Friday, December 12, 2008

Ascent Roadmap Propellant Decisions

The Ascent Roadmap choice of propellants is currently (and tentatively) liquid oxygen and kerosene for all lower stages of all Ascent Roadmap concepts, and liquid oxygen and either liquid ethylene or liquid natural gas as the propellants for upper stages. These propellants were chosen by a process of elimination which was not documented until now. Propellants were eliminated in the following order, even before After Columbia Project had decided to work on its own detailed launch vehicle concepts in 2005. Part of the reason for that particular decision was the discovery that accurately modeling a commercial launch vehicle for simulation purposes without the help of the vendor is as difficult as starting a basic concept from first principles and modeling that instead.

Solid Propellants: It really is a no-brainer, but the reasons I've had for shunning solid propellants have varied over the years. First, I thought that it was impossible to make a non-detonable solid propellant with enough performance. This probably took a circuitous route from the Bureau of Alcohol, Tobacco, Firearms and Explosives in the United States. This myth was first shattered by learning from NASA sources through egroup discussions that the Shuttle's SRB propellant is not detonable under any circumstances. It just burns. It is also very difficult to get it to burn vigorously enough to be a rocket propellant; it takes a four-stage igniter, itself generating a thrust pulse of 27kN, just to start the SRB. Standard issue model rocket propellant is not quite as benign, because it is easier to get it to burn vigorously, but it is still non-detonable. It has been tested with blasting caps and detonation cord, which can't even ignite the stuff! But, the Bureau of Alcohol, Tobacco, Firearms, and Explosives in the United States has set itself against the amateur rocket community, and has very inaccurately opined model rocket fuel as an explosive, and very inappropriately classified it as one. Still, ground safety is an important reason why solid propellants suck for large launch vehicles. The reason why is because your rocket motor comes with its full load of propellant, and weighs as much (more if it has remove-before-flight (RBF) items attached to it) as it will at lift-off. This isn't a big deal until complete rocket motors start weighing about 25kg or 50 pounds or more. This is the threshold where you can start wrecking your back in lifting operations even with the proper technique (some people, like me, using proper technique, can lift up to 100lb without any risk of long as I'm wearing safety toes.) Some further territory can be gained by dollies, pallet jacks, forklifts, cranes, bobcats, zoom booms, etc. (“MHE”), but gas pressurized transfer and transfer pumps get cheap fast compared to all this fancy stuff. Once you're into the Kilder scale, where you need all this fancy stuff anyway, it is a heck of a lot cheaper, if it need only lift some 8-10% of the lift-off weight of a stage, and the only advantage to having solids is that it saves you from having the pumps and tanks for handling liquid propellants at the launch site. Also, with solids, the propellant still has to be loaded somewhere, and mixing the non-detonable propellant mixture of tire rubber, aluminum dust, and ammonium perchlorate oxidizer is still quite dangerous. This is why the National Association of Rocketry frowns on rocketeers mixing their own propellants and making their own motors (which are small pressure vessels with plenty of complications in their own rights. If they are done poorly, they can be very dangerous.) Not having to do this at the pad is the big advantage of solid propellants over liquids (which are actually mixed during the motors operation, not prior.)

In flight, solid motors really suck. Once started, the only way to make them stop is to blow them apart. In the Shuttle program, this has actually been done once. During the ascent of STS-33, the final mission of Challenger (better known by the manifest number 51-L), the guidance system lost communications with the solid boosters, and then about half a second later, lost electrical power. These were natural side-effects of the vehicle breaking up. Both solid rocket boosters survived the breakup of the Challenger and the burnt-orange colored External Tank hanging from her belly. The right hand solid rocket motor had an estimated thirty inch diameter hole stemming from a leak in her aft field joint, but what is less well known is that another one was just opening up in its forward field joint after it broke away from the disintegrating External Tank. Of course, this difficulty in segmenting and the more catastrophic consequences of leaks is yet another nail in the coffin of large solid-fuelled motors. As the two solid boosters separated from the Shuttle, they had no guidance and in an aerodynamic sense, are statically unstable, making their flight totally unpredictable. The left hand solid booster was approaching the edge of the flight corridor when the range safety officer hit the switch. This sent a radio signal to both boosters and the External Tank (which had already broken up) to set off a set of linear shaped charges designed to cut them into pieces. By this point, the External Tank's radio receiver, battery, and explosive charges were several hundred feet apart, floating separately into the Atlantic, so it did not respond to the signal (the External Tank's charges were later recovered, removing any doubt that they caused the disaster.) The intact solid fueled boosters obediently exploded, reducing their chamber pressures enough that they quit burning. Pieces were recovered with the last ten seconds of fuel still attached. There was initially some argument as to whether shutting down a solid motor could be done safely, and in fact, Shuttle's solid boosters had been designed with such a system. The implementation was having two hatch-like blow-off panels near the nose of each booster, positioned to direct the rocket exhaust away from the Orbiter. These had to be far larger than the throat of the nozzle to vent enough exhaust to shut the motor down. When they studied what the resulting reverse thrust forces would to the aborting Shuttle stack, they gave up. Imagine if your vehicle had the brakes and tires to actually stop on a dime from highway speeds. What would actually happen is that your wheel assemblies really would stop on a dime while your vehicle's body, with you still in it, would bust loose and skid, flip, roll, and break up all over the next couple hundred feet of the highway. Even if your whole car could handle such a stop, you'd be a splat mark all over your windshield and dashboard, even with a seatbelt and airbag. This was the sort of problem the engineers were facing. They were stuck with solid boosters already, that choice having been made in ignorance (or defiance) of all the stupidities of solid motors. On a positive note, solid motors are excellent for military applications. Typically, these sit in a stockpile or silo for years, must be ready in minutes (seconds for the ones on the battlefield), and must then work perfectly when the trigger is pulled. They can also be made strong enough to survive the violence of shutting down a solid motor. The Minuteman missile has this facility in the third stage, to prevent it from flying too fast and overshooting its target area. Most satellites do not have, nor can acquire, the strength required to survive a solid shutting down, so solid satellite boosters keep from overshooting by either fishtailing to waste energy, or by having a liquid propellant final stage which is easy to shut down (and often, restart for further maneuvers.) If the final stage is a spin-stabilized solid, which can't fishtail to adjust its performance, it is built to very tight tolerances, aimed very accurately by the last guided stage, and the result is that the cost belies its simplicity, much like a $100 ball point pen (a $0.10 ball point pen (with a lid) actually has one more moving part than a typical spinning solid motor stage!! A ball point pen with a retractable tip has three more. The only moving part on a spinning solid motor stage is the valve for the thruster that keeps it from wobbling.)

The exception to these considerations for the Ascent Roadmap is very reasonable: the first stage of the Prochron. One module of this stage has a mass of just 6.6kg, well under the lifting hazard threshold, and the advantage of being able to grab a mass-produced, off-the-shelf, mostly reusable piece ofuire equipment for this role that can be used with very little preparation and no modification is a major bonus.

Solid Fuel: That meaning hybrids with a liquid oxidizer and solid fuel. Operationally, these hybrid motors have most of the disadvantages of the solid motor, and most of the disadvantages of the liquid propellants, along with some advantages. First, hybrid motors can be throttled, shut down safely, and even restarted (although not as easily as a liquid-fueled motor.) The main disadvantage that they inherit from solids is that they are still heavy, perhaps half that of an equivalent all-solid motor. The main disadvantage that they inherit from liquid-fueled motors is that the oxidizer (sometimes more than one) still has to be loaded into the vehicle. Often, a separate oxidizer is used for ignition, in addition to the flight oxidizer. The second disadvantage they inherit from liquid systems is the relative volatility compared to solid motors. Yes, liquid fueled stages actually blow up more easily than solid ones, all other factors being equal (i.e.: if you don't have leaky rotating double bore tang-and-clevis joint designs in your solid motors that can blow up school teachers on cold days.) This is the main argument for the Ares I crewed launch vehicle NASA is developing. The counterargument for the Atlas V is simply that that booster family has flown over 80 successful launches in a row, proving that the design is solid and the people running the system are not screwing up. The Shuttle has a similar record, but the Ares I is nothing like the Shuttle, and half the people running it will be laid off for about five years between the two. Even those who come back will not be current on any of the systems after a five year break. Even without the five year break, the differences between the Shuttle and Ares I will guarantee that the launch teams won't be as comfortable as confident as those running the Atlas V, and they will be more likely to screw up, even after all that expensive retraining. And of course, rockets are not very forgiving when you screw up, solids are even less so (if you drop the External Tank from the crane, you'll be lamenting the loss of the tank...if you drop an SRB segment, you'll be lamenting the loss of whatever it lands on!)

Liquid Hydrogen: It was easy to rule out liquid hydrogen...its low density (one fourteenth that of water) dictates the need for a tank that is very light for its size, and its low boiling point 250 degrees Centigrade below that of water dictates extreme materials and good insulation. The need for very light tanks dictates a turbopump system, and pumps care about head, not pressure. A hydrogen pump needs to develop fourteen times as much head to generate the same pressure as a water pump, as well as fourteen times as much flow per minute to deliver the same mass. These difficulties don't add up...they multiply. The result is expense far beyond any benefits to be had from improved performance. If you are looking for cheap access to space, don't look to hydrogen. Also, most of these arguments also apply to the “hydrogen economy” envisioned by many environmentalists. Since private industry has more sense than NASA, I'm willing to place bets that the hydrogen economy will never develop where vehicles have to go any substantial distance from the power grid (it might appear in trolley cars and metro trains, allowing some freedom for these vehicles...if propane is not allowed in underground parking, hydrogen never will be, making it a poor choice even for city cars.)

Toxic Hypergols (full names nitrogen tetroxide, nitric acid, hydrazine, mono- and dimethyl hydrazine): These used to be the favorite choice in general, but they have environmental and ground safety problems that make them unfavorable for large commercial boosters, or small ones for that matter. If applied to Prochron, the safety equipment required to handle the propellant will cost far more than the actual launch vehicle. Hypergols do have the advantage of auto-ignition, making them ideal for thrusters once already in space. For this role, the non-toxic (but still highly reactive) hydrogen peroxide is preferable if you are looking to save money. Hydrogen peroxide was ruled out of Ascent Roadmap main propulsion systems because they don't need autoignition, and because hydrogen peroxide is quite expensive to produce. It has not been ruled out for reaction control thrusters.

Nitrous Oxide, Hydrogen Peroxide: Main argument against these is the expense of production, but they are also intrinsically hazardous because they contain inherent technical terms, a positive enthalpy of formation. In simple terms, they can blow up on their own...both have a history of doing rocket propelled vehicles. The most recent such accident was when a SpaceShipTwo propulsion system exploded on the stand in Mojave in 2006, killing three people. Nitrous Oxide is hardly favorable for its autoignition properties, and is famous for its insidious effects on the nervous system (making leaks that much more hazardous) so it has been ruled out for all Ascent Roadmap systems. Hydrogen Peroxide, however, is an excellent reaction control system monopropellant, so it may yet find itself on Ascent Roadmap boosters.

Propane: Propane was briefly considered at the recommendation of MicroLauncher competitor Charles Pooley (who has missed the mark on many more important topics, such as the specifics of the microspace revolution which is now passing him by, and the difficulty of engineering pico-scale interplanetary spacecraft.) The reason why propane was considered, and then only for the Prochron booster, was its vapor pressure. Its volatility means that it will clean up its own spills by boiling away, and its vapor pressure may be usable for feeding it as a propellant. Its inherent vapor pressure is a disadvantage on any scale larger than a single Prochron modules (which is about the same size as the MicroLauncher first stage) because it creates a large, flammable and invisible cloud of gas as it boils off, one that can also displace the oxygen from the air around launch crew workers, killing them even without a spark. Kerosene has a high flash point, and for this and other reasons, a ground covering is highly desirable anyway. Spilled fuel is not necessarily lost under these circumstances. A concrete or firebrick pad (probably both) protects the soil beneath, and, if the pad is not large enough, or if the rocket is being launched from a stand without a pad, a tarp will be used to protect what will almost certainly be a grassy fuel from both fuel spills and burning debris from the Prochron first stage (the only stage in the whole Ascent Roadmap likely to produce any without a failure.) As far as vapor pressurization goes, propane in thermal contact with liquid oxygen through a common bulkhead will not develop enough pressure to keep up with the liquid oxygen. This complicates the pressurization system to the point where kerosene is just as good as propane. With either fuel, either a supplemental pressurant supply is needed for the fuel, or the liquid oxygen tank needs to be vented to bring its pressure down to parity with the fuel tank.

What's currently under consideration:

Liquid Oxygen: Pennies a litre, easy to produce anywhere (on Earth at least), pressurizes itself, and has lots of experience in rocketry. The choice of liquid oxygen is actually very easy. There are only two real competitors as oxidizers for any booster, not just those for the Ascent Roadmap. Nitrogen Tetroxide, which is hazardous (and that hazard makes it expensive) and Hydrogen Peroxide, which would work, except that it is difficult and expensive to produce.

Kerosene: It is the cheapest fuel available, and it has nearly as much experience in rocketry as liquid oxygen. It is easy to store, easy to ignite, and relatively easy to pump with its relatively high density. Most of the competitors are either more volatile or very toxic, neither of which endear them to lower stage applications. All are more expensive, counting all the secondary costs of transportation, storage, transfer, and safety. All fuels (including liquid hydrogen) are currently derived from fossil fuels, so it is very likely that all of them will follow a cost trend somewhat parallel to that of energy prices in general, and petroleum products in particular.

Liquid Ethylene: Properties are very similar to those of liquid natural gas, and it may be possible to make stages that can operate with either fuel (that would be great.) Ethylene is produced either by a Fischer-Tropsh reactor or steam fractionation, making it relatively expensive compared to liquid natural gas (LNG) and especially kerosene. So why? Ethylene is backed by three Mars Sample Return studies as very likely the best fuel to produce on Mars. This is Ascent Roadmap borrowing the long view from the After Columbia Mars Direction. Liquid ethylene is not being considered for the lower stages of any Ascent Roadmap concept. Upper stages are smaller, therefore more similar to the Mars launch vehicles that will be taking off in near-vacuum in less gravity, providing useful experience for them. Being smaller also makes the all-new oxyethylene propulsion developments more affordable, since ethylene has no experience in rockets at all.

No comments: