The big challenge with “small cap” start ups in the space arena is that the big opportunities are “supply chain” opportunities. What makes the capital requirements for a space system high is how much of a new space system has to be custom rather than coming out of catalogs. Valves. Bits of electronics. Actuators. The traditional supply chains do not serve these needs well because they are set up to handle massive quantities and the customer base is still heterogenous. But where modern manufacturing techniques and CAD can change the game for lower-tier suppliers is that today it ought to be possible to, for example, design a family of valves and make-to-order the size/configuration someone needs for a reasonable price. There is room for cleverness there which does not require huge capital investment. Not so glamorous as complete systems, but very important to rebuilding a 21st century aerospace component supplier chain.
” …it ought to be possible to, for example, design a family of valves and make-to-order the size/configuration someone needs for a reasonable price. There is room for cleverness there which does not require huge capital investment.”
I can see 3d printing of custom parts being outsourced to a network of people operating 3d printers at home to bring those costs down – in fact, I see that as the future of manufacturing. However, ITAR prevents American space companies from doing that; you can’t share CAD files of a component of a space vehicle across the US border.
I’m thinking part of the issue is that the parts present a catch-22. You won’t generally use off-the-shelf industrial parts because they’re not optimized for absolute minimum weight, and rockets are so expensive that you need everything as light as possible. But that means you use all custom parts, and that makes the rocket so expensive that you can’t afford the mass penalty of using standard industrial parts. Riding along with that problem is that the parts and the rocket are so expensive that you want to make absolutely sure that each one will work, which requires much more testing, checking, certifying, and documenting, which adds to the expense.
I think the root of the cost spiral is the engine, and that cost comes from complexity, the turbopumps, and trying to conduct massive heat fluxes through the chamber wall. The way we initially attacked the pump and cooling problem works, but it’s a very expensive solution to design and build. I’ve got some alternate ideas, but they’re going to take a lot of doodling, fiddling, and trial and error to see if they’re even potentially workable.
There’s a widget called the ‘Sandia Cooler’ developed for heat sink cooling that may be of interest. It’s a quite clever method ‘to conduct massive heat fluxes through the chamber wall’.
They aren’t even vaguely in the same class of ‘massive’, but it really seems like it should do wonders in a much wider array of situations than the ‘cool a CPU’ problem. I have no idea if it fits your specific case, but the core concept is brilliant IMNSHO.
Ooo… That is neat, but I’m thinking they might not have a good solid-solid thermal contact to the chip unless they use a fluid coupling.
The issue in a rocket chamber is a bit different, because ideally you’d like a tiny coefficient of convective heat transfer between the gas and the chamber walls, but you’re stuck with one that gets bigger as you up the chamber pressure and flow rate.
So, just to move things along, my method is to move the chamber wall like a conveyor belt so it’s only exposed to the heat for 10 or so milliseconds, and then takes a long bath in the coolant, producing a small duty cycle between heating and cooling, or a small ratio of heating area to cooling area, making it trivially easy to cool a chamber at extremely high flux rates because the chamber wall is moving steel at a mass flux rate comparable to the propellant flux, carrying the heat away as hot steel (or your favorite combustion chamber wall material).
Needless to say, making an enclosed vessel out of moving steel is geometrically challenging, and making it not leak ever more so, and making it hold against pressure excursions is even more fun. I think I can do all of those things, but I won’t yet swear to it. The numbers on the heat transfer look excellent, and the sketches of various ways to plug up the leaks seem workable. By adjusting the RPM on the rollers, you can even keep significant heat from making it through from the liner to the rollers, so those can be aluminum. The tricks you have to pull with pressure drops to keep the roller bearing loads under control are interesting.
Might work. Might not. But it will be cheap, use off-the-shelf parts, and the only required tools are a saw, a drill press, a Bridgeport mill, and probably a lathe.
The big challenge with “small cap” start ups in the space arena is that the big opportunities are “supply chain” opportunities. What makes the capital requirements for a space system high is how much of a new space system has to be custom rather than coming out of catalogs. Valves. Bits of electronics. Actuators. The traditional supply chains do not serve these needs well because they are set up to handle massive quantities and the customer base is still heterogenous. But where modern manufacturing techniques and CAD can change the game for lower-tier suppliers is that today it ought to be possible to, for example, design a family of valves and make-to-order the size/configuration someone needs for a reasonable price. There is room for cleverness there which does not require huge capital investment. Not so glamorous as complete systems, but very important to rebuilding a 21st century aerospace component supplier chain.
” …it ought to be possible to, for example, design a family of valves and make-to-order the size/configuration someone needs for a reasonable price. There is room for cleverness there which does not require huge capital investment.”
I can see 3d printing of custom parts being outsourced to a network of people operating 3d printers at home to bring those costs down – in fact, I see that as the future of manufacturing. However, ITAR prevents American space companies from doing that; you can’t share CAD files of a component of a space vehicle across the US border.
I’m thinking part of the issue is that the parts present a catch-22. You won’t generally use off-the-shelf industrial parts because they’re not optimized for absolute minimum weight, and rockets are so expensive that you need everything as light as possible. But that means you use all custom parts, and that makes the rocket so expensive that you can’t afford the mass penalty of using standard industrial parts. Riding along with that problem is that the parts and the rocket are so expensive that you want to make absolutely sure that each one will work, which requires much more testing, checking, certifying, and documenting, which adds to the expense.
I think the root of the cost spiral is the engine, and that cost comes from complexity, the turbopumps, and trying to conduct massive heat fluxes through the chamber wall. The way we initially attacked the pump and cooling problem works, but it’s a very expensive solution to design and build. I’ve got some alternate ideas, but they’re going to take a lot of doodling, fiddling, and trial and error to see if they’re even potentially workable.
There’s a widget called the ‘Sandia Cooler’ developed for heat sink cooling that may be of interest. It’s a quite clever method ‘to conduct massive heat fluxes through the chamber wall’.
They aren’t even vaguely in the same class of ‘massive’, but it really seems like it should do wonders in a much wider array of situations than the ‘cool a CPU’ problem. I have no idea if it fits your specific case, but the core concept is brilliant IMNSHO.
https://ip.sandia.gov/technology.do/techID=66
Ooo… That is neat, but I’m thinking they might not have a good solid-solid thermal contact to the chip unless they use a fluid coupling.
The issue in a rocket chamber is a bit different, because ideally you’d like a tiny coefficient of convective heat transfer between the gas and the chamber walls, but you’re stuck with one that gets bigger as you up the chamber pressure and flow rate.
So, just to move things along, my method is to move the chamber wall like a conveyor belt so it’s only exposed to the heat for 10 or so milliseconds, and then takes a long bath in the coolant, producing a small duty cycle between heating and cooling, or a small ratio of heating area to cooling area, making it trivially easy to cool a chamber at extremely high flux rates because the chamber wall is moving steel at a mass flux rate comparable to the propellant flux, carrying the heat away as hot steel (or your favorite combustion chamber wall material).
Needless to say, making an enclosed vessel out of moving steel is geometrically challenging, and making it not leak ever more so, and making it hold against pressure excursions is even more fun. I think I can do all of those things, but I won’t yet swear to it. The numbers on the heat transfer look excellent, and the sketches of various ways to plug up the leaks seem workable. By adjusting the RPM on the rollers, you can even keep significant heat from making it through from the liner to the rollers, so those can be aluminum. The tricks you have to pull with pressure drops to keep the roller bearing loads under control are interesting.
Might work. Might not. But it will be cheap, use off-the-shelf parts, and the only required tools are a saw, a drill press, a Bridgeport mill, and probably a lathe.