This doesn't follow. Engineering complexity is not a limit on payload to orbit, it is a fundamentally different parameter. Yeah building a mile tall rocket would be hard, but it would get a shit ton of payload to orbit. There is no maximum beyond which making a bigger rocket starts to reduce your payload to orbit.
> But then you have another curve that I think starts small and increases near-monotonically, which is the complexity/likelihood-to-fail factor to the size of the rocket. It's (relatively) easy to launch a toy rocket, (fairly) simple to build a missile-sized sub-orbital rocket, difficult to build a small-to-medium orbital rocket, and apparently very difficult to build a Saturn/N-1/Starship-sized rocket.
Complexity does not increase with size, people just become more risk averse with size. Toy rockets fail all the time, just nobody really cares. No one would bet the lives of multiple people and hundreds of millions of dollars on a successful toy rocket launch. If complexity increases, it is with capability. If you want to land on the moon, you need something a bit more advanced than a hobby rocket. There is no reason to believe a floatilla of physically smaller rockets capable of achieving any given mission will be less complex in aggregate than a single physically larger rocket.
> This doesn't follow. Engineering complexity is not a limit on payload to orbit
At this point I'm merely talking about size (which I think is clear from the words I use. I don't think "building a mile tall rocket would be hard" adequately describes the difficulty when we haven't even built a mile tall building.
Sea Dragon[1] was only envisioned as 490 feet tall, and as near as I can tell even the Super Orion[2] would only have been 400-600 meters tall. And of course, neither of those was even close to implementation. Therefore I stand by my statement that a mile tall rocket is, for all practical purposes, impossible, and thus has a payload to orbit of zero. If you disagree then add a zero -- surely you agree we can't build a ten-mile-tall rocket?
As far as complexity, I'm not sure what to say. Toy rockets might fail all the time, but the point was complexity, and a toy rocket can be constructed from under a dozen parts. Even larger model rockets have at most a few dozen to a few hundred parts. The part count of the Falcon 9 has to number in the thousands, if not tens of thousands (9 merlin engines with at least several hundred parts each?).
To be clear, I agree with you that complexity increases with capability.
But also, to push back a bit, I don't think complexity aggregates the way you're saying it does. A box of hammers is not more complex than a nailgun, even if it has more parts in total.
1. https://en.wikipedia.org/wiki/Sea_Dragon_(rocket)
2. https://en.wikipedia.org/wiki/Project_Orion_(nuclear_propulsion)I was assuming you were using a comical example to illustrate a "nightmare to engineer." The comparison to a building doesn't actually work at all. The practical limitation on how high we can build buildings is how fast we can make elevators. Just making something tall is not a problem.
> Sea Dragon[1] was only envisioned as 490 feet tall, and as near as I can tell even the Super Orion[2] would only have been 400-600 meters tall. And of course, neither of those was even close to implementation. Therefore I stand by my statement that a mile tall rocket is, for all practical purposes, impossible
First, the optimal design for a rocket is not to just keep making it taller, and second, size was not the obstacle to either of these projects not being built. That does not at all prove that it is impossible. What kind of world would we be living in we presumed anything that hadn't already been actively pursued was impossible?
> and thus has a payload to orbit of zero.
My point was that this does not equate to a payload of zero. Surely you wouldn't argue that the weight of this mile high rocket is zero, and therefore that there is some curve for the weight of rockets where making the rockets larger starts to make them lighter. Just as we can calculate the weight for something without actually building it, so too can we calculate the payload, and it can increase far beyond anything we can actually implement.
> If you disagree then add a zero -- surely you agree we can't build a ten-mile-tall rocket?
I agree it would be impractical, but not that it would be so non-physical that we couldn't calculate what its payload capacity would be were it to be built.
> Toy rockets might fail all the time, but the point was complexity, and a toy rocket can be constructed from under a dozen parts. Even larger model rockets have at most a few dozen to a few hundred parts. The part count of the Falcon 9 has to number in the thousands, if not tens of thousands (9 merlin engines with at least several hundred parts each?).
Falcon 9 is a liquid rocket designed to take people into space. That is the source of its part count. You could scale up a solid rocket motor to an arbitrarily large size while keeping the parts count exactly the same. It's probably not the optimal way to make a solid rocket of that size, and you'd be missing out on a lot of capabilities that are important for a real rocket, but if you just wanted a toy no more capable than what you buy in a hobby store it would be no more complicated. Conversely, try to make a fully functional falcon 9 complete with 9 working liquid rocket engines small enough to hoverslam on your desk and you have an immense engineering challenge on your hands.
> But also, to push back a bit, I don't think complexity aggregates the way you're saying it does. A box of hammers is not more complex than a nailgun, even if it has more parts in total.
I concur that part count is not the same as complexity, but that point is in my favor. Making something bigger is like adding hammers to a box of hammers. The quantity goes up, and at some point you're going to need to make some improvements to the box if you want to keep adding more hammers, but conceptually it is simple. Making something more capable, like a nail-gun, is much harder.
So there's some sort of curve, zero at both ends, between overall rocket size and the payload to orbit. The question is where Starship sits on that curve, and to your point it seems likely that it's looking good on that metric alone.
But then you have another curve that I think starts small and increases near-monotonically, which is the complexity/likelihood-to-fail factor to the size of the rocket. It's (relatively) easy to launch a toy rocket, (fairly) simple to build a missile-sized sub-orbital rocket, difficult to build a small-to-medium orbital rocket, and apparently very difficult to build a Saturn/N-1/Starship-sized rocket. More props to the crazy '60s team that pulled it off.