I’m back! (Back! Back.) And for the first of many new scienceposts, I bring you a quick précis of what I just spent the last week doing. I feel somewhat tawdry doing this since it’s duplicating my space school research more than a little bit, but what the hell – it’s my research*, and since I had to aim the task at eleven year olds I didn’t get to use all of it.

We’ll start with this really great chart featured on wikipedia’s page on space colonisation.

Delta-v is the fancy space scientist way of saying “change in velocity”. Objects that are stationary with respect to the thing you are trying to reach or get away from have zero delta-v. A spacecraft sitting on the launchpad on Earth has zero delta-v with respect to Earth orbit. To get it up to LEO, we are going to have to change its delta-v by around 10 km s^-1 to accelerate it to the orbital velocity of 8 km s^-1 required to stay in low Earth orbit. Where do the additional 2 km s^-1 of delta-v come from, if we only need to get it up to 8 km s^-1? That’s extra delta-v needed to counteract the effects of atmospheric drag and gravitational forces during the spacecraft’s ascent. Delta-v is basically a measure of the amount of “effort” needed to move from one orbit to another – or one place to another – on the part of the spacecraft engine, which typically burns propellant to create reaction mass to produce delta-v.

The thing about delta-v is that it’s not an absolute measure. It takes no notice of spacecraft attributes like payload weight or fuel weight which make the amount of thrust required from the spacecraft engine different from spacecraft to spacecraft. It tells you nothing about the specific amount of effort you’re going to have to make to get a specific spacecraft into space; in order to get that we’d have to take these specific spacecraft attributes and multiply them by the delta-v required in order to get a specific required thrust value unique to that spacecraft. Once we’ve done that we can’t then take that thrust value and apply it to another spacecraft because its weight, engine, structure etc. will be all different. We’d have to recalculate the thrust value using that spacecraft’s own attributes.

What delta-v does do, though, is provide a quick and easy scalar quantity that allows spacecraft designers to estimate how difficult a particular orbital maneuver will be relative to what it’s already done, or has yet to do. The above chart is a summary of the delta-v budget required to get to various places in the solar system from other places in the solar system. By looking at it we can see that getting from Earth to LEO is rather hard, requiring a delta-v of 9.3 km s^-1 (yeah I rounded up before, so sue me), but that getting from the LEO altitude of just 300 km up all the way out to the Moon requires a total delta-v of just 4.8 km s^-1. In other words, no matter what spacecraft you’re dealing with and how heavy it is, it’s always going to take twice as much effort to cover the first 300 kilometres from the surface of the Earth to LEO as it is to travel the next 380,000 km out to lunar orbit.

The reason I like this chart so much is because it illustrates a key restriction on travelling around the solar system that I’ve brought up more than a few times before on here: by far the hardest part of space travel is getting out of the Earth’s gravity well. Once you’ve done that you can get about on relatively little fuel because the subsequent delta-v requirements are comparatively low, but while the Earth provides a rather pleasant living environment for human beings – it doesn’t burn us, freeze us, crush us, poison us or suffocate us, unlike literally everywhere else in the universe – being stuck at the bottom of its gravity well is actually a crippling handicap as far as space travel is concerned. When the amount of energy you have to expend travelling those first 300 km is only slightly less than the amount of energy you’d expend getting from there to freakin’ Mars, you know you have a problem.

 As long as spacecraft manufacturing and launch facilities remain Earthbound – as they are likely to do for the foreseeable future – human exploration of the solar system and beyond is always going to be held back by this massive delta-v requirement to get out of the gravity well. Earth’s gravity is not something that can be counteracted, cancelled out or removed; as long as we launch stuff from its surface we’re always going to have to use these massive, inefficient rockets to provide that large initial delta-v investment. The only thing we can do about it is move somewhere else, building and launching our spaceships from locations which are a little kinder in terms of delta-v requirements such as geostationary orbit or even the surface of the Moon. As you can imagine, creating the sort of offworld industrial base required to do this is going to be a bit of a stretch considering we can hardly claim to have conquered even LEO with the ISS.

Some additional notes about the chart: “C3” is another way of saying escape velocity, so the Earth C3 and Mars C3 points on the chart are the Earth and Martian escape velocities respectively. The reason the Mars C3 point comes right after the Earth C3 one is because you can follow the chart back in the other direction, using it to go from Mars to low Mars orbit to Mars escape velocity, and that allows you to get a rough value for the amount of delta-v required to get back to Earth.

 Another crucial point about delta-v is that you need it to slow down as well as speed up a spacecraft. Accelerating up to the Martian escape velocity is no good whatsoever if you don’t have the fuel left to slow the spacecraft down again once it arrives back at Earth. Fortunately, if you happen to be going to – or even passing by – a large planetary body with a sufficiently thick atmosphere, you can slow the spacecraft down effectively “for free” expending just a bare minimum of fuel by using a maneuver called aerobraking. Here we make atmospheric drag forces work for us for once; by adjusting the spacecraft trajectory so that it skips through part of the planet’s atmosphere, the drag forces that arise from this will slow the spacecraft down for us with no further effort required on our part to change its velocity. The amount of time it spends within the atmosphere depends on how much we want to slow it down; if we want to stop it entirely then we just do a stock re-entry procedure. This free delta-v is one way only – you can use it to get from LEO back to Earth, but you can’t use it to get from Earth to LEO – and that’s denoted on the chart by the giant red arrows.

Finally, now that we understand how delta-v works it should be obvious why we use the quantities of impulse and specific impulse to describe rocket engines and their thrust efficiency rather than force or energy or any of that jazz. Thrust exerts a force on the spacecraft according to F = mass times acceleration. To achieve a certain change in velocity we’re going to have to accelerate it over a sustained period of time — so to get to 1 km s-1 we accelerate it at 1 m s-2 for 1000 seconds or something — and since impulse equals force multiplied by the time you apply that force for, once you factor out the spacecraft mass (so divide the thrust force F by m) you end up with acceleration multiplied by time, and that is just another way of saying delta-v.

*By which I mean “It’s my shameless cribbing from Wikipedia.”

The post To Vee Or Not To Vee. appeared first on The Scientific Gamer.

And now, the thing that indirectly led to last week’s post on Race Into Space: the Kerbal Space Program.

KSP has been in development for a while now. I first played it this time last year, when it was a free alpha and had a grand total of two rocket engines, two couplers and one capsule. Even then its potential was clear, because even though that version was the most basic bare-bones version of the concept, the concept just happens to be making a pseudo-accurate simulation of building and launching your own space rockets. It’s fantastic. No, really.

These days you have to pay $15 for the latest alpha version (although the old one is still free). You totally should, though; a year of development has added two moons you can land on, an improved UI that indicates atmospheric pressure and your projected orbital — or escape – trajectory, and a whole host of new rocket parts including RCS thrusters, jet and ramjet engines, wings, control surfaces, landing struts, fuel lines… it’s a very long list and the game isn’t even close to being finished yet, with a slowly-expanding BARIS-style space centre that remains mostly non-functional at the moment while the developers nail down the sandbox element of the game.

In Kerbal Space Program you are the omnipotent designer, builder, launcher and pilot of the Kerbal race’s Heath Robinson-esque attempts to escape the gravity well of their home planet, Kerbin.  You start your spaceship design with a command capsule containing three suicidal Kerbal astronauts. As your first rocket you might decide to keep things simple by adding just a fuel tank and an engine of some kind underneath the capsule; the builder is fairly intuitive with the modular rocket parts automatically snapping to pre-determined connection points, so this wouldn’t take more than thirty seconds or so. Then you rush it out to the launch pad to see what happens.

Now, I can’t speak for what will happen to this simple rocket design, but I can tell you what would probably happen. You’d ignite the engines, increase the throttle, lift off – and then nosedive into the ground because you forgot to include any control surfaces or the automatic SAS stabilisation system to keep the rocket stable as it ascends through the atmosphere. Or if that doesn’t happen, you burn your engines too hard and they overheat and explode. Or if that doesn’t happen, you run out of fuel well before you get anywhere near orbit, the rocket starts dropping back towards Kerbin, and you suddenly realise that not only is there no way to separate the command capsule from the main body of the rocket, but that even if there was the three Kerbal astronauts would be doomed anyway because you didn’t include that most basic piece of spacecraft equipment: a parachute,

There’s a lot of ways you can fail in KSP, and seeing just what kind of unanticipated disaster will befall your spacecraft next is half of KSP’s fun, especially when the failures are so spectacular. It can honestly take a bit of work just get a spacecraft design off the launchpad when they have a penchant for falling to bits if not structured correctly. Rocket engines that run on liquid fuel need some way for the fuel to get to them, either directly or through a fuel crossfeed. Somebody new to the game might be tempted to attach some solid fuel rocket boosters onto their spaceship, but while these have the great virtue of being simple they also  have the slight drawback of essentially being an unstable lump of explosive material stuck to the side of your rocket. Even if you manage to get it off the ground you have a hell of a task ahead of you just controlling your heading, orientation and fuel burns to get into orbit. KSP is a game where succeeding in your chosen goal for the first time – suborbital, orbital, moon landing, whatever — does not come easily. However, behind the succession of comedy failures lurks a surprisingly deep iterative learning process which mimics the way real space programs are developed.

(Except those tend to have less dead astronauts, obviously.)

KSP is a game that – for now, at least – fosters experimentation. At the moment it consists of a basic sandbox mode where all rocket parts are free and you can have as many as you want on your spaceship, while pile-driving your three astronauts into the bottom of a burning crater carries no penalties besides having to take that particular rocket design back to the drawing board. You are free to try as many launches as you want in order to make it work. And slowly, gradually, you’ll start to figure out what you should and shouldn’t be doing. It becomes apparent that more engines do not necessarily equate to more thrust when your burn all six of them at once and your rocket only gets about a foot off the launch pad, and that experimenting with explosive decouplers that separate a rocket into discrete stages is far more efficient. How many stages do you want, though? How many boosters per stage? Do you really need that tri-coupler there? Fine-tuning this, as well as the fuel/payload ratio, is what occupies you during the first couple of dozen launches. Eventually, though, you’re going to refine your design to the point where it’s capable of escaping Kerbin’s gravity well, and it’s at this point that you encounter a whole new challenge: getting into orbit.

This is not as simple as you might think. Orbit is often described as “falling without hitting the ground” and hey, falling’s pretty easy, right? It’s just a matter of giving yourself enough sideways velocity so that you fall towards the planet you’re trying to orbit at the same rate as the ground falls away from you due to its curved surface. Surely, then, orbit must be a matter of going up high enough and then turning your spacecraft ninety degrees to the side and making a full burn of all remaining engines.

Uh, no. Spacecraft kinematics are actually really difficult to get your head around when you’re trying to manage throttle and keep the thing pointed in the right direction without exploding like a giant firework. I did manage to make a successful orbit using this very stupid method – even got the astronauts back to Kerbin safely afterwards – but it looked like this:

A highly, highly elliptical orbit with the periapsis about 100km above the surface of Kerbin and the apoapsis out beyond the orbit of the sodding moon. It took over a day to complete one orbit. My problem was I went up to a height I thought would be sufficient to fall from without hitting the ground and then made my sideways burn, but I forgot that a spaceship isn’t a car and that it still had a vertical velocity of over 2km s^-1. The sideways burn didn’t help either because it was still pointed slightly up, hence the ridiculous orbital altitude.

Clearly I have some way to go before I can even think about attempting a moon landing. Here, again, I have some inkling that it’ll be far trickier than I think because if you point a spaceship directly at the moon and blast away for 48 hours then by the time you get there the moon will have moved on in its own orbit and you’ll have missed. A successful moon landing requires you to get yourself into an orbit that crosses the moon’s orbital path, and then for the two orbiting bodies – spacecraft and moon – to occupy the same area of space at roughly the same time. The patched conics system that makes up KSP’s map makes this a little kinder on the prospective lunar explorer (although it has been done without by a crazy person, which incidentally demonstrates how much easier it is to get out of a moon’s gravity well compared to a planet’s gravity well) but it’s still a challenge that is going to result in lots of failures, and hence lots of hapless Kerbal astronauts floating helplessly around the solar system inside their steel coffins.

KSP is something that everyone should try, I think. That alpha I played last year is still free, after all, and you’ll get some idea of what it’s like to try to launch a rocket into orbit. The full (well, fullest) version adds spaceplanes to the mix, giving you a whole new type of craft to ram into the ground at 5000 km h^-1, and the next release seems to be adding in EVAs for the little Kerbals giving them a physical presence in the game beyond the three perpetually-terrified portraits in the lower right-hand corner of the screen. It says something about Kerbal Space Program that I’m currently trying to figure out some way to tie it into genuine science education; it’d be excellent for demonstrating why staged rockets are superior, along with a lot of other basic concepts of rocketry. It’s nowhere near a 100% accurate simulation, but then I don’t think I’d want it to be. KSP successfully walks the fine line between player enjoyment and technical verisimilitude without ever falling off of it, and that, I think, is its greatest accomplishment. Kerbal Space Program  makes rockets fun.

The post In Praise Of: Kerbal Space Program. appeared first on The Scientific Gamer.