They're developing this thing in stages; they've developed and tested the engines, they've developed and are testing launching the thing, they've developed and are testing the re-entry flop. They have partially developed the whole "landing" thing and if you're going to be testing the belly flop, you might as well put whatever you have with regards to landing through its paces as well.

By having a small team develop and test individual parts in stages, you avoid needing a giant team, and you avoid needing to have all of the schedules line up to be ready to go all at once. And it's the engineers who develop the rocket that is expensive, not the manufacturing.

IMHO the benefits of having a completely flexible schedule greatly outweigh the expensive of assuming failure.

Pretty much every NASA engineer that has worked with SpaceX has been incredibly impressed by their methods. Literally every single one of their competitors, including nation states have started to change their who launch strategy because they saw no other way to compete with Spacex (and they are still failing) all of them have expressed how impressed they are with SpaceX.

You also seem to have a very trivial understanding of what SpaceX is doing and how it is different then what has been done before.

SpaceX is not just developing a the biggest rocket ever (double Satrun V lift-off thrust), or just the most advanced rocket engine ever, or just a new method of landing (very different from Shuttle), or a new type of heat shield (again different from Shuttle or SpaceX Pica), or a method of mass producing rockets. They are fully vertically integrated and they build and work on all of these and many other things at the same time.

Having one prototype explode is fine, because they are not just testing the prototype, they are testing the production of the rocket and optimizing the operations and everything mentioned above. The goal is to have the most advanced rocket of all time by a long margin, and be able to produce it cheaper then a current generation rocket, and operating it as cheaply as a airliner.

You can not simulate or plan anything so complex and get it right the first time. They are constantly learning and improving every single aspect. The engine team uses every test fire for new inputs and adjust their simulation and improve the design, the engine manufacturing team is trying to improve the cost per engine and the production quality, the rocket manufacturing engineers are constantly improving the way the rocket is built and later prototypes have these improvements, the materials team invents new materials and is measuring their performance on the prototypes, the operations team is trying to reduce the amount of work it takes set up a launch and I could go on and on like that.

Literally every prototype coming of the manufacturing line is a incrementally improved and usually cheaper then the one that came before it. The current SN8, SN9, SN10, SN11 are only small differences and with them they want to test the landing. They will be museum pieces once the landing is done, them blowing up generates real world data that can then improve their simulations. One of SpaceX core believe with this project is that the manufacturing system is 10-100x more difficult then the design of the rocket itself, and that their goals can only be achieved if these two things are co-designed. A test of a prototype is not just testing the design but also the manufacturing and quality control.

When you try to design something that has never been done before and is incredibly complex, and you have the requirement that it must be incredibly cheap and incredibly reliable there is no way you can simply design it on a white sheet of paper, simulate it and then simply have a mass production system of that vehicle. That is simply impossible.

And your claim that this is just 'cash burn' for no reason is equally wrong. SpaceX approach to engineering has routinely beat everybody in every single competition and usually by quite a lot. Nobody knows better how to develop something with a small budget compared to SpaceX. The whole Starship program so far, including new revolutionary engines, has cost less half of the much smaller Ariane 6, or less then a single year of the SLS program.

SpaceX is simply universally acknowledged as the most advanced rocket company in the world, and its not close. They are doing it this way for a reason, and they are a lot smarter then you.

There is a whole community that has been following this project since day 1. There are people filming the production site every day. The analyze every part they see and try to figure out what changed. Some people in the community can get questions to Elon Musk and others.

The reason you are getting down-voted is that you seem to have not studied it at all what SpaceX is actually doing, and simply jump into the comments and make strong statements that what SpaceX is doing is wrong.

And your justification for this opinion is totally unconvincing for anybody that has actually studied the problem.

> I don't care who you believe they are beating in the market place and how magnificent you believe this rocket is.

Its really not about believe. Its a simple fact that everybody has acknowledged.

Sure, and the same goes for you. Empirically, though, we know that when the US/NASA throws a ton of money at something and has a good engineering culture, we can get brilliant things like Apollo. When that culture erodes and the appetite for burning cash fades, we get things like the Space Shuttle. Yes, amazing, but mediocre when compared to the achievements of the Apollo program, and ultimately ending in failure and decommissioning.

Empirically, we have SpaceX, which has basically invented the "reusable rocket" category, using a hybrid development method that includes simulations and building and testing prototypes early and often.

Empirically, we (still) have Boeing, which has stuck with the old model, overengineering from the start. What do we have? A development process that is years behind SpaceX's, for a lot more money.

Learning from failure is fine when the only cost is money and materials. SpaceX's progress so far seems to have proven their development model, even before Starship.

I do find it really funny that you start out by deriding everyone for just speculating, but then finish your post loudly denouncing everyone who supports SpaceX as some sort of idiot... which is exactly the same sort of speculation... speculation which is not supported by the facts on the ground.

Yeah, they were doing high fidelity high performance computational fluid dynamics simulations and aerodynamic simulations by 2015.

So I think they've done the sims, and now they're doing it in real life.

Some day we'll have good enough models for air flows in all necessary modes and good enough computers to run those models; we aren't there yet. Testing is still required.

We cannot fully predict turbulent flow from first principles. I mean, we can, but when constrained by real-world computing hardware, we can only do it in very tiny domains. Turbulence requires way too many orders of magnitude in spatial resolution to fully resolve it in "real world" problems.

Therefore, any practical simulation for engineering purposes involves some way of modelling (read: approximating) turbulent behavior at sub-grid scales. There are tons of these models suitable for all kinds of different applications, domains, flows, etc. - but they are all just models. Part of the engineering effort is finding the right model for the task on hand, and for that you need experimental data.

This of course doesn't apply only to fluid dynamics (but also e.g. the overall dynamic behavior of the entire assembled rocket and rocket engine, which could interact with the fluid flows in a nontrivial manner), but the point is that simulations aren't automatically accurate because you do more of them or throw more money or compute at them (that is also a good way to burn cash) - they are still approximations of unknown accuracy until you actually compare them with data from reality to confirm whatever modeling assumptions you made. The "least cash expenditure" approach will thus involve both simulations and experiments, and you will eventually reach the point where putting money into making rockets (potentially) crash will give you better lessons than running yet another simulation with (unbeknownst to you) subtly flawed assumptions.

Simulations are always going to be "assume a spherical cow in a vacuum" to a degree, as the real world is unbelievably complex, and you have to use models at some point. These models _can_ be wrong, and the only way to test them is through experimentation.