# Why are optimal launch TWRs so different in KSP than on earth?

In Kerbal Space Program (KSP), optimum thrust-to-weight ratio (TWR) is around 1.5-3. When looking at NASA, SpaceX, Blue Origin spacecraft, the launch TWR is around 1.25-1.3.

Is it because the atmosphere is much larger, and an optimal Goddard solution is achieved higher in the atmosphere?

• Define "optimum." Commented Oct 6, 2016 at 2:27
• @RussellBorogove I would assume he means delta-v expended to reach orbit. Commented Oct 6, 2016 at 20:22
• That's not generally what real world rockets are optimized for. Commented Oct 6, 2016 at 20:49
• Short answer: game is "wrong". Get some mods: Realism Overhaul, Real Solar System, Real Fuel and so on. Then in-game numbers will be similar to real world.
– PTwr
Commented Oct 7, 2016 at 13:41
• What are you trying to optimize, expelled ∆v to get into orbit, launch vehicle cost, or something else? Commented Nov 28, 2016 at 1:20

It is no longer generally considered optimal to use a launch TWR of 1.5-3 in KSP. A little background: In KSP players who were optimizing for launcher weight would go for a TWR of 2.0 or higher, so players wishing bragging rights of "I launched 100t into orbit for only X tonnes of rocket" would go for a TWR of about 2.0 or even higher and this worked because a high TWR is extremely effective at mitigating gravity drag, a rocket with a TWR of 1.25 loses 80% of its thrust to gravity, while a rocket with a TWR of 2.0 loses only 50% of its thrust to gravity, a much greater fraction of the thrust is thus available for building velocity rather than fighting gravity and that makes the rocket significantly more efficient in terms of weight.

But as KSP development continued, career mode was released and the cost of rocket parts was balanced. At this time cost per tonne to orbit became a new and more legitimate metric for judging rocket performance. In KSP when you want to build a rocket which launches payload as cheaply as possible, often a launch TWR of around 1.25-1.3 is exactly what you want to use - at least for rockets which aren't using SRBs.

Now, why is this the optimal TWR for making cost-effective rockets both in KSP and IRL? What it comes down to is rocket engines are expensive but fuel and tankage is cheap, relatively speaking. Thus it is desirable to pile almost as much fuel onto the engine as it can lift to get maximum burn time out of the engine.

If in real life, there was some incentive to make launch vehicles which were as light as possible rather than as cheap as possible they would also have a higher TWR.

• Fantastic, well-thought-out answer. Commented Dec 23, 2016 at 4:52
• Seems to me like reusable launch vehicles may be included in that. Commented Feb 8, 2020 at 4:57

There are also other considerations that real spacecraft need to deal with, like gee loading. An initially high TWR will mean that your TWR at MECO will be even higher, potentially causing payload damage. Also, a lower TWR means MaxQ will be at a lower pressure. In real life, engine burn times are also much longer (usually) than in KSP, and generally that relates to a lower initial TWR.

• Are you assuming that the ship cannot throttle down its engines? Commented Dec 21, 2016 at 21:03
• @OrganicMarble: Why go with super-high-TWR (and expensive) engines if you're gonna throttle them anyway?
– SF.
Commented Dec 22, 2016 at 14:01
• So you don't overstress the payload? Commented Dec 22, 2016 at 14:34
• @OrganicMarble: How would weaker (and cheaper) engines overstress the payload?
– SF.
Commented Dec 23, 2016 at 1:12
• Many engines are not throttleable, at least not to the extent that KSP shows. For instance, the Merlin 1D engine that SpaceX uses is only rated to throttle down to 70% (though they've tested down to 40%), and the 1C before that had no throttle capabilities. The SSME engines could operate anywhere between 67% - 109%. To limit acceleration on Saturn V launches, the center F-1 engine was shut down 26 seconds before MECO (giving an effective thrust of 80%) Commented Jan 3, 2017 at 18:00