I started writing this post after the Orbital Sciences rocket explosion earlier this week. Unmanned, nothing hurt except pride and stock price. I was going to begin with a joke about how the sound of a rocket blowing up is spelt.
Then news came of the Virgin Galactic tragedy during test flight on Friday, and it didn’t seem right. Appalling news and certainly not something to joke about. In such moments we have first to think of the family, friends, and colleagues of those affected.
To me both accidents underscore the fact that rocket science is – well, rocket science. It’s gained that repute for a reason. We’ve long fantasised about space-flight becoming as routine as jumping into the car. But the laws of physics – particularly, the way the energy curves rise – tell me that reality is otherwise, especially when we think of going into orbit or beyond. While it’s tempting to put the Antares booster failure down to use of left over Soviet moon rocket motors from the 1960s, the fact remains that all rockets, and especially those required to boost something into orbit or beyond, are high-tech engineering that push the limits of materials physics. And complex systems, inevitably, fail in complex ways.
The problem is the energy curve, which is exponential. On the face of it, rockets are simple – so simple that medieval technology could produce them. Something burns in a combustion chamber, hot gases rush out a nozzle at one end and push the rocket in the other, thanks to the Third Law of our friend Sir Isaac Newton.
Unfortunately it doesn’t scale up well. The Royal Arsenal, at Woolwich, was able to make fireworks into battlefield weapons by the turn of the nineteenth century, the ‘Congreves’ fired at Fort McHenry to produce ‘the rockets’ red glare’ of the US national anthem. But powder rockets were limited by chemistry. Other chemical reactions – liquid oxygen and an oil fraction, for instance – offered more energy, and by the first decades of the twentieth century, engineers were working on liquid fuelled rockets.
The problem engineers hit was mass ratio, the difference between the mass of an empty and fuelled rocket. This is everything in rocketry. The equation is R = (Mpt/Me) + 1, where R is the ratio, M is mass in kg, Mpt is propellant mass, and Me is empty rocket mass. It’s important because of the other rocket equation, Δv = Ve*lnR, where Δv is total change-of-velocity capacity, Ve is exhaust velocity and R is mass ratio. Ln is the natural logarithm of X, the exponent to which the transcendental number e (2.7182818…) has to be raised to equal X, which in this equation is R. See what I mean about rocket science?
OK, enough geek porn. The point being that the lighter the rocket vs fuel mass, at all times, the better off you are. A rocket with half-empty tanks is lugging wasted mass, and that is a killer when it comes to the energy needed to reach orbit.
That’s why Convair’s SM-65 Atlas booster, developed from 1951, was an aluminium balloon that dropped two engines on the way up. It’s why the Saturn V Moon rocket had three stages, renewing that mass-ratio every time it dropped an empty stage.
Achieving this wasn’t easy. In theory, a chemical rocket is simple – a tank of fuel (let’s say kerosene), a tank of oxidiser (let’s say liquid oxygen). Pump both into a combustion chamber, ignite them, and off you go. Actually, problems begin at once. Liquid oxygen (LOX) is super-cold – a light blue liquid that boils at 90.19 degrees Kelvin (-297.3 degrees Fahrenheit, -183 degrees Celsius). You have to pump it into the rocket at the last minute, because it’ll boil off fairly soon even in the best-insulated tank. You have to keep LOX away from its fuel (kerosine or liquid hydrogen, typically) until it’s needed. But sealing joints can be difficult. Early seals used Ulmer leather, which LOX tended to saturate and ignite. Today various exotic compounds are used. Duct tape is NOT among them.
Wait, there’s more. Kerosine and oxygen burns at 3670 degrees K – (3396.8 deg C, 6146.3 deg F). Even titanium melts at 1668 deg C. How do you stop your motor melting? One answer is to make a thin double-wall chamber and nozzle and use your LOX (or liquid hydrogen in a LOX/LH rocket) as a coolant, before sending it into the combustion chamber. Of course, you have to get the rate of flow right to make sure your cryo-liquid provides enough cooling – but relate that to the flows needed to burn properly in the motor.
That takes a ton of geekery. And there’s the fact that being super-cooled on one side and super-hot on the other is a Bad Thing in terms of metallurgy, one of many reasons why big rocket motors have reliable firing times, without maintenance, of minutes.
That’s without considering pressures. Combustion chamber pressures in the F-1 motor that launched men to the Moon topped 70 megapascals – 1,015 psi, or around 69 times the pressure of the air you and I happily breathe at sea level. That makes ‘thin-walled’ a moot term with the only answer being – you’ve guessed it – more mass, even if you do something mathematically clever with curves and ribs to increase relative thickness.
The next problem is firing. Asymmetric combustion can cause shock waves strong enough to destroy the motor. Rocketdyne made their huge F-1 burn properly – and launch men to the Moon on the back of 750,000+ kg of thrust per motor – with tests that included igniting black powder charges inside the combustion chamber during engine firings.
Since the mid-twentieth century, developments in chemistry have offered ways of building solid fuel rockets that approach liquid-fuel energies without the mechanical complexity. But once lit, they can’t be stopped. That worried Space Shuttle launch controllers, who envisaged chunks of burning SRB crashing through the windows of the Cape control centre if there was an emergency abort-and-fly-back.
Wait, there’s even more to rocketry. All the thrust is at the bottom of the stack – like trying to loft a pencil balanced on your finger. One answer is to add gyroscopes (more mass). What about control vanes in the rocket blast, or wings? The former have to stand up to metal-melting blasts. The latter add drag during the initial launch phase and (naturally) mass.
The current approach is to pivot the motors, though this adds mass – and then, how do you instruct that system? Yup – gadgets that add even more mass.
By any measure, the science demands expensive and exotic materials, expensively machined to miniscule tolerances, because the engineering parameters are completely unforgiving. A near-invisible scrap of loose metal in a valve – even an over-tightened screw with a slight burr over which a wire passes and abrades – might be enough to kill a system. That’s why rocketry is so expensive. It’s why I doubt we’ll get ‘gas-and-go’ car reliability for orbital rocket launchers. To do that, we need a different technology.
For me the amazing thing isn’t that rockets fail. It’s that they don’t. Much. They are incredibly complex machines – and they do stuff that, in an everyday sense, is not ordinarily possible. To me that underscores the tremendous skill and work that goes into any launch. Very, very talented people work like Trojans, doing very, very smart things, to make sure. I salute them. They are tweaking the nose of physics, as we currently understand it. And most times, it works.
See why it’s called ‘rocket science’?
Copyright © Matthew Wright 2014