SpaceX on landing a rocket under power

posted in: JustPlainCool | 0

In case you missed it a few weeks ago, SpaceX performed something never before done.  They launched a rocket, with a customer’s payload, and brought the first stage — which is the majority of the whole enterprise that actually gets the payload from Earth’s surface to orbit — safely back to ground.  Not only down; the first stage landed, under power, where they wanted it to.  Why is this important?  Why is this hard?

First, two videos SpaceX has made available.  The first is one they just released, which is an edited version with some very cool camera angles of the landing.  The second is the more or less unedited entire launch video.

Now, why is this hard?  Why is it such a big deal that a rocket landed?

For starters, the first stage, the ‘rocket’ if you will, is the most expensive part of the launch system.  The engines and fuel delivery systems that feed them require some serious rocket science engineering, and none of that is cheap.  The first stage requires a lot of very precise and extremely high tolerance construction so it performs properly; everything has to withstand high Gs, high temperatures, shocks, and so forth.  It’s very expensive, and quite wasteful to have to toss it aside after each use.

At this point, you might wonder why don’t they just strap some parachutes or something to the thing if they want to recover it safely.  Well, that has been considered and tried throughout the history of spaceflight.  You can do a lot of reading about it if you like, but the short answer is what we’re trying to safely land is heavy.  Very heavy.  And the parachutes, heat shields, glider sails, or anything else that might work to suspend and lower all that weight are also very heavy.  Every pound you add to a rocket launch adds multiple pounds of fuel that’ll be necessary to successfully launch the mission into orbit.  I can’t find a convenient metric right now, but adding three or five hundred (or more, I think, was what they actually found out) pounds of parachutes was so expensive and detrimental to the launch profile that it’s not a workable solution.

So that’s the “why we want to” of landing a rocket under power.  At this point, some of you might still be wondering “why is this hard.”  Ah, that’s the rub.

To help sort of illustrate the difficulty of this, I’ll turn to Kerbal Space Program.  I became aware of this ‘game’ after this XKCD strip appeared:

Explaining Kerbal Space program
Explaining Kerbal Space Program


I’d heard Kerbal popping up here and there for a while, but that strip really got me wondering.  I know Randall Munroe is a pretty intelligent guy.  The concept of a ‘game’ being more illustrative in understanding orbital mechanics than actually being employed at NASA required I take a closer look.  So I did.

Now, ask anyone “how do you launch a rocket”, they generally tell you the rocket has to go up.


I know, it sounds stupid.  After all, we talk about “sending rockets up” and “going up to space” and stuff like that.  Those are the terms we use.  And yes, space is above us here on the ground on Earth.  But to “send a rocket up”, the hard part isn’t actually sending it up.

This is what I found when I started screwing around with Kerbal.  I figured out how to build a simple rocket and get it launched, and how to control it, and then I started trying to get into orbit.  I failed.  Repeatedly.  Some of the earliest failures were things like losing control of the rocket and letting it veer and yaw off course, breaking itself up under the wild stresses of being so out of control.  But once I got that stuff more or less sorted out, here’s what I kept having happen to me when I tried to “send a rocket up”.  This:

Bad Rocket Launches

Yes, I know that picture is fucking horrible.  I’m a writer, not a visual artist.  But it should be enough.  My rockets would launch, rise through the atmosphere, and then (at some point) I’d run out of fuel and crash back down.  The only really useful thing I learned during this period was how to make sure my parachutes for safely landing the crew capsule worked.

Eventually I gave up and did some reading.  Okay, getting into orbit requires going fast enough around the center of the gravity well (e.g., Earth, in this case) to stay up rather than crashing down.  So I tried some more launches.  My trajectories shifted to things like this:

Bad Rocket Launches-2

So all I’d really accomplished was to start crashing in the Atlantic or over in Europe rather than somewhere close to Florida.  Let’s keep trying, I thought.  After all, while there is a lot of real (though simulated) physics in the game, it is a game.  One doesn’t have to be a full fledged rocket scientist to figure it out; right?  Well, I hit a wall after a while and retreated back into research.  This time I reached for Youtube and Let’s Plays of Kerbal.  I found a few videos, but I quickly found Scott Manley.  Any of you who are already familiar with Kerbal are also very likely already familiar with Scott as well.  He’s a great Kerbal player with a lot of charm and personality, but also is smart and an actual rocket scientist.  Or, at least, he has degrees in things like Physics, Astronomy, and specialized subsets of general Physics.

Scott was the hero I needed when I got tired of bashing my head against the “how do I send a rocket up” problem.  Here’s the video I found where he both explained and demonstrated the process:

Beginner’s guide, yup that was me.  So he repeated what I’d read in my lookups about Orbital Mechanics; to stay in orbit you have to go fast.  Yeah, I know that.  I can’t figure out how to build a rocket that has enough fuel to get up and then go sideways fast and rapidly enough to stay up in space before it crashes and/or runs out of fuel.  And then he demonstrated the “ah ha” moment.  Kerbal players call it the gravity turn, and as far as I know it’s what actual rocket scientists call it too.

See, to launch into orbit, the hard part isn’t going up.  It’s staying up.  And to stay up, you have to go around and around the center of gravity — the planet or moon or whatever — faster than gravity can pull you back down.  By doing this, you are “in orbit” and don’t crash.  Launching things into orbit does this by basically launching sideways.  That is, when the rocket launches, it doesn’t go up so much as it’s going off sideways at something like a 45 degree angle off vertical.  I don’t know the actual angle, and I do know it changes a lot at various points during the launch.  For example, the first few moments of the launch will actually be vertical, but then it’ll start tipping over and throwing thrust towards going more and more sideways until it’s basically thrusting fully sideways at a completely perpendicular 90 degrees to the ground below.

The extremely elementary mistake I’d been making, in my complete ignorance, was the sideways part.  And when I learned about the sideways part, I figured you got up into space and then went sideways.  That’s probably theoretically possible, but it would be extremely inefficient; requiring an overbuilt rocket with way more fuel than is needed to simply launch properly in a sideways direction while still way down deep in the atmosphere.  Scott explained it and demonstrated, and I was able to duplicate it and finally “send rockets up”.  That stayed up until I wanted them to come down.

Which, incidentally, was a whole other learning experience.  You’re in orbit, and want to come down, so you thrust down, right?  Well, you can, but that just means you’ll enter the atmosphere going really fast sideways.  Wind resistance, friction heating as a multi-tens-of-thousands-per-hour speeding capsule hurtles through the air; not a great idea.  I had several capsules break up or explode until I figured out “oh, you thrust backwards to your orbital direction; not down.”  Because you have to kill your velocity, which automatically drops you down, and is much safer than just going down.

And, one final note; having now fiddled around with Kerbal, XKCD was not kidding.  I’m not, nor will I ever be, a physicist.  But I’m a nerd, I’m a smart guy, I have a lifelong love of science and science fiction and all sorts of things like that.  I’m a guy who will read Wikipedia entries just for fun when the mood strikes me.  But it took actually trying to guide rockets from pad to space in a real simulation to really “get” how it works.  When someone like Randall Munroe says, even jokingly, that his understanding of orbital mechanics catapulted upon playing Kerbal, I believe it.  Have absolutely no trouble at all believing that at all.

But let’s bring this all the way back to SpaceX and why it’s so hard, and such an accomplishment, that they landed the rocket successfully.  As we all know now (or already did and have spent the past paragraphs laughing hysterically at my failings) you’re really launching sideways-not-up when you launch a rocket.  And fast; the rocket is getting into some multi-mach speeds to get its payload into orbit.

When the first stage separates, to land it that first stage has to kill all its sideways velocity.  And then add some more backwards velocity to come back to where they want it to land.  And then, finally, correctly apply thrust to kill all the sideways and downward velocity so it can lower itself gently and correctly down to the landing pad where it will land and not fall over or come down so hard it breaks up or explodes.

That’s what SpaceX has demonstrated they can do.  The rocket launched, and separated.  The second stage with the payload (real payloads, 11 Orbcomm satellites were successfully deployed into their correct orbits by this mission) went up — er, sideways — into space.  The first stage flipped and reversed direction, thrusted further to put itself on a downward arc back to the landing pad, flipped again to orient thrusters down to the pad, zeroed all of its lateral velocity, and came down on the pad without crashing.  And then shut itself down and sat there like nothing’s nothing.

Personally, the unedited video with the room full of people cheering, that gets me right in the feels.  It did on the day of the launch, and it still does when I look at it again.  I’m not entirely sure who everyone in the room is, but I’m guessing most of them very probably are SpaceX employees who were part of the braintrust that figured all this out.  Who crunched the numbers and scienced the science and built the rocket.  Who made history happen.  When they go nuts with sheer unbridled joy at that rocket calmly landing itself on the pad, I just fucking love it.  People in sports stadiums cheer wildly for stuff that’s basically unimportant, like a goal or a touchdown or a great catch or whatever; this was a crowd going wild for something that truly counts.  And it’s just a “feel the feels” moment for me.  Watch the video, look at their faces.  Look at how incredibly happy and excited they are for this!  What a great moment.

Incidentally, SpaceX has had time to run engineering surveys of the landed stage.  It’s ready to be fueled and sent back up.  This one probably will never fly again; it has enormous historical significance.  Hopefully it’ll end up somewhere like the Smithsonian to inspire generations to come.  But the first stage is ready to be reused.  SpaceX has said they’re going to run some live-fire tests on it, which will mean putting fuel in it and firing the engines.  And then they’ll survey it some more.  Because science, obviously.

But it’s been done.  It’s not a theory, it’s not a wish, it’s not a dream.  The rocket can be launched and recovered, checked over and refueled, and then put right back into use.  This drops the cost of getting from the ground to space considerably.  And the major hurdle to getting into space is cost, not science.  We’ve known how to get into space for decades.  The reason it doesn’t happen more often is cost.  This advance chops a huge chunk of change off the bill to put people and stuff, anything, into orbit.

Science rules!  We’re in the 21st century damnit; it’s well past time we started acting like it!