Design Tips for Intermediate Players

[Aegolius13]

After a few years of hanging around the Gameplay Questions forum, I’ve collected some general tips that may help intermediate players get a little more out of their rockets.  This guide is geared primarily toward extending the effective range of a rocket, but it segues into a few related ideas as well.  This is also focused on rockets rather than spaceplanes, though some (but not all) of the suggestions can apply to planes too.  I'm hoping to make this a work in progress, so let me know if there's anything you'd like to see added or expanded.  

Note: By “intermediate player,” I’m thinking of a player who’s gotten to orbit, has played around with most of the part types, is familiar with the basic jargon of the game (TWR, specific impulse, etc), and has some acquaintance with delta-v. 

 

The Rocket Equation: Live it, Learn it, Love it.  Do you want your rocket to go farther?  It seems like a complicated proposition, but thanks to the Tsiolkovsky Rocket Equation, it’s actually a pretty simple matter.  The equation solves for the delta-v (i.e., effective range) a rocket can achieve using three, and only three variables:

1.       Dry Mass: the mass of everything in your rocket other than fuel.

2.       Fuel Mass: the mass of the fuel in your rocket.

3.       Specific Impulse: the efficiency of your engines.

The first two can even be condensed into a single number called the mass fraction (the ratio of dry mass to "wet mass," which is wet mass + fuel mass), giving you only two variables to work with. 

So if you want to improve delta-v, the Rocket Equation gives you a few general guidelines:

1.       Decrease Dry Mass: this is arguably the most important point, and one I’ll touch on in greater detail below.  But cutting unnecessary fat can pay enormous dividends.

2.       Increase Fuel Mass: that’s a fancy way of saying “add more fuel.”   Simple enough, as long as the rest of your rocket can handle it.  But be aware that adding fuel runs up against diminishing returns pretty quickly, because that extra fuel equals extra dead weight up until the point it’s burned.  This is the core of the “Tyranny of the Rocket Equation.”

3.       Increase Specific Impulse: this just means selecting an engine with favorable ISP numbers for the task at hand.  Again, simple enough on the surface.  But (again as discussed below) ISP is just one of many factors you’ll have to balance, and can be a bit of a trap if over-prioritized.

There’s one more range factor that’s not expressed in the classic Rocket Equation, but obviously has a large effect on your ship’s range.  That’s staging, and while it’s an important topic, I’m not going to get too far into the theory here.

 

Design from the Top Down.   It’s almost always good design practice to start building your ship with the last stage, and work your way down from there.  Think about what your final stage needs to do (e.g., lift off from the Mun, return to Kerbin, and land safely).  Then build it with what you need to accomplish that task.  Then think about what you need in the prior stage to get the final stage where it needs to go.  And so on and so on, until your rocket is complete.  If you start from the bottom up, you’re likely to end up with stages that are too big, or too small, than needed for what comes next. 

 

Keep it Simple: This goes back to the “reduce dry mass” point above.  It is almost impossible to overstate how large an effect dry mass can have on your ship’s delta-v.  I don’t want to get too technical, but a very significant fraction of your ship’s delta-v comes when your fuel level is low, at which point the rocket is very light, and every unit of fuel burned can produce a lot of acceleration.  Keeping dry mass light really helps amplify this effect.  This is the major reason why real-life rockets can achieve much greater delta-v per stage than their KSP equivalents – their fuel tanks and engines are much lighter.

Reducing dry mass can pay other dividends as well.  You will get better TWR out of your engines, which at a minimum can improve maneuver efficiency, but can sometimes allow you to switch to lighter (or fewer) engines – further reducing dry mass in a virtuous cycle.

So with that as background, it’s always good practice to approach a rocket with the question of “what can I cut?” rather than “what can I add?”  I will get into a few specific examples below.  But I’d suggest starting with a clear concept of what a rocket’s purpose is, and dictating the design around that.   

-Leave whatever you don’t need.  If a part is not necessary for a mission (or reasonable contingencies), it should probably be left behind.  If a pilot will be on duty for the whole mission, you probably don’t need a probe core.  And you might not need an antenna either unless you’re transmitting science.  Your Mun lander doesn’t need a ladder if you’re comfortable using the jetpack (or a jump) to get back to the hatch.  As I’ll touch on below, you likely won’t need any RCS or monopropellant on many craft.   

-Don’t overbudget resources.  If you’ve got no major electrical drains, a small battery and a couple of the smallest solar panels may be plenty.  If your command pod has more than enough monopropellant for docking, you can probably leave that extra tank at home.  It’s still prudent to leave a reasonable safety margin, but if you find you’re packing 10x as much of something as you end up using, chances are you can slim down without any problems. 

-Look for lighter parts.  The game frequently offers “small, medium and large” options on various components.  I’ve talked a bit about engines already and will talk more below.  But other components like landing legs, reaction wheels, decouplers, and docking ports, try to use the smallest/lightest ones that your design will allow.

 

Don’t Get Too Impulsive: It’s very easy to get into the trap of overprioritizing specific impulse.  Like the MPG on a car, ISP seems like a relatively-straightforward measure of efficiency.  So higher ISP is always better, right? 

Try swapping engines on a very small craft, like a satellite, and checking out the delta-v readout.  A tiny engine like the Ant might provide better delta-v than the Spark, which might beat the Terrier, which might beat the Poodle – even though each successive engine has better delta-v.  This goes back to the discussion above about the deceptive importance of dry mass. 

And of course, you need thrust to put that delta-v effectively to work.  This is most obvious on the launchpad – if your TWR is under 1, you’re not going to move.  If it’s so low your rocket is barely moving, then gravity is going to eat up an excessive portion of your delta-v.  (Don’t forget, while going straight up, you lose 9.8 m/s every second to gravity alone).  Thrust is less important in space, but if it gets too low, your maneuvers will suffer and you’ll take some unavoidable efficiency losses.  So sometimes it’s worth trading a little ISP for better thrust. 

If you’re playing in career mode, then cost enters the equation as well.  And it’s often possible to save a lot of money by using less efficient, but cheaper, engines like SRBS, the Reliant or the TwinBoar.  Sure, your rocket might need more fuel to get to orbit, but who really cares if it ends up costing less?  .   

Also keep in mind that ISP is most important in your final stages, and least important on your launch stage.  Why?  The Tyranny of the Rocket Equation again.  Fuel is heavy, and you’re using fuel just to move fuel before you burn it.  Think about a multi-stage rocket like Apollo – the stages get exponentially bigger as you move down the rocket, for this very reason.  So saving a kilogram of fuel in your upper stages can save many kilograms down below, meaning the efficiency gains of a high ISP engine are amplified.  But the opposite is true at launch – you don’t have the booster's fuel that fuel very far before you burn it, so it's comparatively less critical that you burn that fuel efficiently.

We see these lessons in real life with the Falcon-9 and its Merlin engine.  It’s mediocre-at-best in the ISP department, but it’s cheap and packs world-record TWR.  As a result, once cost is factored in, it tends to beat the pants off fancy, expensive, high-ISP launchers like the Delta or Arianne.

None of this is to say that ISP is bad.  All other things being equal, it’s always better to have higher ISP rather than lower.  But the game purposefully offers you tradeoffs, and sometimes the other factors (thrust, cost, mass, etc.) should win the day.

 

Haste Makes Waste: It’s very easy to build a rocket with a lot more thrust than needed.  As mentioned above, you want enough thrust at launch to keep gravity losses to a reasonable number, but your thrust needs level off pretty quickly once you start building horizontal speed.  Once in orbit, you can get a with a TWR of .5 or less, and still get maneuvers done in a reasonable period of time.  And if you’re using low-thus engines like the NERV and Dawn, you may need to go significantly lower.  When you’re landing on another celestial body like the Mun or Minmus, be sure to check your local TWR figures.  You may be able to get away with a lot smaller of an engine than you’d think.

 

Improve Your Attitude: One of the more common “flabby rocket” issues I see is excessive attitude control.  The game offers several tools for attitude control at various points in your journey, including: engine gimbal, reaction wheels, RCS thrusters, and aerodynamic control surfaces (e.g., fins and ailerons).  With all these options, it’s very easy to end up with more attitude control than you need.  And in fact, due to the less-than-perfect physics and AI, excessive attitude control can actually lead to overcorrection and instability, making your problem worse instead of better.

At the point you launch, if you have even one engine with gimbal (and most do), you’re probably got all the attitude control you need.  So you probably don’t need to add a bunch of moveable fins or an RCS thruster system to your launch stage.  If you are depending on a non-gimbaling engine, you going to need an alternative.  Fins and thrusters are options, but there’s one good choice that’s easy to forget – verniers (i.e., small secondary engines with gimbal, like the Spider, Twitch and Cub).  Their stats may not look that great, but they’re surprisingly effective.

When you’re in space, more often than not, reaction wheels are more than enough to meet your needs.  In other words, you almost certainly don’t need RCS thrusters (or any monopropellant) for attitude control.  You want need RCS to provide translational motion for docking, but even that’s not necessary depending on your approach.  But as a general rule, if your ship isn’t going to dock with anything, you can lose this stuff without a second thought.

 

Drag is a Drag: On a well-designed rocket, the effect of drag is almost negligible.  Some parts of drag avoidance are pretty intuitive (pointy parts are good).  But due to the game’s unintuitive approach to drag modeling, it’s very easy to make design mistakes that hurt you far more than you’d suspect.

-Keep a clean profile.  Anything sticking out from the side of your rocket is liable to add significant drag.  Consider putting your accessories in a service bay rather than attaching them radially.  Use inline batteries rather than the surface-mounted versions. (There is at least one exception to this rule – the radial parachutes are better than the top-mounted ones). Or, if your payload is unavoidably draggy, put it in a fairing. 

-Avoid abrupt form factor changes. If there’s one thing KSP’s drag model really hates, it’s abrupt changes from one form factor to another.  (By “form factor,” I mean the various diameters of part sizes, plus special cases like Mk 2 and Mk 3 plane parts).  If you need to go from one form factor to another, use an adapter (either “structural” adapters or fuel tanks can work) or an interstage fairing. 

-Avoid fake occlusion.  Sometimes you might think a part is protected from drag, when it really isn’t.  Be careful when placing parts in cargo bays, especially if you attach parts directly to the bay’s walls.  Avoid the variable-length structural tubes, as they don’t even shield the nodes above and below them, let a lone anything inside them.

-Beware of secretly draggy parts.  Mk 2 plane parts look sleek but they generate outsized amounts of drag.  Struts are also draggier than their small size might lead you to believe.   

Finally, it’s worth it to run the numbers.  The game offers a couple vital tools to diagnose drag issues.  Under the console (alt-f12 on PC) > physics > aero, you can open an “aeroGUI” which, among many other datapoints, can show you how much drag your ship is generating at any time.  At the same place, you can also enable an option to see the drag on an individual part by right-clicking on it.  This is handy for picking out particular trouble spots like those “fake occlusion” situations.

 

Edited February 27, 2019 by Aegolius13
Clarified wet mass vs fuel mass per Tyko's comment.

[Tyko]

Great guide!

Note though, in the rocket equation - Wet Mass is the total weight of your fueled rocket. You have it listed as the weight of the fuel.

In aerospace engineering, mass ratio is a measure of the efficiency of a rocket. It describes how much more massive the vehicle is with propellant than without; that is, the ratio of the rocket's wet mass (vehicle plus contents plus propellant) to its drymass (vehicle plus contents).

Edited February 25, 2019 by Tyko

[Aegolius13]

On 2/25/2019 at 9:44 AM, Tyko said:

Great guide!

Note though, in the rocket equation - Wet Mass is the total weight of your fueled rocket. You have it listed as the weight of the fuel.

In aerospace engineering, mass ratio is a measure of the efficiency of a rocket. It describes how much more massive the vehicle is with propellant than without; that is, the ratio of the rocket's wet mass (vehicle plus contents plus propellant) to its drymass (vehicle plus contents).

I'm using the term loosely (maybe technically incorrectly), but I think there's a helpful distinction. I.e., you can increase "true" wet mass by either adding fuel or by adding other stuff, but they're going to have very different effects on delta-v.  Maybe "fuel mass" is a better term for above?  

[Tyko]

 

Here's a another explanation with some wording that would be more common: In aerospace engineering, the propellant mass fraction is the portion of a vehicle's mass which does not reach the destination, usually used as a measure of the vehicle's performance. In other words, the propellant mass fraction is the ratio between the propellant mass and the initial mass of the vehicle. 

 

Edited February 27, 2019 by Tyko

[Tyko]

So....there are other errors in your info. I don't want to pile a lot of critiques on unless you want them.

If you want this to live on as a useful guide it's probably worth getting a review by members of the community though, be that me or others.