Gravity Turn

[RizzoTheRat]

An advantage of messing about with kOS is I can do more or less repeatable actions and log data while I'm doing it.  So I thought I've have a look at my gravity turns.  I usually end up cranking the ship over to something that looks right, and switching SAS to stability assist if I think it looks like Prograde is sending it a bit too low.

So I set up my SSTO launcher with a payload slightly within its load capacity, went straight up to 50m/s, held a specific deflection until 100m/s and then held prograde the rest of the way to orbit, but limiting MaxQ to about 0.25 atmospheres.  I then ran it a few times for different deflection angles, and tried limiting the engines a bit to look at the impact of TWR.  So I now have graphs for 3 different initial TWRs, going up pretty much the maximum angle where it could still make orbit.

The biggest surprise to me is that they're pretty linear, I was expecting gravity losses to decrease, but drag losses to increase (I did consider calculating these and logging them but was too keen to get started) as I increased the angle until it reached the point where larger angle become less efficient.  I guess you need a much higher TWR to find that point.

So for this particular vessel I can determine an optimum starting angle based on the TWR, but I need to do some more experiments to see what impact drag has.  My drag calculation to work out (Cd x A) seems a little bit flaky at the moment though so I need to figure out whats going on there first.

 

I know there's a few on here who like to do their calculations before launch, have you got any insights on working out optimum gravity turn angles?

Edited March 1, 2018 by RizzoTheRat

[OhioBob]

What you are doing sounds like what the mod Gravity Turn does, which is a launch autopilot.  It has the ability to learn from previous launches.  Gravity Turn sets when and by how much the initial turn is.  It makes a more aggressive turn each successive launch as long as the dV to orbit is decreasing.  But as soon as the dV starts to go up (because of excessive drag), it reverses direction, followed by small back and forth adjustments until it finds the optimum.

[RizzoTheRat]

Not come across that mod.  Trouble I found is on all 3 of those lines an extra half a degree meant it didn't make it to orbit, so the point where drag exceeds gravity losses is somewhere around or beyond the point where it causes you to crash

I'm thinking I could do something like switch to direction hold or aim a degree above prograde if the time to Apoapsis drops below a certain threshold level to allow it to recover from overly aggressive turns.

Presumably the optimum angle is going to be a function of TWR, drag, and something involving the position of the CoM.