Calculator for true gravity turn parameters (FAR/1.0)

[conklech]

By gravity turn, I'm talking about a flight path that goes vertical for some distance, executes a single pitchover maneuver, and then maintains a zero angle of attack (follows prograde) at least until exiting the atmosphere.

Choosing the correct pitchover parameters is pretty important in FAR, and presumably in the 1.0 aero model, because you're then effectively locked into that trajectory; for normal rockets, once you pass a few hundred m/s, maintaining an AoA beyond a few degrees is impossible. If your pitchover is too aggressive, you'll never escape the atmosphere; if it's too vertical, you'll incur potentially-huge gravity losses.

Furthermore, if I understand correctly the trajectory is in large part determined by the thrust curve of the launcher, which obviously involves a number of variables. A pitchover maneuver for one launcher is not likely to work for another.

For these reasons, it's been my experience that FAR launches involve either a lot of trial and error for each new launch platform, or throwing a lot of extra delta-v at the problem at taking a conservatively high trajectory. Okay, maybe that's the Kerbal Way; but...

I see too modding-type challenges here; I don't know if either of them has been addressed. One is to reliably execute a pre-programmed pitchover maneuver, which doesn't seem that difficult. (It's not incredibly hard to do manually, but there's a big payoff for precision.) The second, which I'm more interested in at the moment, is to compute the correct pitchover.

It seems that this should be a solved problem in the actual aerospace industry; I haven't yet done enough research to see. It'd also be nice to calculate a good value for "how much delta-V do I need to get into X orbit" rather than the existing rules of thumb.

Is anybody working in this area? If so, I'd like to help (within my relatively limited means).

[conklech]

Now that 1.0 has hit, a gravity turn calculator would be particularly relevant. The MechJeb thread has been spammed with a lot of attempts to get its curve-following ascent guidance system to work, with mixed success. An accurate pre-calculation would make things much more reliable.

[Frederf]

The phrase "gravity turn" is thrown around in KSP rather loosely. A gravity turn is not the most efficient ascent. All it is is minimal steering loses and at high enough speeds, minimal AOA. Only through exceptional design attention are the zero steering loss ascent and the minimum fuel ascent equal. Imagine two rockets of different CG-CT displacements but otherwise identical engines, fuels, weights, etc. The solution for a gravity turn on each would be different but their optimum "fuel to orbit" profiles would be the same. This should show that gravity turn and fuel to orbit optimizations are not the same question.

The cost associated with a non-gravity turn in fuel are rather small, a percent perhaps. This is perhaps enough to buy a new car on a real launch but few KSP players are optimizing to that degree. A huge real life motivation is the strength of the rocket structure which does not cope well with aerodynamic side loading. KSP's rockets are infinitely more robust than their real counterparts and routinely take AOA of 10-30-90 degrees in stride where more than 2-3 would collapse a real article.

It seems like a nit picky distinction but these nits are the size of buffalo at this specificity. It would be embarrassing to receive exactly what you asked for only to realize you asked for not quite the right thing.

The best you'll get out of Squad is some kind of "sandbox" wind tunnel simulator whazzit that you can use to try various procedures yourself and maybe find a better way. It's firmly against the design philosophy of KSP to output "answer" optimizations on behalf of the player.

[Somtaaw]

the usual answer to an "efficient" route to orbit is usually by chucking a crapton of attempts until you get a route that works on the majority of craft. So in the pre-1.0 days it was ascend vertically to 10-14km, pitch directly to 45 degrees until whatever, and then directly to 90 degrees and start circularize burns. It wasn't the absolute tightest efficiency, but it was optimized for nearly all rockets.

It's still possible to use the same route and even with the same rockets, I took a rocket Scott Manley used in his "Landing on Minmus" video from December 2012. I also took it on a 0.25-0.90 ascent pattern, not only did it still work... I made it to Minmus as well.

Sadly it was not enough to actually land, but I did manage a very low <5km flyby and then got stuck on a very eccentric Kerbin orbit for the next 6 months until I could loft a properly built "rescue" shuttle. But it still worked getting out of the atmosphere, and no real changes were needed, and that was on a pre-0.25 version of the game.

TLDR the older ways and rockets still work if you're careful. It just lost a touch of efficiency, not enough to bother most people, just the folks who shave the margins close.

Edit: just reliase you probably didn't know about the older routes players used, but yeah. The optimized thing, is just keep trying different variations on routes. If you blowup or fail during launch, restart and try it again a few times, try different things.

Edited May 1, 2015 by Somtaaw

[rkman]

The cost associated with a non-gravity turn in fuel are rather small, a percent perhaps.

In 1.0 least d-v to orbit is in the range of 2800m/s, vs some 3800m/s for a 'classic' stock ascent profile.