Optimal ascent speed: terminal velocity?

[Rodyle]

Hi there guys,

I'm trying to create a generalised kOS takeoff script. I'm still fiddling around with ascent profiles, but I'm a bit at a loss as how to handle the velocity. I remember that back in the day, there was some discussion about optimal velocities for kerbal ascent, but I don't remember whether or not there eventually came a consensus on this. Some argued that you should try to stick as closely as possible to the terminal velocity, since it should minimize losses due to gravity and atmospheric drag.

 

Since it'd be fairly trivial to write a script which dynamically keeps a rocket at terminal velocity, I was wondering whether this is still thought to be the case.

Cheers

[John FX]

As far as I am aware, when the atmosphere changed it became very hard to reach terminal velocity once you were above about 10km.

did a google search and found this answer which seems sensible

"

Can terminal velocity can still be estimated for a common spaceship? It can be found out experimentally. The terminal velocity of an object is the speed it reaches during freefall. So when you drop the rocket from orbit top-first you can observe the terminal velocity at different speeds. Launch your rocket, but don't decouple the first stage. Let it continue on its sub-orbital trajectory. After reaching the apoapsis, turn the rocket so the top always faces the prograde direction and observe the falling speed at different heights. When it impacts the ground, revert the flight to vehicle assembly. During the fall it is important that you mimic the orientation during ascent. A rocket which flies/falls sideways has a higher aerodynamic drag and thus a lower terminal velocity than one which flies straight. Real-life engineers in the early days of space exploration used wind tunnel tests to optimize the atmospheric drag of their rockets. Today, computer simulations are used.

"

In my experience it does not make a great difference unless your rocket is very over or under powered, a TWR of 2.0 for the first few KM then 100% thrust would probably be close enough to perfect that the difference would not be worth the effort.

[Dunbaratu]

In addition to the fact that terminal velocity is now a lot harder to find, I was told it's also the case that it doesn't matter as much anymore even if you could find it.  It's no longer the optimal speed like it used to be.

[Snark]

My understanding is that terminal velocity is still the optimum speed.  The thing that changed with the new aerodynamics is,

• terminal velocity totally depends on the design of your ship
• ...but in general, a well-designed ship will have a considerably higher terminal velocity than was observed in KSP.

In practice, I've found that for most "efficient" rocket designs, it's simply not an issue because the ship never reaches terminal velocity.  To do so would require having a really high TWR, in which case you're wasting mass by lugging big heavy engines you don't need.  I launch rockets with a launchpad TWR of 1.5, and terminal velocity simply isn't an issue.

That said, if a given rocket did happen to have a very high TWR, and you're working out "what's the optimal profile for this rocket", then it's to keep it right at terminal velocity, at least for the early vertical part of ascent.

[Rodyle]

Thanks for the replies. If I read you guys correctly, it's still the optimal speed, but it's much harder to reach, so you might as well go with as fast as possible. In that case, I'll probably still write the script. it shouldn't be more than 50 lines and I'm interested how much it'll actually save in terms of delta-v.

[Snark]

2 hours ago, Rodyle said:

you might as well go with as fast as possible

Well, if someone has made a very high TWR ship, "fast as possible" won't be the right decision.  In practice, it's "go at terminal velocity" ... but you'll find that in most cases, your script will peg the throttle at 100% and just stay there.

 

[cantab]

To reach terminal velocity in a vertical ascent requires maintaining a TWR greater than 2, probably significantly greater. Which is indeed generally overkill.

As far as ascent profiles go, my own generic approach: Vertical until 75 m/s, pitchover to chosen angle, hold prograde. The starting pitchover sets the whole rest of the ascent which is then at maximum efficiency and minimum drag. There's a wrinkle though with when you switch from surface to orbit reference, especially in inclined orbits.

[wumpus]

28 minutes ago, Snark said:

Well, if someone has made a very high TWR ship, "fast as possible" won't be the right decision.  In practice, it's "go at terminal velocity" ... but you'll find that in most cases, your script will peg the throttle at 100% and just stay there.

Unless you are on Eve, I'm not sure you can manage to exceed terminal velocity without physics hacks (at which point you might as well do it inefficiently).  Programming a mechjeb/kOS system for "terminal velocity" will at least get you Eve for free, while being right for everything else.

[Snark]

Just now, wumpus said:

Unless you are on Eve, I'm not sure you can manage to exceed terminal velocity without physics hacks (at which point you might as well do it inefficiently).  Programming a mechjeb/kOS system for "terminal velocity" will at least get you Eve for free, while being right for everything else.

Yeah, terminal velocity's a real concern on Eve.  Those first 20 km of vertical climb are a real bear, and a script to follow terminal velocity there would be a big win.

In practical terms, it never comes up anywhere else, mainly because the next thickest atmosphere to launch from is Kerbin, and at the kinds of TWR that an efficiently-designed rocket has, you never catch up to your terminal velocity.

But you could easily hit it on Kerbin if you had a very high-TWR rocket (well above 2), or if you had an extremely non-aerodynamic one.

[cantab]

I'm surprised you can get enough thrust to hit terminal velocity on Eve, considering the double whammy of high gravity and lowered engine thrust from the ambient pressure.

[Rodyle]

Thanks again for the replies. One last question because I'm starting to doubt myself: should the total velocity of your rocket be as close as possible to terminal velocity, or just the vertical part of your velocity?