It is demonstrably possible to develop high performance aircraft with a low rate of vehicle loss. Using rockets doesn't fundamentally change this. It's too early to say what the lessons of the SS2 tragedy will be, but I'm pretty confident that it won't be "the only safe way to develop a rocket plane is to fly without crew".
High performance aircraft crash all the time, but things like ejection seats prevent most accidents from being fatal. This year there have been non-fatal crashes of both a Russian T-50 and an American F-35.
As for T-50, I'm not aware of any hull losses in the trailing year. If you could provide the date of the mishap, it would help.
Quote from: zaitcev on 11/03/2014 01:52 amAs for T-50, I'm not aware of any hull losses in the trailing year. If you could provide the date of the mishap, it would help.I might have remembered wrong. My point was that most hull losses for high performance aircraft are non-fatal thanks to feature such as ejection seats (don't quote me on statistics though).
People are talking in this and the SpaceShipTwo threads as if an autopilot is a simple piece of technology that you can bolt in under the dash (next to the 8-track player) and have it fly a vehicle through a test program - the purpose of which is to characterize the performance of the vehicle and find out where the edges of the envelope are. It turns out that people are much better at flying by "feel" than computers are, which is what you need to do to learn what the flight characteristics are.This was 100% true 30 years ago; it may be 100% false 50 years from now, it is what it is today.
Yeah, as far as I can tell, neither SS2 nor Lynx are good fits for remote or autonomous control.If /I/ were to build a spaceplane, I would absolutely make it fly-by-wire and, if it were suborbital (and thus easily reused), then I'd test it 100 times unmanned before putting anyone in it. Good marketing for safety, plus would make the regulators happier and it wouldn't cost me that much (assuming I wanted autonomous or remote capability anyway, which I would).But XCOR has the anti-fly-by-wire religion. They want a piloted freaking first/second stage (i.e. the one /on-top/ of the airliner, not just the very final stage) for their orbital design, from what I remember. I think that the as-likely-as-not eventual test failure will kill an expensive test pilot (really, really good ones probably make $10-20 million in interest-compounded lifetime earnings) and do damage to the company's image as well as the inevitable year-long stand-down such a high profile failure will cause. If you're going to invest hundreds of millions of dollars on a spaceplane, a dozen million on a remote/autonomous capability is definitely not too much to ask, IMHO, and would allow you to fly the spacecraft packed just with unmanned cargo/experiments, too.I don't think it's /wrong/ to fly SS2 or Lynx piloted-only, I just wouldn't make that decision if I were in charge.
I suspect the reason why Scaled and XCOR avoid fly-by-wire is the complexity if its not strictly required. The physical component cost is not as much of a limit, although the development cost may be the limit.
The people at XCOR came from building rocket powered helicopters.
Quote from: QuantumG on 11/05/2014 06:10 amThe people at XCOR came from building rocket powered helicopters.Manned rocket-powered helicopters...Right from the beginning, Lynx had a much more extensive test program in mind than SS2. XCOR planned to have a test program that mimicked what they did during the Rocket Racer development, with several flights per day gradually opening up the envelop from rocket-powered taxis to black sky. This would be after gaining many, many hours of ground test time for the engines. In contrast, VG/Scaled has been extremely reticent to fire SS2's engine, either on the ground or in flight. Ironically, that inexperience of how the vehicle handled at supersonic speeds could have been a major contributor to the accident.Manned test flights are always risky, but XCOR has already been doing much more than VG/Scaled to retire as much risk as possible.
If /I/ were to build a spaceplane, I would absolutely make it fly-by-wire and, if it were suborbital (and thus easily reused), then I'd test it 100 times unmanned before putting anyone in it. Good marketing for safety, plus would make the regulators happier and it wouldn't cost me that much (assuming I wanted autonomous or remote capability anyway, which I would).
Quote from: Robotbeat on 11/05/2014 01:27 amIf /I/ were to build a spaceplane, I would absolutely make it fly-by-wire and, if it were suborbital (and thus easily reused), then I'd test it 100 times unmanned before putting anyone in it. Good marketing for safety, plus would make the regulators happier and it wouldn't cost me that much (assuming I wanted autonomous or remote capability anyway, which I would)....And you'd totally miss the point of MANNED suborbital spacecraft :)
We don't fly manned aircraft "unmanned" on test flights only UNMANNED ones because we are testing them AS WE WOULD FLY THEM which is the main, and pretty much ONLY point.
By flying "unmanned" you are in fact NOT testing or "proving" your safety as much as you don't TRUST the vehicle to fly safetly manned :)
Marketing wise you've just lined up all your PR effort against a wall and shot it :) Both Burt and XCOR realize and except that if you WANT a spacecraft that is "treated" like an aircraft you have to test AND operate it like it IS an aircraft. It still doesn't quite work that way in real life (and the FAA has made the call that it will in fact NOT be treated that way) but its a start.
