So How Does A Rocket Fly Sideways, Anyway?

It’s often said that getting into orbit is less about going up, and more about going sideways very fast. So in that sense, the recent launch conducted by aerospace startup Astra could be seen as the vehicle simply getting the order of operations wrong. Instead of going up and then burning towards the horizon, it made an exceptionally unusual sideways flight before finally moving skyward.

As you might expect, the booster didn’t make it to orbit. But not for lack of trying. In fact, that the 11.6 meter (38 feet) vehicle was able to navigate through its unprecedented lateral maneuver and largely correct its flight-path is a testament to the engineering prowess of the team at the Alameda, California based company. It’s worth noting that it was the ground controller’s decision to cut the rocket’s engines once it had flown high and far enough away to not endanger anyone on the ground that ultimately ended the flight; the booster itself was still fighting to reach space until the very last moment.

Astra’s rocket on the launch pad.

There’s a certain irony to the fact that this flight, the third Astra has attempted since their founding in 2016, was the first to be live streamed to YouTube. Had the company not pulled back their usual veil of secrecy, we likely wouldn’t have such glorious high-resolution footage of what will forever be remembered as one of the most bizarre rocket mishaps in history. The surreal image of the rocket smoothly sliding out of frame as if it was trying to avoid the camera’s gaze has already become a meme online, arguably reaching a larger and more diverse audience than would have resulted from a successful launch. As they say, there’s no such thing as bad press.

Naturally, the viral clip has spurred some questions. You don’t have to be a space expert to know that the pointy end of the rocket is usually supposed to go up, but considering how smooth the maneuver looks, some have even wondered if it wasn’t somehow intentional. With so much attention on this unusual event, it seems like the perfect time to take a close look at how Astra’s latest rocket launch went, quite literally, sideways.

Perfectly Balanced

Everyone knows that sending a rocket to space is hard, but unless you’ve read up on the Tsiolkovsky rocket equation, you might not realize just how true that statement really is. Even if everything goes right, which is already a lot to ask for, the physics that govern chemical rockets are particularly brutal. Put in the most simplistic of terms, the “Rocket Equation” governs how much of a rocket’s liftoff mass is dedicated to propellant versus what it’s actually carrying. In the aerospace community, this is often expressed as a booster’s payload fraction.

The Space Shuttle only managed a payload fraction of around 1.5%, while the SpaceX Falcon 9 is capable of slightly more than 4%. Astra hasn’t provided detailed specifications for their current launch vehicle, but given how small it is and some back of the envelope math, we can assume it’s operating at a payload fraction of not more than 1%. We can further estimate that, at the time of liftoff, the vehicle’s thrust-to-weight ratio (TWR) must be quite low.

The destructive failure of one of the rocket’s engines, captured in just four frames from the live video.

Which is precisely what we saw in the live video from the August 28th launch. Right after ignition, one of the rocket’s five first stage engines failed. This reduction of thrust brought the TWR of the vehicle so close to equilibrium that, instead of climbing vertically, it entered into a remarkably stable hover. As luck would have it, the failed engine happened to be the one diametrically opposed to the launch tower, which meant the asymmetrical thrust of the remaining engines moved the vehicle laterally away from the launch pad. Had one of the other engines failed, there’s an excellent chance the rocket would have struck the launcher or some other piece of ground support equipment.

The rocket, now clear of any structures and traveling horizontally, was still consuming propellant at an incredible rate. At around five seconds after engine startup, it had burned through enough fuel and oxidizer that the mass of the vehicle was once again below the total thrust of the engines. With a positive TWR, the vehicle began to climb away from the launch area, albeit at a far lower rate than normal. During the live stream, you can see the obvious confusion of the camera operator, who expected the rocket to be much higher at this point in the flight and needed to pan back down to keep it in frame.

As the rocket climbed higher and faster it started to look like a fairly typical launch, and after about a minute of flight time, you’d never have known how close the vehicle came to being obliterated at liftoff. Its reduced thrust and skewed ascent meant it would never make it all the way to space, but it certainly made a valiant effort.

Just before the mission elapsed timer clicked over to 2:00, an astonished member of ground control could be overheard on the live stream saying that the rocket was actually approaching its nominal downrange trajectory; a sentence that seems destined to become an Easter egg in the next Kerbal Space Program.

A Truly Remarkable Failure

It’s easy to look at this launch and see it as a major setback for Astra. Indeed, the company’s stock price plummeted nearly 20% when trading started on Monday morning. But while clearly not the ideal outcome, the reality is, things could have gone far worse. For one thing, the fact that the launch site was completely unscathed is nothing short of miraculous. When an Antares rocket exploded a few seconds after liftoff in 2014, it took nearly two years to repair the launchpad at a cost of approximately $20 million.

The team at Astra was also undoubtedly able to collect valuable data about the vehicle’s performance during this truncated flight. Remember, the rocket technically never failed. It was still flying strong, though not necessarily on the correct flight path, when the command to activate the flight termination system was given. At that point the rocket had flown for nearly two and a half minutes and reached an altitude of 50 kilometers (30 miles), an impressive accomplishment for a fledgling company that lofted their first vehicle just two years ago.

Finally, this event is definitive proof that whoever is handling the guidance and navigation systems on this rocket is well worth whatever they’re being paid. An avionics package that can not only keep the vehicle balanced vertically while it’s translating laterally but eventually find its way back to its prescribed trajectory afterwards could only be designed by the most steely-eyed. If there was ever a question as to whether or not you could build a rocket-powered Segway, it seems the fine folks at Astra just answered it for us.

43 thoughts on “So How Does A Rocket Fly Sideways, Anyway?

    1. Same here. I don’t quite understand why the stock price went down by 20% instead of up by 50%. :) Clearly these guys have an amazingly experienced team of software developers. Given the right means, this team is going to go far.

      I mean, this flight control software seems ready to actually land the vehicle if they so wanted. Not sure how long Astra already exists, but it seems their software is already close to the level of that of SpaceX’s software.

      Ok, they could also have been lucky, of course. :) But I think not.

      1. Because investors are not rocket scientists, what they saw was $millions in capital explode without a (visible to them) positive result.

        We, as rocket enthusiasts and technically minded people see all the things that went right in the context and go wow. They see an explosion and go Aw.S**T — There goes a whole bunch of money down the tubes and our investment is probably a waste. Let’s sell it for what we can, quick.

        IMHO, this is a good opportunity for smarter investors to acquire the stock, but I’m in no way an investment advisor. Heck, experience would suggest that I’m not even a particularly good investor.

  1. “Naturally, the viral clip has spurred some questions. You don’t have to be a space expert to know that the pointy end of the rocket is usually supposed to go up, but considering how smooth the maneuver looks, some have even wondered if it wasn’t somehow intentional. ”

    We can kind of blame SpaceX for setting our expectations about “unusual” rocket behavior. Before we would have blamed it on aliens.

  2. There was only a minor glitch at the beginning of the flight. While it’s a complex mess to understand exactly why this happened, it seems exceptionally clear to me that this company will be successful in the future. If not for the faulty start, the rocket would have made it to orbit. Astra is turning out to be the next SpaceX.

      1. “ What would you expect? ”

        A civil response to someone who asked a fair question about a subject they’re not an expert in. The fact that they asked about gimbaling shows they thought about the problem.

        Keep thinking and asking, Ren.

        1. Well said Queeg and thank you. The only purpose of a pissy comment such as that made – “What would you expect” – is to show how stupid the recipient is and thereby how truly wonderfully intelligent the person who made the comment actually is. I know who I’d rather have as a mentor or a friend.

  3. The fact that an engine was able to explode without taking out the rest of the engines is also a testament to rather spectacular engineering (or incredibly good luck, hard to know which and might be both) by Astra. One need look no further than what happened when one of the SRBs on the shuttle exploded due to bypass of frozen O-Rings and the chain reaction thereafter to realize just how unlikely it is that an exploding engine would not cause significant (catastrophic) damage to the other nearby engines and eventually the entire vehicle.

    1. What’s really pissy on that one was they watched it burn for a while before it took out a propellant tank. Had THEY aborted before failure the incident would have been an annoying faux Pas instead of a RBFD.

      1. Not necessarily. If I remember correctly, the shuttle was in a window during that time where they could not abort safely no matter what. Combo platter of you can’t separate the SRBs while they are burning at full, and the shuttle doesn’t have enough energy to recover and fly to any kind of safe landing. Remember, the shuttle had a glide ratio somewhere between a brick, and a brick with control surfaces. The shuttle at that point had full fuel, supplies, satellite payloads, not the configuration it typically landed in. Basically, they had to sit and hope for the best, by design unfortunately.

  4. “Ladies and gentlemen, due to a minor issue with this launch pad, we will be proceeding via surface routes to an alternate launch pad where we will continue our liftoff. We apologize for the inconvenience and we thank you for flying Astra.”

  5. When the engine failed / exploded, did fuel continue to be pumped to it?

    Consider: If you have, for example, 5 tons of fuel, for 5 engines to use in 4 minutes.
    Suddenly, for whatever reason, you have only 4 working engines.
    Could the fuel be redirected, allowing the remaining engines to thrust for an extra minute each?

    1. Nope. Pressure drops and valves close. And if the flight wasn’t terminated, the 4 remaining engines would have burned longer. But because the thrust-to-weight ratio at launch was very close to 1, the rocket burned fuel ‘for nothing’ for a bit, and that made it not have enough fuel to get to orbit.

      This is why rockets go sideways after they’re out of the thickest part of the atmosphere. They need speed to reach orbit, and gravity fights against that speed. So they change the flight vector to compensate for gravity while going sideways as fast as possible. This is more efficient than going straight up:

      https://en.wikipedia.org/wiki/Gravity_loss#Vector_considerations

      (this is also why rockets have a thrust-to-weight ratio of 1.1-1.5, you need to get out of the atmosphere and gravity well quickly, but not with too much weight for rocket engines)

      1. It’s not just the vector and gravitation losses that dictate that it go at an angle but also (and probably more so) the simple fact that if you want to achieve orbit, you have to go sideways. If you went perfectly vertical (ignoring the rotation of the earth), you’d come back down right where you started. Instead, you need to follow a ballistic path that has a horizontal non-zero vector component, like an ICBM would do. The only reason ICBMs hit the ground is that they don’t achieve as much speed. But you don’t fire them straight up, either (not after the first few seconds, anyway.)

        1. In theory if you went straight up for 35,000km you would be in geostationary orbit and could stop there. In practise to “go straight up” (staying above the same spot on the Equator) to that height would require a lot of lateral speed anyway.

          1. Nope. At geosync, you need to be moving sideways at about 6750mph. Try tying an elastic band to a ball and then hold the other end and throw the ball out to one side. It just comes back and hits you in the face. Now take the same band and ball and start spinning around and throw the ball out. The ball stretches out the band and remains out at a distance but if your hand is moving around a circle with your body at the middle and thus going a distance of 18′ every rotation, then the ball that’s 3′ away from your hand is going a distance of 36′ every rotation (roughly speaking.) Since they both rotate at the same RPM, the ball has to go twice as fast sideways as your hand or else it’s STILL going to come back and smack you in the face. But the band will also be wrapped around your throat first so you’ll probably black out and die before you get the black eye.

            So no matter WHAT sort of orbit you want, you have to impart sideways motion. The earth can assist with that as you can take off as close to the equator as possible and then can angle out to the east to take advantage of the initial 900mph or so sideways momentum that you started with (Florida is the most southern you can get in the continental US so gives the best initial start, not to mention that if something happens and the rocket has to abort, there’s a lot of empty ocean to fall into.)

        2. In theory, if you go “straight up” for any definition of “straight up” that means a trajectory that keeps the back end of the rocket pointed directly at earth and the rocket accelerating in a direction perpendicular to the tangent of it’s current location over the earth’s surface, there are two possibilities…

          One, you fall right back where you started (adjusted for wind, coriolis, etc.).
          Two, you eventually escape Earth’s gravity well and go off into space.

          Unless there is some altitude above geostationary where one can achieve orbit based on the lateral momentum imparted by earth’s rotation at the launch site, orbit just won’t happen by going straight up.

          1. I was ignoring the whole escape velocity thing as we’re talking about going to orbit. And ya gotta remember that all an orbit is is a satellite that’s constantly falling past the center of gravity of the other object and then “being flung” back up and then falling again. So if you went straight up and then started the fall, then there’s only one way you manage to go past the center of gravity of the earth (with some sort of BFG that can make ya a hole through the center before you hit the ground.)

          2. (I am no expert, will likely be obvious) Since this is a continuum, it seems to me that between your “One” and “Two” there must be a point when the spacecraft is not falling back and not escaping. When that point (geostationary or not) is reached attitude thrusters could be used to maintain it … What am I missing? (I have no clue if it is technologically possible to have a TWR greater than 1 in such a “vertical” launch)

          3. This is actually in reply to Marc’s question.

            If you hit escape velocity, that means that no matter how long of a time passed, you’d never return to the planet (we’re going ignore the other planets, the stars, etc that screw up the calculations.) If you don’t quite hit that, then, given enough time, eventually you’d return. Yes, there’s theoretically a middle point but if you hit that, it simply means that given an infinite distance and infinite time, eventually you’d wind up at 0 velocity relative to the earth. But you’d be infinitely far away from the earth.

            You could maintain your position with thrusters, sure. You could do so at 1″ off the ground or 100 miles off. But you run out of fuel eventually. The only stable flight path is an orbit around the earth and that (again, ignoring air friction, other objects, etc) path would keep you orbiting for all time. But you need a sideways motion vector in that orbit.

  6. Fortunately in the brief period before the engine explosion, it managed to lift the nozzles clear of the base ring of the launchpad. Had it failed a fraction of a second sooner the nozzles would’ve hit the inside of the base ring. Dunno if the flight control software could have handled that. Most likely it would have tipped over, smashed down and exploded.

  7. How to clean salt water for cheap and at scale (and manage salt)? how to provide food in a sustainable way? how to reduce pollution ?

    I prefer that we spend time and effort on theses questions. Rocket science…. not my priority even tho it gave us satellite but voilà, it is already a solved problem – enough until we have fixed the other critical ones..

    1. They are problems for other companies, governments and NGO’s to solve.
      Astra? They’re a rocket company, and they’re solving rocket problems. Well, a bit at a time, and they’ll either get there or bust. So I say Good Luck to them.

    2. We need to be ready for many different eventualities. The things you mention are very important, but we need to be ready for the next steps and hope that our technology is ready when we need it. Just because we don’t need it now doesn’t mean we might not need it in the future.

    3. All the things you mention need monitoring. The best way to do this monitoring, both technically and economically, is from orbit. Smaller satellites have begun to take over these jobs. The Astra launcher is aimed directly at this market. Such launchers will make the cost of small sat launches cheaper, and thus reduce the cost to farmers and others who need the data. Farmers now rely on better weather forecasting and crop monitoring afforded to them by various satellite services. They can watch the weather building in their region and select the best times to plant and harvest. They can also monitor the health of crops.

      This is not an either or situation. By reducing the cost of getting both monitoring and science into space, it becomes cheaper al round. Astra and the others in this market sector are actual working on the very things you talk about! You need to adjust your priorities.

    4. Your comment, which amounts to ….. Let’s not do this because we could do that instead because it’s better / more useful / more ecological / closer to what I think is right ….. is generally not a solution to anything except living up trees and eating raw veg for breakfast. Oh and picking the nits off our neighbours. For example, let’s not use precious trees for making paper to make books so that we can transmit knowledge more easily so that we can treat disease and learn how to grow crops better rather. You can’t see what will be more useful in the future and neither can I but what I do know is that without knowledge we are all stuffed.

      1. The trouble is Dave our disease treating is getting too good, we are in a pandemic and barely 0.05% of the population has died from COVID. Compare that with the good old Black Death which wiped out maybe 50%. So demand for food will continue to soar with the burgeoning population.

        We’ll still be vulnerable to the ‘big one’ i.e. an asteroid/comet impact that was big enough to wipe out most of the fauna on the planet. We need to be off this rock and populating somewhere else as a backup system. So – go gung ho at space exploration in whatever form it takes.

        Sorry to nit-pick! :-)

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.