Apparently there was an attitude control failure toward the end of the burn. That could have been a vehicle (and pilot) killer if it had happened earlier.

I believe that is also what is mentioned in an Aviation Week article about the flight. But it's a little hard to keep track of the problems. Right now I think that they had three main problems during the flight: a hard roll during the ascent, which was probably caused by high altitude winds, the loss of attitude control, and a trim problem. Also the bang.

The wind problem could be really dicey, because unlike traditional rockets, SS1 has some hefty wing area. Either they are going to have to be careful with upper altitude winds, or they may have to find some other solution (like automating the flight controls for powered ascent). Unless someone has another idea?

Thing is, good as the simulations may get, you can never really prove a craft other than by flying it. There will be more problems in future flights. It is pointless to hope against those. You can only hope that they are all survivable and the development process continues.

I don't think we have enough data to usefully speculate on what the problems were, or their solutions. I doubt if automating the FCS will be one of the latter, though. Winds aloft is only one possible explanation for the rolls. If that had been a problem, White Knight, which has much more wing area, would certainly have noticed it, particularly after dropping its payload.

The roll excursions early in the boost were probably a transonic shock impingement on the elevons- the timing is about right for that, and after the mach number went up some more, Melvill got it under control. This instability may be hard to eliminate given the elevons' position outboard of the wing, inherently in the wake.

The 120 degree uncommanded roll at engine shutdown was caused by the trim actuator failure, as noted in several sources. This can be easily fixed, although it is certainly very important.

Reports vary on when the loud bang aft (fairing failure) occured, but the fix for that is obvious- beef up the fairing design and fly again.

The early uncommanded engine shutdown was presumably caused by the safety system detecting an impending burnthrough. This might be a hard but unpredictable limit on the total impulse the engine can deliver, thanks to uneven regression of the fuel grain. This, too, might be difficult to fix.

This might be a hard but unpredictable limit on the total impulse the engine can deliver, thanks to uneven regression of the fuel grain. This, too, might be difficult to fix.

It's not that hard to fix. Just use a liquid engine...

It's not that hard to fix. Just use a liquid engine...

The thing to remember though is that currently Spaceship One doesn't use liquids.

What's the deal with laughing gas and rubber? All rocket motors have some form of "fuel" and an "oxidizer". In solid rocket motors the oxidizer is embedded into the fuel (like an Estes rocket motor) and when lighted will burn until depleted. In liquid rockets the oxidizer is usually liquid oxygen and the fuel another liquid like hydrogen or kerosene. In our hybrid motor we use Nitrous Oxide (N2O or laughing gas) as an oxidizer and hydroxy-terminated polybutadiene (HTPB or rubber) as the fuel. Both of these can be safely stored without special precautions and will not react when put together. Finally N2O has the nice quality of self-pressurizing when at room temperature so that the space ship doesn't need complicated turbo pumps or plumbing to move the oxidizer into the combustion chamber.

As they say, that means you don't have to engineer for cryogenic fluids or a system (pressure fed is another alternative to pumping above) for pushing these fluids into the reaction chamber, and the engine is safer to operate.

Going into orbit might change things. A hybrid would need to use liquid oxidizer, and it appears to generate less ISP than a liquid rocket. Eg, the theoretical top for hybrids seems to be 280 secs (which I can't verify). This is attainable by kerosene/lox engines. The current SpaceDev engine has an ISP of 250 secs.

OK, according to the latest Av Week (via Clark Lindsey) the roll was caused by wind shear, as Dr. Day suggested. It still doesn't follow, though, that automating the flight control system would do anything to help this. The solution is more likely more peformance margin (and that's only important for specific things like winning the Ansari X-prize--it wouldn't really matter for a vehicle whose only goal was to spend significant time in suborbital space).

The failure of the cowling from heating shows the value of integrated propulsion tests. I don't think they ever did a full (or even a partial) burn of the hybrid on the ground in the vehicle, which might have shown that the thermal environment was too extreme for their materials. Such a test would be probably more affordable with a liquid system.

I read somewhere that their hybrid engine costs $100k to replace between each flight... Is that true?

If true, that is pretty shocking IMO... I would image such a per-mission cost would be a non-starter for most X-prize contestants.

And it would also explain why they test and fly so rarely.

Rocketplane reaction control system

Doesn't the oxydizer of the SpaceShipOne hybrid rocket engine serve double duty as a cold gas thruster for exoatmospheric flight control?

The rocket equipped NF-104 used a reaction control system that had problems too. There is an interesting section in The Right Stuff about how the inadequacy of the NF-104's reaction control system almost killed the test pilot.

I wonder if the X-15 had any similar problems. I imagine not, since the X-15 was such a successful program with so few crashes.

Hybrid rocket engines

For such a dirt simple engine as SpaceShipOne uses, an ISP of 250 doesn't seem too bad at all. I understand that a different hybrid engine of wax fuel and liquid oxygen is just as good as a liquid engine using kerosene and liquid oxygen in terms of ISP. I would think that refuelling a wax hybrid engine would be easier than using rubber. A wax hybrid also gives a whole new more literal meaning to the old saying of "light that big candle."