In every endeavour. Yet the viability of the whole can not always be deduced from them.
[youtube=http://www.youtube.com/watch?v=yRPH1DHdT7I&feature=channel]
[youtube=http://www.youtube.com/watch?v=miRV-SgYx7Q&feature=channel]
[youtube=http://www.youtube.com/watch?v=AV1eUeo27tc&feature=channel]
The Armstrong-Whitworth Ensign, Bristol Brabazon and Saro Princess all flew but none operated for real. Yet machines like DC-3, Lockheed Constellation or even the Boeing 707 were successful. What made these differences? Is a certain amount of failures always needed? Could the waste be avoided beforehand? Interesting questions.
Very interesting! I had never heard of the Brabazon but your post made me do some Wikipedia-ing on the design. I couldn’t help but draw some parallels between what the designers “thought” passengers needed and what they were actually willing to tolerate versus what NASA “thinks” a crew needs to survive a trip to Mars and what it might actually require.
Definitely great questions. I believe that it is almost always the case with complicated new technologies that several attempts must be made before one can find the right combination of features, performance, and reliability for success. There is an evolutionary, survival of the fittest process that finds the winner.
We see it happen often with lower cost technologies. E.g. the Newton was a big failure but the Palm Pilot was a huge success. There were lots of MP3 players before the iPod but it became the smash hit.
When someone claims that the Shuttle proves that RLVs are not viable or that the Concorde proves that supersonic planes cannot be economical, I like to point out that the DC-3’s success only came after the failures of several prior attempts to create practical passenger planes. Single data points shouldn’t be used for big generalizations! If after many attempts, there is still no success, then one can make the grand conclusion that the goal lies at an evolutionary dead end. (E.g. see long distance airships.)
The problem with RLV/Supersonic airliners, of course, is that there is some large minimum investment involved in creating each data point. That slows down the evolutionary process considerably as compared to PDAs and MP3 players.
One interesting note is that DC-3 and Lockheed Constellation were both designed as response to specific airline requests, while these three British projects were government ideas IIRC.
The Brabazon looks like a church inside. A very tall cabin. Who on Earth thought of one flight per week in super luxury, all developed with taxpayers’ money? No wonder the labor government was voted in soon.
It is hard for me to grasp the British mindset at times.
There were lots of trial and error things in the USA as well. Lockheed Constitution for example – a single huge airliner built just post-war, that ended up as a bill board in Vegas.
“…these three British projects were government ideas IIRC.”
Government projects, of course, have the additional problem that they are difficult to kill once they start, even if huge problems are spotted early on. We see that with NASA, e.g. the Shuttle, Ares I, MSL, etc.. NASA will quickly dump a small project that runs into problems no matter how promising it is (e.g. electrodynamic tethers) but once a major mission program gets rolling, it is unstoppable regardless of how many serious problems it has. They just throw more and more money at it and repeatedly delay its completion.
OTOH, if the government is funding several competing entries from outsiders, as with COTS, then the fittest can be selected based on performance and cost. I think we would both agree that there is no way that a ridiculous contraption like Ares I would win in a fair fight competition with EELVs and Falcon 9.