There's a demon in the skies over the Mojave Desert, and it's taken another life and another vehicle. In his latest Firewall, Bill Whittle draws on his experience as a pilot to help unravel the psychology of the presumed pilot error behind the crash of Virgin Galactic's SpaceShipTwo.

TRANSCRIPT:

Hi everybody. I’m Bill Whittle and this is the Firewall.

Well, there’s been a crash in the Mojave. 

There’s been a lot of crashes in the Mojave desert, over and around Edwards Air Force Base – including the crash of a Northrup B-49 Flying Wing, piloted by Captain Glen Edwards over the site formerly known as Muroc Air Force Base. I learned to fly in those very same skies, over that very same desert. I never looked at that lakebed without remembering the men who were killed to get the information to allow us to fly almost 30,000 domestic commercial flights per day, without a single fatality on a major US carrier in 13 years. 

Now on October 31, 2014, Virgin Galactic’s first SpaceShipTwo vehicle, VSS Enterprise, was dropped from its carrier aircraft for it’s fourth powered flight. Piloting the mission was Pete Siebold, a veteran of Burt Rutan’s SpaceShipOne program. His co-pilot was Mike Alsbury, with 1600 hours of flight test experience. 

After release, the vehicle ignited the rocket engine, and about 11 seconds into powered flight the ship broke up in mid-air. To understand what happened that day, you have to understand SpaceShipTwo, and to understand SpaceShipTwo, you have to understand SpaceShip One. And if you are going to understand SpaceShipOne, you have to understand its designer, Burt Rutan. 

Here’s an airplane. 

Here’s a Burt Rutan-designed airplane.

Here’s an airplane.

Here’s another Rutan airplane. And another. And another. And another. 

They don’t look like other airplanes, and that’s because I’m convinced that the only thing Burt Rutan could not design and build would be the simple box that he could think outside of. 

So when considering his essentially home-built space program, Burt had a problem. It’s the same problem the designers of the X-15 had a generation before with their rocket plane. 

On the way up, you want your spaceplane to look like a dart: aerodynamically smooth, because your enemy is not just gravity, but air resistance, called drag. A dart punches cleanly through aerodynamic drag. 

But! On the way down you have a problem. That sleek dart is made for speed, and on the way back into the atmosphere, it is going very fast, which means it gets very hot, which means heat tiles, weight, complexity and money. 

Burt realized that what he needed was two different aircraft: a dart on the way up, and a parachute on the way down. So that’s what Burt Rutan did: he built two airplanes into one. Burt Rutan decided to break the aircraft in half, rotate not just elevators or stablizers but the entire back half of the vehicle an entire ninety degrees. This configuration meant that when it began to come into contact with the air once again, SpaceShipOne would not re-enter pointy-nose first – not a fast, clean, deep, hot dive into the pool of air -- but rather bottom first: a flat, shallow, high-drag belly flop. 

A dart like The Space Shuttle has to renter the atmosphere at a very precise angle; just a degree or two low and it comes in too fast and steep and burns up. But this feathered design is foolproof. 

So how could this foolproof system designed to save the vehicle, destroy the vehicle?  Well, if you’re going to break an aircraft in half and put it back together again – in the air – you need to do it when aerodynamic pressure – we call that “q” – is very low. At the top of the flight, both SpaceShipOne and SpaceShipTwo are at essentially zero q – there’s no air in space. And the high-drag reentry means that the plane can be unfeathered back to an aerodynamic dart very high, while air is thin and q is still low, in order to glide to landing. 

There’s two parts to the feathering mechanism on SpaceShipOne, and I assume that to be the case on Virgin’s SpaceShipTwo: one mechanism unlocks the tail section, and a second hydraulic system actually moves it. 

Shortly after regaining consciousness in the hospital, Britain’s Mail Online reports that Pete Siebold saw co-pilot Mike Alsbury unlock the feathering mechanism early and then frantically try to shut down the engine. Unlocking the feathering mechanism – especially right at that transonic moment when the airframe is passing through max q, the highest loads on the airframe – likely resulted in aerodynamic forces overcoming the hydraulics, and pitching the vehicle belly-first at something like four times the aerodynamic loads it was designed to withstand.  

So if that’s the case – IF! – then what could have made an experienced test pilot like Mike Alsbury make such a mistake? Was it lack of training?

Well, as a thousand-hour, instrument rated pilot with some time in experimental aircraft myself, if I had to guess, I would say Mike Alsbury made that error not in spite of how he was trained, but because of it. 

When we do our checklists, we are trained to reach out and touch the various controls and instruments we are referring to. This builds muscle memory. It conditions us to focus our attention where we need it. 

Another thing we are taught – especially in instrument training, is pretty simple: what are the next two things. You have what seems like a million things to do, in order, so what are the next two things? Tune and identify a VOR frequency, turn the VOR to the 167 degree radial, let’s say. 

That’s what I think happened. I think he was worried about the next two things and muscle memory did the rest.

When pilots crashed in the Mojave trying to break the sound barrier, they talked about a demon that lived in the air, just past Mach One. That demon had destroyed a lot of airplanes and killed a lot of pilots. 

There is a demon up there, in the skies. It’s not the aerodynamic shock waves that locked elevators and likely deployed the unlocked feathering mechanism on SpaceShipTwo. It’s not re-entry heat. The demon is not even gravity. 

The demon in the air is us. It’s lemurs and monkeys trying to fly faster than a rifle bullet and climb out of the air entire and see the world from the weightlessness of by-God outer space just because we want to and just because we can. But we carry with us the demons of psychology of microseconds and mach numbers and thrust chamber PSIs. We weren’t made to do this. We have to learn how. 

It sounds simple – tragically simple. But it’s just one more thing we did not know before that flight, and now we do know. We won’t make that mistake again. We can go ahead and make new mistakes now, and those will kill and injure test pilots, and destroy vehicles as well. Test flying is not just to work out the bugs in the hardware. It’s to work out the bugs in our heads. We don’t know what we don’t know. We have to go up there and find out.