Slashdot Mirror
Maverick Rocketeers Pursue Space Access
Mad.Scientist writes "This article at Space.com is about mavericks who are trying to lessen the cost of going into space. One of the companies, Armadillo Aerospace, is founded by John Carmack, who is also a founder of Id Software, and the brain behind games such as Doom or Quake. I just have to say, godspeed to all." Carmack is only one of the people mentioned in this story, but see our previous story for more on Carmack's rocketry habit.
I am helping a hardware vendor optimize the E3 build of Doom right now, but I'll make a pass of replys and comments later on tonight...
(yes, the Id net connection is slashdotted at the moment)
John Carmack
I am going to have to save the parent post, because it is such a perfect example of the mindset that has made progress in aerospace so damn slow. I couldn't have said it better if I was trying to intentionally construct the stereotype. This ties directly in to the quote I had in the article -- "rocket science" has been mythologized out of all proportion to its true difficulty.
First, you are severely understating the achievements of the Wright brothers. They had to invent almost everything from scratch, including much of the theory, and there was no existence proof to show that it was possible at all. I'm really not an aviation buff, so I'm sure someone else can recount the challenges better, but it is worth noting that at the time the Wrights did their work, there was also a high profile, government funded effort underway headed by Samuel Langley. With the "best minds in the country" and government resources behind it, they still didn't make the breakthroughs.
I contend that building and flying an X-Prize class vehicle (100 km suborbital, three passengers, reusable) today is a much less daunting task than the original invention of the airplane.
We have existence proofs of what is being attempted. There is no question that it is possible, because it has been demonstrated in many different forms. The only question is how cheap it can be done.
There is a massive amount of information available. Today, anyone can go read up on things that NOBODY knew back when they were building the early rocket systems.
Obviously, computers and electronics are vastly better. Our current electronics box has all the necessary sensors and actuators for flying a spaceship, and it cost less than $15,000 to put together (yes, it runs linux).
It isn't as blatant, but other manufacturing areas have also made great progress. I had a batch of a dozen small motors made at a CNC job shop for only $1000. Even counting everything that goes into them, the total cost including valve is less than $300 each. These may well be better than the peroxide thrusters used on the Mercury capsules. It was amusing to hear the NASA pad manager tell stories about having to go bang on the Mercury thrusters with a wrench to get them to stop sputtering. Don't think that all NASA hardware performs as designed.
Pressure vessels are significantly improved. A common all-carbon-fiber tank for natural gas vehicles has a better compressed volume to mass ratio than anything that could be built in the sixties. Filament winding can make large structures that are both stronger and cheaper than the classic welded structures.
There are direct spinoffs from government rocketry development. To drill the tiny, high aspect cooling passages for the Agena upper stage engine, they had to invent brand new machining technologies. Today, you can get the same techniques done at standard industrial job shops. As far as expensive materials go, the Agena engine was made out of aluminum.
The general industrial infrastructure is also a heck of a lot better. I can order damn near anything I need for our work from McMaster-Carr at 4:00 in the morning, and it shows up two days later.
NASA spent $50 million to set up the tracking and telemetry networks for Mercury. You can get far, far better results today with a GPS and satellite modem. There are billions of dollars of space based assets already at our disposal.
I could go on for quite a while on why we would have an easier time today just replicating the efforts of the past, but that is only part of the issue. What we are aiming for in the near term is far smaller in scope than any of the projects that the public normally associates with space. Even with all the advantages of today, it would be absurd to think that we could put together a space shuttle or a Saturn V. I hesitate to make analogies, but we are effectively working on building little microprocessors instead of big mainframes. 100 km straight up and down (that gives a 5G reentry, which, while not for your grandmother, doesn't take a superhero) is just not all that hard.
Yes, there are lots of challenges to be met, and we will doubtless run into all sorts of things that we haven't even considered. We will solve them as we go. People do hard things all the time, in many different fields. The reason "rocket science" looks so much harder is just lack of familiarity.
Because the existing way of doing things in space costs tens to hundreds of millions of dollars a shot, there just isn't an opportunity for radical experimentation. The optimization problem is slowly trending towards a stable local minimum, with little chance of getting out to the global minimum. Imagine trying to develop software if you only got to compile and run your app four times a year. Imagine how much that would slow down progress, and what contortions you would go through if $100 million was riding on each run.
Build fast. Test often. Stay flexible. Mind the critical path.
John Carmack
He came to Space Access to meet with us, and it was interesting talking with him. He is certainly not an engineer, but he is actually building a lot of hardware, which is more than can be said for most folks in the space crowd.
The abject stupidities in his original design that got him a lot of flack (Fins at the top! 1.2 T/W ratio without guidance!) are now gone, and he has decided to have a testing plan before launching himself, so I think he has a decent shot at flying something and living to talk about it. I wish him luck.
An interesting question: is it easier to motivate a learned individual that never does anything, or educate an ignorant individual that actually produces things?
John Carmack
>Anyone know if any of these milestones were achieved?
>Or if not, what armadillo's latest estimates are for the same things?
The estimate from day one was:
Year one: VTVL demonstrator
Year two: manned rocket ships
Year three: space shots
The VTVL demonstrator went faster than expected, and it looked like we were going to lift a person off the ground before the end of the first year. We had a couple crashes and redesigns that set us back a bit, and we were forced to make a major change in our catalyst packs to allow us to get enough back-to-back flights without changes before putting a person on it, so we haven't yet made that "milestone bunny hop".
However, while we were waiting for some things along that development path, we wound up developing some other technologies that weren't even in the original plan -- our recent work on biprop engines wasn't really scheduled until year three or later, and the rocket rotor work is looking like it will allow some big improvements in our upcoming designs.
The current goal of record is to set some of the manned aviation 3000 meter time-to-climb records before the end of this year.
John Carmack