After the last chapter I received a comment from Luiz Eduardo, asking for more tips. I must admit I flew enough, read enough and became so good at getting into orbit that I don’t notice difficulties. But I see I did say how to calculate your 𝛥v, but not how to build a ship and how to pilot it. I decided to fill this gap with this special chapter.

Chapter 4. How to Build and Fly a Rocket into Orbit

Design Tips

Install Flight Engineer Redux and add its part to your ships if you haven’t yet. As we discussed on the forum, it does not add any supernatural abilties, but rather fills the instrument gap that KSP did not implement.

Here are the values that I kept in my flight engineer:


With this engineer and the advanced 𝛥v map you can build any ship you need. Still, there are many details of getting into orbit from the atmosphere.

Here are all design tips in one image:


Let me explain some tips:

  1. We need a lot of RCS fuel to correct the ship in ascent. RCS tanks weigh a lot, so I prefer to drop one with the ascent stage, and have my main RCS tank full after that. The easiest way is to lock it on the pad.
  2. Reaction wheels on the command pod do not do much but oscillate and bend the rocket. I turn them off too untill before circularizing.
  3. To cope with space debris, I prefer using an interplanetary stage, and the rest will just fall down on a ballistic trajectory. If this is not possible, then I add a pod, RCS thrusters and a solar panel to the last ascent stage. I’ll decouple and deorbit it.
  4. If you see own TWR and specific impulse of Skipper and LV-T30 engines, the T30 are superior, even in 4 and with an adapter attached. LV-T30 is better than LV-T45: despite having no vectoring its much lighter.
  5. For low-weight missions I prefer 6-8 narrow side tanks with LV-T30. 2 or 4 wide tanks on the sides, with bigger engines, increase weight a lot.
  6. TT-70 decouplers leave more space to add some side objects, but need struts to keep side tanks from bending.
  7. Because of poor aerodynamics model, a pencil shape does not reduce drag at all, so initial TWR of 2.0-2.2 is a good choice. More thrust with TWR > 2.3 will be lost in drag. At high altitudes you may have high TWR, but such engines are not necessary: just maintain vertical speed, and you have enough time to achieve horizontal velocity. I set my upper ascent stages to have TWR about 1.0-1.5, which is enough. The weakest I made was 0.5, but it had hard time to circularize.

Here are some screenshots from assembly building:


6 side stages in asparagus configuration.


Quadruple LV-T30


Solar panel to power the pod before deorbiting. The pod should stay either inside the stack, or, if not possible, close to the center.


I set an action group to turn torque off to not bend the rocket.

RCS thrusters are a big question and I made a simple diagram to understand where to put them:


Addition on 18.11.2013: After this article was published, I found out that the trick of putting an RCS thruster on a long lever was used in real life.

Flying Tips


Reality Check

To show that these advices really work, here’s a sample take-off with this münar rocket. Showing the best optimization and lowest possible weight, I have optimized it so that when I go too steeply, I can’t circularize, falling short by 50-100 m/s. I had to make gravity turn and steer a lot and level my rocket (i.e. burn horizontally) yet at 30 km.


Locking the top tank to keep its fuel for the mission.


I maintained terminal velocity when taking off, there was no big difference with going full throttle.


Still about terminal velocity.


Starting the gravity turn.


Slight turn by 15 km


45 degrees at 25 km


Maintaining time to periapsis at 40+ seconds. See that at 33 km of height, I already burn horizontally.


Next flights I just kept burning horizontally. Time to apoapsis lowered to 45-50, but then started to grow. So, no need to add vertical speed.


I got (falsly) worried that apoapsis wouldn’t raise and raised pitch a bit


Apoapsis raised to 75 km. At 47 km, drag is negligible.


Set maneuver node to be sure.




Some fuel is left and the top RCS tank is full.


Turned retrograde, decoupled from the ascent stage and moved upwards to let it pass.


Ascent stage is deorbiting itself.


It will now crash into Kerbin.


Münar ship is assembled and ready to depart.

When I tried going full throttle climb. After gravity turn I had much more vertical speed and could burn horizontally right from 30 km. It did save some fuel, but within human error margins.


Going faster at climb


Apoapsis is far enough, should lower pitch


Steering is not that fast


Finally burning horizontally


Apoapsis increases slowly but steadily


...and time to apoapsis grows even faster, which means we’re safe


...and then we stop the engine.


About 1/3 of the tank left, or 2.62 tonnes, or 397 m/s. After circularizing, about 600kg would be left, which is 100kg (20 m/s) more than in the previous example. About the same.

Next time I tried turning earlier and a bit quicker, and lost fuel. So this is the best I can get.

In reddit KSP community some enthusiasts tested different parameters and got tiny improvements. So this is almost all you can save. 4350 m/s of 𝛥v is about the best possible take-off budget with autopilot. An experienced human can spend about 4450 m/s, and a newbie may just arrange the interplanetary stage to circularize, and fly comfortably. It’s reasonable to say, we’ve reached the perfection.

Hope this was helpful. Next time I’ll sure write about the Duna trip! Fly on budget and enjoy flying!

Table Of Contents


comments powered by Disqus

Previous topic

Chapter 3. When Do We Need a Lander?

Next topic

Chapter 5. Planning the Duna Mission

This Page

Updates feed

See also

One Giant Leap
a photo gallery of Apollo program (with comments in Russian)