effective to just increase the material used on the shell of the tank to solve your problems. Whereas in space, every kg that can theoretically be removed from the launch vehicle is worth investing in, even if it raises the cost of engineering and materials.

So yes it seems reasonable to me that these are groundbreaking challenges.

>where mass is the dominant cost factor.

The compounding importance of mass is one factor here, but only half the story. The other half is how SpaceX' quick iterations and ability to retire vehicles early allows ample room to experiment and also to shave off safety factors.

In the more mundane aerospace engineering mass is quite important, but the manufacturing, servicing, longevity etc. are pretty important considerations too. A typical airplane is designed for manufacture of some 200 ... 2000 units and to remain in use for 1 ... 3 decades. This, together with the usual requirement for human flight ratings, ends up dictating rather high safety margins for designs.

Contrast that with SpaceX' vehicles which are supposed to perform a dozen unmanned flights each at most, and where one-way mission of a vehicle is a perfectly feasible strategy for testing while still making decent revenue.

Quick iterations get shit done in aerospace. Among historical examples, that's how MacCready's Gossamer Albatross[1] record-breaker came to be, where other teams failed.

[1] https://en.wikipedia.org/wiki/MacCready_Gossamer_Albatross

>I'm not in that area, but I struggle to find an analogue of a large pressurized tank on Earth where mass is the dominant cost factor.

If you stretch the definition of "tank" and "on Earth" then modern pressurized airliner cabins definitely qualify.

>modern pressurized airliner cabins

It's a good reference point, even if there are some key differences worth mentioning.

First up, an airliner fuselage is specced for tens of thousands of pressure cycles; needs to sustain the wear & tear, accumulated stress, and distortion. The rocket needs to handle maybe a dozen pressure cycles at most, and it is feasible to only handle one cycle for first few experimental vehicles.

Secondly, the airliner fuselage has much more complex shape, due to conflicting needs for aerodynamics, attachment of wings & control surfaces, etc.. This, together with varying materials and large openings concentrates stresses in key points. The rocket tank is as close to a big dumb uniform pipe as you can get. Even better, a rocket may rely on internal pressure for structural rigidity, while airliner needs to handle depressurization.

Lastly, the Starship is designed to survive re-entry, thus heat resistance is a major concern there; not anymore for an airliner ever since the Concorde SSTO got retired.

>So yes it seems reasonable to me that these are groundbreaking challenges.

Isn't weld strength something engineers would know/understand in advance, though? How would it come as a surprise that the welds would fail under pressure?

It seems strange to me that Musk could be spending so much time working on Starship, and somehow a "badly designed, badly built, and badly checked" rocket makes it to the platform and explodes.

Welding by hand is both an art and a science. There is going to be variability in the quality of the weld, which will introduce variability into the strength. Machines can be far more precise and consistent compared to humans.

They probably designed their tolerances assuming perfect welds and then didn't have them.

>They probably designed their tolerances assuming perfect welds and then didn't have them.

This sounds like terrible engineering practice.

It's perfectly normal engineering practice. Camshaft, gearbox profiles, and bearing surfaces in your car have very tight tolerances. A one slightly unbalanced turbine blade in your airliner engine would assure catastrophic failure. etc etc

Do you not see a difference between the intricate components of an engine and the welds on a tank as far as complexity?

Is it even possible to discuss anything Elon Musk anymore without 100% praise of his engineering prowess? Every single comment gets down-voted into oblivion, with very few replies. And, due to the "cool off", this prevents any balanced discourse.

Yes, but these items (camshaft, gearbox etc) are not made "by hand".

Camshafts were turned then ground manually on line lathes and surface grinders for many decades before advent of CNC, with about the same degree of dimensional precision as today.

It is more an issue of quality control and inspection routines.

Necessity is the mother of invention?

There are additional constraints not found in most welding projects. They can’t just make the material and weld thicker/heavier as that would impact the rocket performance.

So it has to be as thin and fragile as possible but no thinner. That makes it very hard to get right, and the welds have to be done to very fine tolerances on thin material, you could say it is rocket science.

>They can’t just make the material and weld thicker/heavier as that would impact the rocket performance.

I understand. But the stresses that materials and welds can withstand are not unknown variables. We're talking about a pressurized tank here. You aren't going to build a rocket (designed to carry people) that has tolerances so tight that it's touch and go whether it explodes as soon as it's pressurized.

They are currently not primarily working on the rocket, but in the production process.

> Can any engineer speak to whether this is actually progress, or even necessary?

No, they probably can't. It would be really weird for an outside engineer to have a better insight on this than the engineers from the actual rocket company.

Reply on news.ycombinator.com