The Best Bad Idea

When I was a kid I came across a quirky half-reality-TV, half-documentary show called Plane Crazy. You can now watch all three parts on Youtube here:

Plane Crazy tells the story of how Robert X. Cringely, a tech entrepreneur, columnist, and fraud (who lied about getting a PhD from Stanford), attempted to build and fly an airplane of novel design solo in 30 days. The story does not go well. By the end of the 30 days homebuilt airplane legend Peter Garrison estimates that Cringely’s completed 5-7% of the total work the airplane needs to be ready to fly. The first plan is scrapped, as first plans so often are, and instead he enlists the entire staff of the Fisher Flying Products company to help him build an existing kit in 30 days. That works better.

A few moments in the show are particularly notable. In retrospect the breathless exclamations about having an internet connection, receiving a hundred emails a day, and scouring the web for information seem archaic. It’s hard to believe that a time in which I was alive seems so ancient, but I’m now in my 30s, so I suppose the time has come for radical acceptance that the 20^th Century really is past.

Cringely put some effort into coming up with something truly novel for the first attempt at the project. He describes the powerplant configuration of the Bell P-39 Airacobra, which placed the engine behind the pilot near the center of gravity and ran a shaft to a nose-mounted tractor propeller. This was done to accommodate a 37 mm cannon that fired through the propeller shaft. He leaves out that the design resulted in poor handling qualities and was largely unsuccessful in its intended use as an interceptor, though the Soviet Union found a good use for P-39s busting German Panzers on the Eastern Front during World War II. Cringley’s interest in using this configuration for his airplane is that the landing gear, engine mount, and wing carry-through structure can share the same primary piece of structure, which could be a good piece of weight-saving cleverness if you could make it work. The “if” is key. Making the contraption work in 30 days is probably insane on the face of it. Garrison was unimpressed.

Martin Hollmann is tasked with producing the piece by piece detail design of the aircraft, and he dismisses Cringely’s concepts more or less out of hand. After Cringely faxes (ah, the 90s…) a rebuttal defending his conceptual design, Hollmann’s response is terse. “You’ve got a lot of bad ideas.” Cringely deserves this to an extent, but Hollmann also botches the design of the lower fuselage and simply mails Cringely a design update with no prior discussion or explanation besides that the first design was just wrong. After practicing as an engineer for seven and a half years I think I can say with some professional confidence that neither winds up looking great in this exchange (Peter Garrison, on the other hand, seems like a savant).

Watching this again now, over 20 years after I first saw it, I wonder what it was the appealed to me so much at the time. I had an enthusiasm for aircraft bordering on obsession, and I’m sure that had something to do with it. Seeing the inside process of designing and building an airplane (two airplanes, really) was also thrilling. I wonder if seeing this was part of me deciding early on that I wanted to be a designer of machines that fly.

In retrospect the process and Cringely look awful. He complains that people would be much better if they were simply automatons who followed his instructions like computers. He boasts about working 500 hours a month, no problem, which is great if you can do it, but sounds to me like an indictment of anyone who made the choice to start a family and so needs to have a human schedule instead of an insanity schedule. In the end it doesn’t work out with the first airplane, so there’s a bit of schadenfreude in seeing this burnout-ready philosophy actually burn out before your eyes.

I hadn’t thought about Robert X. Cringely or Plane Crazy in years, until he was brought to my attention by this post. Like so many other tech-savvy and aerospace-enthusiastic people with the cash to burn, he’s started a launch company. I’m neither a Tom Mueller nor a Peter Garrison, but I feel like I have some expertise to evaluate what he’s doing now.

The basic idea is technically sound. Eldorado, Cringely’s new company, proposes dropping a small multi-stage solid rocket from an F-104 in supersonic flight to place satellites up to 12 kilograms into low Earth orbit. No one is going to say that’s not possible, which is more than can be said for some newspace ventures. Whether it’s a “good” idea is a much more complicated question.

Cringely claims Eldorado will be launching rockets by the end of the year. That’s not going to happen. Let’s move on from the topic of schedule.

The price tag is given at around one million dollars. This is an order of magnitude less than the price of a LauncherOne mission with Virgin Orbit, which is singled out several times. With the de facto demise of Pegasus and LauncherOne’s maiden flight last month, Virgin Orbit is currently the only game in town when it comes to air launch to orbit. If Eldorado can hit that price point that would make them the lowest-cost dedicated mission option available. If your payload fits in the fairing, the appeal is obvious. It’s not turtles, it’s “ifs” all the way down.

A number of claims in the post are either suspect or demonstrably false. Cringely claims the F-104 is a cheaper airplane to keep flying than a 747. It’s smaller and burns less fuel. In every other respect the F-104 would be a more difficult airplane to keep in the air. According to a quick internet search that I made zero effort to validate, over 400 747s are still flying as of this year, compared with a handful of F-104s. The 747's appeal from a spare parts standpoint is obvious. The 747 was designed to be maintainable and flyable by a civilian crew at minimum cost. The F-104 is a hot rod of an airplane that’s now antiquated as well as temperamental. I don’t think Cringely is right that Eldorado will spend less on their air launch platform than Virgin Orbit will spend on Cosmic Girl.

Releasing the rocket at a 45 degree angle and at Mach 2.2 is listed among Eldorado’s competitive advantages. Air launching at more than twice the speed LauncherOne gets started will certainly help. But the claim that LauncherOne releases in level flight is incorrect. Cosmic Girl’s pitch-up maneuver just prior to release provides a substantial increase in payload, though I need to imitate Freeman Dyson talking about thermonuclear bombs here. I don’t know how much benefit VO gets from their pitch-up maneuver, and if I did know I wouldn’t tell you.

Responsive launch, roughly defined as the ability to launch an arbitrary payload to an arbitrary orbit with minimal delay in sourcing a rocket and analyzing the mission, is a hot topic in the industry today. Air launch has some appeal on that front. It frees the system from the logistical and regulatory hurdles of working with ranges subject to the whims of the Air Force and Space Force. Weather also becomes much less of an issue, and a given site might be able to launch to a wider range of inclinations with an air launch scheme compared to a typical ground-launched rocket. The details matter, of course, and in this case it depends mainly on how far the mothership can fly out before dropping. The F-104 was not designed for long range.

Despite what Cringely claims, Eldorado will not be able to replace GPS satellites within days of the baddies shooting down the constellation. I don’t know the exact number, but the mass of a GPS spacecraft is somewhat more than 12 kilograms. The GPS constellation is also located much higher than low Earth orbit, thousands of miles further out in medium Earth orbit. Eldorado’s payload will not be 12 kilograms to that orbit. It will be zero. Cringely is either ignorant enough about the space industry to not understand the distinction of missions and his vehicle's limitations, or assumes that the reader is ignorant enough that he can sell them the space equivalent of a bridge in Brooklyn. Neither inspires confidence.

I take most issue with Cringely’s discussion of solid rocket motors vs liquid rocket engines. “We already have plenty of liquid rocket engines,” he states, elaborating that new engines are an exercise in vanity. I don’t know the exact magnitude by which solid-propellant rocket production outnumbers the liquid variety, but I would guess it’s in the range of 100:1, even with SpaceX now churning out hundreds of Merlins, Dracos, Super Dracos, and Raptors each year. Eldorado’s vehicle will require new solid rocket motors, at least two and probably three stages’ worth. Vanity comes in a variety of forms.

Of course, making a new engine or motor is not an exercise in vanity. It’s an exercise in engineering. Rockets might be at the high end of the technology readiness level scale but they are not a mature technology. They are too expensive, too unreliable, too hard to manufacture, too hard to maintain, and too expendable. There’s plenty of room for necessary improvement without even thinking about performance improvements, which would always be great too.

The appeal of going with solids for an air-launched vehicle is certainly understandable. But what doomed Pegasus wasn’t the technical challenge of air launch. Orbital Sciences, then Orbital ATK, then Northrop Grumman lost the ability to control the cost of making the vehicle by surrendering control of the motors that powered it. They were made in a quantity that required a Falcon 9 price for Falcon 1 performance. The best technical idea is a bad idea if it doesn’t earn its keep.

All that aside, clearly this is an approach that can be made to work, and it may even find some success on the market. I don't envy anyone trying to bring a startup online in the midst of a pandemic and depression, or anyone going to work for Robert X. Cringely, but if they're successful it'll be all the more impressive. It's easy to poke holes in new ideas and much harder to come up with a genuinely good idea, especially one made to attack a problem as difficult as space launch.

I'm reminded of a scene in the movie Argo, when a high-ranking official in the CIA is briefed of a plan to smuggle a handful of Americans trapped in Iran out of the country by giving them forged passports and teaching them to pose as a Canadian film crew. The official shakes his head. "This seems like a bad idea. Don't we have any good ones?"

"No sir," comes the reply "This is the best bad idea we have."