How to design heavy lifter rocket

What exactly makes lifter rocket Heavy?

It is impossible to give precise cutaway point for heavy rocket, but there are several properties you might notice in them:

• cargo size, its big, its ridiculously big. Its not space station module, its whole station. In single launch. 60 tones is quite big, but not yet spectacular.
• thrust become more important than fuel efficiency when you launch Titanic sized payload.
• thrust become more important than aerodynamic shape. When engine is big enough, even barn doors will fly.
• staging stops being advantage. Weight cost in structural elements for humongous rockets can be higher than few empty fuel tanks.
• you start using cheatsengineering to combat game limitations. (Engine clustering, part clipping)
• you realize part mods are very useful.

What do I need to design good Heavy Lifter rocket?

• Imagination
• Engineering mods
• Part mods
• Moar boosters

Imagination - you are leaving area of sane engineering, there is no connection to real life rockets anymore. Think big, think boosters.

Engineering mods - Knowing how to do rocket math is good. Doing rocket math on your own is a waste of time. KSP "physics" is quite different from real world, so real tools are mostly useless. Numbers and equations are bothersome to extract from game and online tools tend to be obsolete. However there are several mods that will greatly help with engineering:

• Kerbal Engineer Redux - DeltaV, thrust ratio, dry mass, atmospheric and vacuum readout, and more.
• MechJeb - Humans are inherently imprecise controllers. Precise maneuver nodes, precise maneuver execution. Comes with quite good ascent autopilot, which you can further edit on your own. Do not fell bad for using it, we use autopilot since Sputnik ;)
• RCS Build Aid - Allows for correct usage of RCS without wild guessing its positions.

Note: All suggested mods are either working in KSP 1.1 or will be updated shortly, are in active development and work fine on 10 year old laptop ;)

Part mods - At certain size default parts are just too small. Engines too weak. Fuel tanks too small. High part count is heavy hit toward structural integrity (which affects in-flight stability) and greatly affects FPS (fly-by-hand is even more imprecise when game lags).

• SpaceY Heavy Lifter Parts - Big parts for big rockets.
• SpaceY Expanded - Even bigger parts for even bigger rockets.
• Modular Rocket Systems - although it contains mostly small parts, there are very useful structural parts inside. And Quad-NERVA engine for your large transfer stages.

Note: KSP 1.1 (currently in open beta test) performance is greatly improved over current KSP 1.0.5. All suggested mods are balanced against default parts (no sci-fi magic parts here).

Rocket building rules

Lets reiterate over standard rocket building rules:

• Reduce dead weight by multiple stages
• Match your engines to atmospheric pressure
• Build with aerodynamic shape in mind
• Keep your max TwR below 2:1
• First up, then gentle gravity turn when air becomes sparse

However...

At certain size aero-braking (it occurs on ascent as well as on descent) stops being effective. Long story short, it is exposed_surface * atmospheric_pressure * rocket_speed versus rocket_speed * rocket_mass. Rocket mass grows faster than its exposed surface so after breaking 100 tons air stops being a concern. And rockets tend to leave low atmosphere rather rapidly anyway.

Additionally big rockets will benefit more from stable thrust than from fuel efficient engines, so this point can also be ignored.

Lets build a stock rocket!

Note: You will need very high thrust unless you will perform aggressive gravity turn. I picked gravity turn as question implied Career mode, where funds are limited.

Hint: Do not bother with solid fuel boosters unless using bigger parts (eg. SpaceY mod)

So lets lift 60 tons. With huge error margin. Here is quick and dirty design:

1. Define your test cargo with some error margin.
2. Add strong, yet simple, central stage that will burn all the way to orbit, 2000m/s DeltaV is good bet.
3. Add boosters that will increase DeltaV over 3000m/s and will be dropped quite soon.
4. Add some struts, solars, SAS modules and wings. (SAS/winglets can save fuel over control with engine gimbal alone)

Under Mechjeb control (aggressive gravity turn, no concerns for aerodynamics) HL 60t can reliably reach 100x100 orbit with ~3200m/s DeltaV, which leaves ~200m/s spare for error correction (~1000m/s used for ascent, 2200m/s in orbital velocity). Payload fraction is around 16%.

Note: I planned to turn boosters into drone-shuttles capable of controlled descent, initial tests were promising.

Lets optimize our rocket!

Note: do not bother with limiting throttle in low atmosphere (Limit Q in MechJeb). For rocket of such size it will actually yield worse result than full throttle all the time.

Its too strong! 4:1 TwR is a bit high, maybe we do not need Rhino burning all the time? When Rhino got ignited after boosters were detached, rocket reached orbit with 350m/s in tank. We learned that we have too much fuel (dead weight) and too much thrust (dead weight in engine mass).

Not using engine is making it into dead weight, so lets shuffle some thrust from boosters to center (Rhino = 2xVector, Mamooth = 4xVector). We can also try to turn boosters into fuel tanks for central stage with fuel lines.

After switching engines it will reach orbit with 190m/s in tank. We also dropped price from ~350k to ~160k. (1xMamooth+4xSkipper)

Can we go further? Yes, yes we can.

By performing gravity turn we remove need for turning rocket around rapidly, and after many test flights we know this design is very stable. No need for SAS and wings. No SAS means no need for solars. By increasing boosters thrust we can decrease main stage thrust (5xMainsails this time).

Price dropped to 150k (20k payload, 130k lifter alone), 370m/s left in orbit.

Spare fuel means dead weight. We can also replace those nice looking boosters with old and boring orange fuel tanks.

Over 300m/s left in orbit. So we can downsize a bit more.

Puny 50m/s left in orbit. Cost dropped to 127k (20k test payload, 107k lifter). Payload mass ratio increased to ~19%

And here we will end after ~40 launches (wouldn't bother doing it without MechJeb and ALT+TAB ;). Lifter price dropped well below rewards from missions requiring such rocket and as a bonus we dropped to tech level 6.

Hint: If you plan to launch manually, better add some spare fuel. Thrust margins should allow for ~10% increase in Delta-V.

Note: Record payload mass ratios for rockets in KSP are around 25%.

As always, designs can be downloaded from my SE.Gaming KSP repository.

I got the idea to add the lifters below the payload rockets after watching Scott Manley video. And thanks for the great hint from @PTwr to empty the fuel tanks

Here is the solution:

• Empty the tanks of your payload. Refuel them later when they are already in orbit.

• Asparagus is great to send the ship above 30Km before starting to turn the ship. Doing that before will cause the ship to lose control.

• T1 "Dart" did not give enough torque. The best fit here was RE-I5 "Skipper".

• Structure to give enough space between one fuel tank to another. Otherwise, during the separation, they may (will) touch the engines and destroy them.

• Finally, a lot of EAS-4 Strut Connector or the payload will wobble like crazy.

NOTE: In this very specific case, the science module is a reusable detachable part, so it should be delivered separated from the rest of the body. It would spare an extra 2.5 tons from the entire payload.

Empty the fuel tanks of the last stage and launch it into a low Kerbin orbit.

Then launch a second rocket which is just a large fuel tank, dock it with your first stage and use it to fuel it.

You can also launch and attach additional fuel tanks to it. You can then drop these one after another during the mission as soon as they are empty.