If you're interested in learning more about this kind of thing, you want to look stuff related to "ullage".

Fun fact: it doesn't actually take that much time without forwards acceleration for ullage to become an issue. For example, the Saturn V included solid-fuel rockets on the interstage between the first and second stages to provide enough forwards acceleration for the second stage engines to ignite, even though the gap between the first stage engines shutting down and the second stage engines igniting wasn't more than several seconds [0, page 14]. You can see the exhaust from those motors on the outside of this video of the Apollo 4 first stage separation [1].

Another fun fact: instead of using smaller motors to provide acceleration between stages, (some?) Soviet rockets ignited upper-stage engines while the lower-stage engines were still firing ("hot-staging"). This is why some soviet rockets had a lattice (or lattices) between stages instead of solid metal.

Yet another fun fact: it doesn't take much acceleration to sufficiently settle the fuel for ignition. The Saturn V's S-IVB stage used its equivalent of its RCS system, which is normally used for attitude control, to settle its fuel prior to firing for trans-lunar injection.

If you're interested in installing KSP mods, RealFuels includes ullage simulation, so you'll need to have some method of providing acceleration to get larger rocket motors to ignite reliably.

[0]: http://web.mit.edu/digitalapollo/Documents/Chapter5/saturnas... [1]: https://youtu.be/iJIY4n5oXjc?t=8

Edit: Fix description of hot-staging for Soviet rockets. Some Soviet rockets hot-staged their third stage and not their second stage, making the description incorrect.

Another fun fact: you can replace the need for ullage acceleration with a Propellant Management Device that uses surface tension among other things to hold propellant enough to fuel an impulse that will settle the rest of the tank.

http://www.pmdtechnology.com/Index.html

For very low accelerations, they can even draw propellant "upwards" against the acceleration.

Ha brilliant post, I always wondered about those lattice structures, I figured they were for mass saving, thanks!

This classic clip of the Saturn launch shows three ullage rockets firing before the main (liquid) engine kicks on:

https://youtu.be/jPDkagD50XU?t=95

One thing I don't know if it is correct is that, on KSP, I've managed to deal with ullage using the overpowered reaction wheels (essentially using centripetal force). Pretty sure that would not work properly in real life, with real materials...

It's physically possible, but it almost certainly wouldn't be practical. The mass needed for a reaction wheel of the right size would probably make it too expensive, never mind the control and mechanical issues of dealing with an accelerating spinning rocket.

Yeah, I don't think a rocket would like this lateral acceleration at all, even if we replaced the reaction wheels with something less far-fetched.

Oh, but KSP does teach you this. It just requires a mod: "Engine Ignitor". Then you have to care about the fuel sloshing about, ullage, limited number of ignitions depending on engine type, things like that.

https://forum.kerbalspaceprogram.com/index.php?/topic/168424...

If you watch some of the old webcasts you can see shots of the LOX tank in 0g transitioning to firing. The trick is to use another source of thrust (e.g. cold gas thrusters) to accelerate the rocket slightly and settle the fuel.

I've always wondered if you couldn't do the same thing by spinning the rocket, and a spiral in the tank. Advantage: doesn't require any reaction mass. Disadvantages: requires a reaction wheel (might be complicated engineering), and spiral in the tank might impact fuel pumping/flowing performance.

Spinning rockets are extremely hard to control due to precession/conservation of angular momentum. Any change in attitude would introduce a wobble. Unless you make your tanks spherical and rotating independently of the rocket motors on a gimbal you will have trouble with changing orientation and compensating for minor offsets of thrust.

Ah, you're right. You'd need to spin until just the moment when you start the engines, and then stop pretty much instantaneously, which in practice would cause all kinds of instabilities and problems...

Spinning the rocket sends the liquid to the sides, not the bottom where the pumps are.

That's why you have the spiral inside the tank, so that it "scoops" the fuel to the bottom.

Not if you spin the rocket end-over-end

But then the rocket isn't consistently pointing the engines in the right direction.

Depends on how fast you need to spin it. If you had a reasonably long period, the lighting of the engines and the cancellation of the angular velocity could occur fast enough that the engines would act like they were consistently pointed in one direction.

I think a bigger issue might be that all the mass is at the engines, so the engines would stay nearly fixed with the other end of the rocket rotating rapidly, and pretty much all the propellant would go to the wrong end of the rocket.

That doesn't sound right. I would expect that the centripetal force will send the fuel to the end of the rocket (or ends, depending on where the pivot point is).

And how do you spin up the rocket?

Using a reaction wheel.

https://en.wikipedia.org/wiki/Reaction_wheel

Basically, an heavy enclosed wheel that you use a motor to spin. Because of the conservation of angular momentum, the rocket will spin in the opposite direction (more slowly, because it's (probably) heavier).

Not going to work at that scale.

SpaceX is using the RCS (Reaction Control System, https://en.wikipedia.org/wiki/Reaction_control_system) of their rockets for this job (i.e. as Ullage motor, https://en.wikipedia.org/wiki/Ullage_motor).

This shows the contents of a tank (I think it's the oxygen tank as there's no transfer tube in the middle of the tank as it would be in the RP-1 tank (that's short for "rocket propellant-1" a sort of refined kerosine - and yeah, very imaginative name!)) sloshing around during SECO-1 (Second Engine Cut-off 1) during the CRS-4 mission.

https://youtu.be/7YkCh7uOw1Y?t=1442

It does, just not the vanilla version. If you want realistic fuel and engine behavior, you should try with RealismOverhaul mod, where ullage and engine restarts are critical parameters during vehicle design and mission planning.