Why Are There So Many Rocket Designs?

August 20th, 2020

2020, despite all its drama and misgivings, has been a monumental year for space exploration. Three separate space programs have launched missions to Mars this past month, including two - China and UAE - that have never successfully visited the Red Planet before; SpaceX, the private aerospace giant run by flamboyant CEO Elon Musk, became the fourth entity and first ever non-government organization to successfully launch humans into space; several space startups, including small-sat launch providers like Astra, Virgin Orbit, and Firefly, have or will attempt their inaugural launch to orbit this year, and other space ventures like SpaceX’s Starlink network and Virgin Galactic’s space tourism service are also reaching initial operational status. This buzz of activity is an exciting prelude to a decade of rapid commercial development in space.

After watching all these launches, one thing stuck out to me: there is such a large breadth and diversity of engineering designs utilized to put payloads into orbit. Rockets all roughly have the same basic cylindrical design, but variations balloon from there: some rockets only need one stage to reach orbit (known as SSTO - Single Stage to Orbit) whereas some, like the monstrous N-1 rocket pictured above, have five stages; some rockets are purely liquid fueled, whereas others utilize solid rocket boosters (SRBs) or entire solid rocket stages to boost them to orbit; even the rocket propellants utilized are far from uniform, with some burning liquid oxygen and RP-1 - rocket-grade kerosene - and others liquid hydrogen, liquid methane, or the hugely toxic UDMH-nitrogen tetroxide combination. Some decide to utilize a plane-like lifting body design, like the Space Shuttle and Virgin Galactic’s suborbital spaceplane, whereas others decide to stick with a more traditional booster design. For a class of vehicle that has such similar design goal - namely getting to orbit safely - engineers seem to have a wide range of solutions for the same problem. I started to wonder what are the key factors that affect the design of rockets from a high level.

Some might immediately point out that rocket designs varied because different rockets are meant to go to different places. Rockets designed to go farther afield are likely bigger than smallsat launchers destined for low Earth orbit, for example. This is true, but it only tells part of the story. The difference between the Saturn V and N-1 rockets illustrates this best: both rockets were designed to take humans to the Moon and have comparable payload capacities, but the N-1 has five stages to the Saturn V’s three, and the N-1’s first stage has a staggering 30 NK-15 engines to Saturn V’s five F-1 engines. Just visually the difference in engineering approaches is apparent.

A more contemporary example can be seen in SpaceX’s Falcon 9 rocket and other medium-lift launch vehicles, like ULA’s Atlas V and China’s Long March 2F. All three are capable of launching crewed missions to low Earth orbit (the Atlas V will be launching the Boeing Starliner as part of the Commercial Crew Program) and have roughly comparable payload capacities, but designers have taken different paths to solve the same problem. In terms of propellants, SpaceX utilizes RP-1 for both first and stage stage, ULA uses RP-1 for the first stage but liquid hydrogen on the second stage, and the Long March vehicle utilizes UDMH-nitrogen tetroxide for both first and second stage; ULA and Long March rockets also make use of SRBs, whereas SpaceX does not.

This illustrates the complexity of the question at hand that cannot be simply answered by categorizing rocket designs with their “intended destination,” even though it is a key consideration. There are three main mutually-interactive considerations of rocket designs that inform all other sub-considerations: payload destination, reliability and cost. Other factors, like reusability, fuel choice, launch location, launch method, and more, can all be examined through the lens of these three primary considerations.

Destination - Payload Capacity

Like I previously mentioned, destination is the most obvious consideration in rocket design. While it is not the only factor, payload capacity to orbit is doubtlessly the central consideration impacting rocket designs as the whole purpose of a launch vehicle is to have sufficient power to bring a payload to its intended orbit.

Rockets that need to launch larger payloads farther from Earth will require more energy, but it isn’t as simple as loading additional propellants to reach a higher orbit - the efficiency of each type of propellant and the engine that burns such propellant is equally important. A rocket that uses more efficient fuel and engine will be able to launch vehicles farther with less propellant than one that uses a less efficient fuel and engine, much like how the mileage of a car will vary with how efficient the engine is. In rocketry, the measure of an engine’s efficiency is measured by its specific impulse, which is typically measured in second. RP-1, which is rocket-grade kerosene, is readily available and easy to process, but it is a less efficient rocket propellant than a lighter fuel like liquid hydrogen. This is because how much energy a rocket obtains per molecule of exhaust is dependent on the exit velocity of the molecule - the lighter the exhaust product (hydrogen engines mainly exhaust molecular water, or H2O, which is lighter than exhaust from RP-1, which is mainly CO2 and other hydrocarbon molecules) the faster the engine can exhaust the molecules and thus the more energy it provides to the rocket on a per molecule basis. This is why ULA’s hydrogen-powered upper stage engine has a higher specific impulse than SpaceX’s more powerful but less efficient kerosene engine, allowing it to launch heavier payloads to other planets.

Notice how the Falcon 9 upper stage is still more powerful than the Centaur, however. Even though RP-1 is less efficient than liquid hydrogen and its engine will have a lower specific impulse, kerosene is magnitudes more energy-dense than hydrogen, so the SpaceX rocket will be able to store more RP-1 on its upper stage than liquid hydrogen on the Centaur (ULA’s upper stage engine), making up for its efficiency deficit for low Earth orbit missions. It is missions that fly farther afield that bring out the advantage of liquid hydrogen or other cryogenic fuels.

Destination also impact other considerations as well, such as launch method and location. Rockets that incorporate more reusable features, such as a reusable first-stage like the Falcon 9 or the space plane design of the Space Shuttle, will have to make sacrifices to its payload capacity as these features reduce the amount of fuel used for launch or increases the overall launch mass, both of which negatively impacts how much payload you can launch. For example, the Energia rocket, designed by the Soviet Union as the launch vehicle for their Buran space shuttle, can around launch 100 tons into low Earth orbit on its own, but the combined Energia-Buran system can only launch 30 tons in comparison due to the mass of the space plane itself. These are trade-offs that, in conjunction with reliability and cost, play into design decisions of every rocket that considers adding reusability capabilities.

Similarly, launch location also has an impact on payload capacity as different launch locations restrict the number of orbits a rocket can reach given its fuel. The most efficient launch location is at the equator, where a rocket can fully utilize the Earth’s momentum to reach prograde orbits (orbits that go in the same direction as the Earth’s rotation); this is why most launch sites that services such orbits are close to the equator (or as close as geography allows), though rockets heading for polar or Sun-synchronous orbits will have no such incentive as the orbits are perpendicular to the Earth’s orbit and will not benefit from an equatorial launch site. Stationary launch sites can only be so close to the equator, however, and several startups have designed rockets that are meant to be either sea or air-launched so they can be put into orbit from the equator itself from ships or planes, maximizing their payload capacity.

Fundamentally, it is the payload capacity to different orbits that is at the top of every rocket designers’ priority lists. However, as the SpaceX-ULA example has illustrated, there are many different ways to reach the same orbit, and one important factor considered by rocket designers in addition to payload capacity is reliability.

Reliability and Cost

This is a factor that should be fairly obvious as well - no matter how efficient or powerful a rocket is, it will have no customers unless it is safe and reliable enough to be used. This is especially true for crewed vehicles as human lives are at stake - NASA mandated SpaceX’S Dragon and Boeing’s Starliner systems to have less than 1-in-270 chance of loss of crew; the Shuttle was found to have around 1-in-90, or around 1.1%, chance of loss of crew per launch, which contributed to the downfall of the program.

However, reliability is also an inseparable factor from cost. The two are often considered in tandem when designing rockets, with an increase of one necessitating a sacrifice in the other. This is based on the common-sense economic theory of “reducing cost requires cutting corners,” which in turn reduces reliability. While this is true to some extent, newer players within the aerospace industry have demonstrated that cost-cutting is possible without compromise in reliability and, in many cases, can improve reliability. The two will be discussed in conjunction, however, due to the frequent association of the two in popular discussion.

Historically, reliability has been the main selling point for most traditional aerospace players like ULA and EU’s Arianespace, where rockets are designed with additional safety margins and techniques at the added expense of development time and cost. Tory Bruno, the CEO of ULA, has repeatedly emphasized that his company’s rockets are significantly more expensive than competitors like SpaceX due to their pristine and extensive launch record - known in the industry as flight heritage. This is important for launches of large and expensive payloads like satellites and exploratory missions: NASA went out of its way to acquire a European launch vehicle - the Ariane 5 - to launch its James Webb Space Telescope, due to the rockets reliable record and the monumental importance of the mission (JWST costs almost $10 billion and is the primary replacement for Hubble). The US military paid hundreds of millions of dollars in premium to ULA to launch their most prized satellites until newcomers like SpaceX and Blue Origin filed anti-competitive lawsuits against the force to open up the military launch market, arguing their rockets can reliably launch satellites at a fraction of the price. Regardless of the merit of these complaints, reliability is doubtlessly worth paying for and significantly impacts vehicle designs in various ways.

One way, as mentioned previously, is that traditional aerospace designs favor legacy, tried-and-tested hardware over more novel spacecraft designs. Space Launch System (SLS) is a massive, 111-meters tall behemoth that NASA is designing as its next generation launch vehicle after the Space Shuttle. It is designed to eventually carry over 100 tons into low Earth orbit and enable permanent human settlement of the Moon and, eventually, Mars. One feature of the SLS is that it looks remarkably similar to the Space Shuttle it is replacing; indeed, a large portion of the SLS Block 1 (the first version of the rocket) utilizes Shuttle-derived hardware, such as the RS-25 main engine (literally the same engines that previously flew on the Space Shuttle) and the SRBs (also previously used to launch Space Shuttles). Block 1 will also utilize a version of the Centaur upper stage found on the Delta IV rocket, built by ULA, and use RL-10 engines that have been in use since the 1960s. One may imagine existing hardware to be more affordable than newly developed ones; this cannot be farther from the truth, however. SLS costs $2 billion per year to develop and an additional $1 billion per launch, and have costed $17 billion by the end of 2020. While many of these design decisions are politically motivated - the SLS has been dubbed the “Senate Launch System” by its opponents due to Congress mandating the system to use legacy Shuttle hardware, effectively protecting Shuttle suppliers and their employment in key congressional districts - the utilization of hardware with flight heritage is a respected tradition within the aerospace industry. If it isn’t broke, don’t fix it, right? To many, SLS’ huge price tag is worth it as it buys reliability that is backed by decades of flight heritage from the RL-10 engine to the venerable Space Shuttle hardware.

Startups in the launch industry disagree. Aforementioned companies SpaceX and Blue Origin are designing rockets in a radically different fashion from traditional aerospace companies with the promise of producing safer products at more affordable prices. One such way to achieve this promise is placing an increased emphasis on the simplicity of design. This can be seen in the overall design of the Falcon 9 rocket - it is an incredibly minimalistic rocket. The first and second stage uses the same fuel and same baseline design, allowing both stages to use the same engines; the only difference between the two is that the first stage uses 9 Merlin 1D engines optimized for sea level, whereas the second stage only uses 1 Merlin engine that is vacuum-optimized. Falcon 9 also forgoes any strap-on SRBs, unlike previously mentioned Atlas V, Long March 2F, and SLS, given the added complexity of vehicle - each additional component is another possible failure point - as well as their non-reusable nature. Some may point out that the utilization of 9 engines on the first stage does not play toward simplicity as it provides additional points of failure; after all, the aforementioned N-1 rocket failed due to its complex array of 30 engines, whereas the Saturn V has a 100% flight success rate. This is a valid point, though this is where a compromise between absolute simplicity and relative simplicity with reliability in mind is warranted, as the 9-engine setup allows for one engine to fail and still result in mission success, as opposed to a one-engine setup. The rule of thumb then, it appears, is pursue simplicity as much as possible but over-engineer when necessary, with the former being the norm.

The quest for simplicity can also be seen in SpaceX’s decision to forgo isogrids when building the Falcon 9: isogrid is a type of structural pattern that significantly increases the strength-to-weight ratio of the material, similar to how grid-patterned steel is stronger than a singular I-beam. However, isogrids are incredibly expensive and complicated to manufacture, requiring precision machinery and wasting about 95% of the raw material. SpaceX instead utilized other tank-strengthening techniques like stir-welding stringers rather than expending resources on isogrids, reducing manufacturing complexity and cost. Thus, SpaceX illustrates how simplicity reduces potential failure points while reducing overall costs.

Another innovation that startups argue will improve reliability while decreasing cost is reusability. While SpaceX and Blue Origin, in addition to startups like Rocket Lab, incorporate reusable boosters as part of their core business model, more traditional players like ULA and Arianespace are only beginning to dedicate research and development efforts to the field. The hesitation from traditional players is part of a wider debate within the aerospace industry regarding the desirability of reusability. Many see reusable rockets to be technically feasible but much less economically feasible: the European Space Agency’s director general commented in 2018 that “reusability is fine from an ecological point of view. From an economic point of view, I don't know.” This is because while reusability undoubtedly lowers the cost of access to space, it is currently unclear whether there is sufficient demand for such access to warrant a radical shift in rocket design for these players. It should be noted that neither ULA’s upcoming Vulcan nor Arianespace’s Ariane 6 incorporate reusability technology, at least in the foreseeable future.

While the economics of reusability is a debatable subject (excellent subject for another article), it is undeniable that reusable vehicles make for safer vehicles, for a very simple reason: reusable vehicles can be retrieved and inspected, allowing potential problems with the rocket to be found and rectified for future designs. This is a point that SpaceX has repeatedly emphasized in their messaging and Musk has even tweeted about how it is cheaper to insure a Falcon 9 than any other rocket (though an industry insider called this misleading at best). This point may appear dubious, especially considering SpaceX has suffered two launch failures in the past decade, both related to the Falcon 9’s composite overwrapped pressure vessel, or COPV, that holds pressured helium used to pressurize propellent tanks. These failures don’t detract from the reusability argument, however, given the first failure was related to a defective strut and the second helium-loading conditions and unrelated to reused hardware. NASA has also given a vote of confidence in SpaceX by allowing it to reuse its Endeavour spacecraft on the Crew-2 flight to the International Space Station after previously mandating the use of a new vehicle, which was a decision likely taken after inspection of flown hardware from previous test flights Demo-1 and the Inflight Abort Test. NASA, as well as SpaceX’s long list of customers that utilized their reused boosters, stand as testament to the power of reusability in driving down launch prices without compromising reliability.

I have by no means exhausted the sub-factors that fall within these three main categories rocket designers consider. I simply hope to illustrate how the three can effectively encapsulate most, if not all, of these sub-factors. Balancing all of these factors is a fascinating puzzle in itself as well, and it is one that is tackled everyday by space programs around the world. At the end of the day, the end goal of every rocket is to push the limits of human engineering to reach ever greater heights and achieve ever greater discoveries, whether it is to a scientific, military, or capitalistic end. For me, it is simply exciting to see the innovation being pursued in the field due to the never-ending quest for reliable, heavy-lifting rockets at an ever-cheaper price.

2020 has been a remarkable year for rocketry, but much like biplanes from the 1920s, the industry still has a long way to go. It is incredibly exciting to see where we will be given a few more decades.