How to determine if you have hit terminal velocity // Find optimal throttle?

[Peppe]

Coming back to vanilla gameplay and missing an acceleration readout I had in modded play.

The g-force meter to the side of the navball seems to be the only tool/readout to represent acceleration. Is there a rule of thumb to read it and keep your craft on an optimal throttle for ascent?

Is there an indicator you are accelerating at terminal velocity and should reduce throttle?

Edited December 17, 2013 by Peppe

[Sirrobert]

G-force IS a readout of acceleration, it has nothing to do with terminal velocity.

If you want to know what your terminal velocity is, drop something and see how fast it goes. Than remember that number.

Or just go back to your mod

[Taki117]

There is a table on the wiki that lists the terminal velocities for various altitudes, try to keep your acceleration at, or slightly below that number.

Link for you!

[WafflesToo]

My personal rule of thumb is this...

100m/s off the launchpad

Keep the G-meter right around 1g

Aim to hit 240m/s at around 10km

Probably not mechjeb perfect, but it seems to give me consistent and reasonable results.

[Tank Buddy]

Terminal velocity only really applies to an object without a force of it's own, like thrust. If you are having problems with terminal velocity, you must have a wicked TWR.

[koshelenkovv]

Kerbal Engineer Redux mod shows this and much more useful things.

[WafflesToo]

Terminal velocity only really applies to an object without a force of it's own, like thrust. If you are having problems with terminal velocity, you must have a wicked TWR.

Following terminal velocity during ascent maximizes the balance between air and gravity drag.

[Nao]

Aside from following altitude / terminal velocity relation as Taki117 mentioned. You can just build the craft to fly at proper speed on 100% throttle. This is actually more efficient than following terminal velocity with varied thrust.

Just make sure that each stage burnout TWR no bigger than 2.1-2.5 (the lower the next stage initial TWR, the higher your burnout TWR). That's for early, vertical ascent up to ~10km because above that terminal velocity increases really fast and during gravity turn you will need more than 2.5-3 TWR to have acceleration keep up with terminal speed increase. After that the cost of additional engines to keep terminal velocity will make the design severely inefficient, so keep TWR of around 1,5-2.

If you want to play pure vanilla game, to save time on mass calculations, you can "launch" each stage on launchpad to see it's mass in map view.

[thereaverofdarkness]

There actually is a really simple way to estimate it. Any rocket with a TWR of 2.000 will accelerate gradually toward terminal velocity, getting ever closer with time. Now as time goes by and your rocket goes up, the air gets thinner and terminal velocity goes up. Your TWR also goes up as you deplete fuel, but different stages on your rocket may have very different starting and ending TWR. Also, the gravity lessens a bit as you ascend, further increasing your TWR. I find it all balances pretty well if your rocket has a starting TWR of ~2.0 on every stage. ~2.5 is the highest TWR you want to launch with, anything more and you'll be wasting fuel if you don't throttle down. So if your TWR is that high, just add more fuel if you can. Now if you have a TWR of less than 2 due to packing on lots of extra fuel, you may get more total delta-v out of your rocket. But don't go less than 2 unless you have a relatively small payload size.

The in-flight way to check and see how your speed relates to terminal velocity is by looking at the g-meter. If you are going exactly terminal velocity straight upward, then it will show -1G for a very brief moment immediately after you turn off your engines (I think). Now if your rocket settles just above 1G for most of the flight (after you get up to speed), you have a pretty good TWR for efficiency. If you have it showing 1.5G or higher even while you see streams of hot air shooting over your rocket, then your TWR is way too high. In fact, if you see any mach or re-entry effects during ascent, you are going too fast.

Quick reference for launch speeds from my memory:

1000m: 95-110m/s

10,000m: 275-350m/s (this is when you should tilt 45º east for optimal circular orbit insertion)

Stop thrusting at around 1100m/s (LESS for slow rockets, slightly more for fast rockets) to give yourself an apoapsis a little above the atmosphere.

2300m/s circular orbit velocity at 70km.

2250m/s at 100km.

[Aghanim]

How about with FAR installed?

[numerobis]

Just make sure that each stage burnout TWR no bigger than 2.1-2.5 (the lower the next stage initial TWR, the higher your burnout TWR). That's for early, vertical ascent up to ~10km because above that terminal velocity increases really fast and during gravity turn you will need more than 2.5-3 TWR to have acceleration keep up with terminal speed increase. After that the cost of additional engines to keep terminal velocity will make the design severely inefficient, so keep TWR of around 1,5-2.

Terminal velocity gets you the least deltaV usage when you're flying straight up, but as soon as you start to turn, the math gets harder. I'm fairly sure you should go slower afterwards, even on strict deltaV grounds.

[WafflesToo]

Terminal velocity gets you the least deltaV usage when you're flying straight up, but as soon as you start to turn, the math gets harder. I'm fairly sure you should go slower afterwards, even on strict deltaV grounds.

Correct. I usually only worry about the terminal velocity thing until I hit 10km or so. The air thins out so quickly after that that it's largely a non-issue. By the time I hit 20-30km I'm far more concerned with my time-to-apoapsis and climb-rate.

[Superfluous J]

Before I installed KER, my rule of thumb was 100 was fine everywhere, 150 at 5km, 200 at 7km, 250 at 10km, and after that the sky's (literally) the limit.

These days I just watch the Atmospheric Efficiency in KER but I notice that if I'm right at 99% or so (which is actually 260 at 10km) and I've had to throttle down a bit to keep it low, throttling up right at 10km pushes efficiency up over 100% for the next 2-5km. So maybe if you hit 260 at 10km, don't just punch it but wait until you're at 12-15km before doing so.

[Nao]

Terminal velocity gets you the least deltaV usage when you're flying straight up, but as soon as you start to turn, the math gets harder. I'm fairly sure you should go slower afterwards, even on strict deltaV grounds.

Correct. I usually only worry about the terminal velocity thing until I hit 10km or so. The air thins out so quickly after that that it's largely a non-issue. By the time I hit 20-30km I'm far more concerned with my time-to-apoapsis and climb-rate.

You are right. But that said, the least amount of dV spent does not mean the least amount of fuel spent.

Having TWR lower than 2.1 required for terminal velocity ascent makes us use more dV to get to orbit, but that is offset by higher dV from lower engine mass netting us an increase in performance. At around ~1.8 TWR we get the lowest fuel spent per payload unit. And if we decrease TWR further, we get lowest craft mass at launchpad at around 1.6-1.7 constant TWR.

So yes the whole thing "ascent at terminal velocity is the most efficient way" is a lie . Similarly to how people think asparagus staging is super important for ascent. But in KSP reality it only nets us like ~10% increase in performance.

Edited December 14, 2013 by Nao

[numerobis]

The term "efficiency" implies an objective function. The efficiency of vertical ascent at terminal velocity isn't a lie if your objective is to minimize delta-V. However, it doesn't follow that minimizing delta-V minimizes mass, part count, simplicity, structural integrity, and other such considerations.

Back to the optimal throttle: In order to reach terminal velocity, you'd need TWR of 1 + sin(pitch). The 1 fights drag at terminal velocity, the sin(pitch) fights gravity. If you keep TWR at or below that, you won't exceed terminal velocity even at full throttle. If your engines have 10% higher TWR, reduce throttle to 1 / 1.1 = 91% and you won't exceed terminal velocity.

At higher altitudes (above 10 or 20 km) you no longer care so much about drag, and what you want instead is to maintain a nice smooth gravity turn. In other words you want to thrust prograde just enough that your flight path will fall exactly the right amount so that at 70 km you'll be horizontal in a circular orbit. That's going to be tricky to calculate (I don't think anyone has quite figured it out yet, so the relevant challenges are still challenging), but again, ideally you don't need to change your throttle during flight, you build so that you can just fly at full throttle.

[WafflesToo]

At higher altitudes (above 10 or 20 km) you no longer care so much about drag, and what you want instead is to maintain a nice smooth gravity turn. In other words you want to thrust prograde just enough that your flight path will fall exactly the right amount so that at 70 km you'll be horizontal in a circular orbit. That's going to be tricky to calculate (I don't think anyone has quite figured it out yet, so the relevant challenges are still challenging), but again, ideally you don't need to change your throttle during flight, you build so that you can just fly at full throttle.

That's pretty much been my experience. I find I start really having to pull the throttle back just over 20km, often as low as 25% to keep my apoapsis from running too far ahead of me (keeping it around 30 seconds seems to give consistantly good results). I'm working on modifying my launch vehicle so the Skipper burns out at about that altitude and stage down to a smaller engine.

[John FX]

My personal rule of thumb is this...

100m/s off the launchpad

Keep the G-meter right around 1g

Aim to hit 240m/s at around 10km

Probably not mechjeb perfect, but it seems to give me consistent and reasonable results.

I came to say something similar. I use 100% throttle until 100m/s then increase speed at the rate so that by 7km I am doing 200m/s. 240m/s by 10km seems on the money as well..

It depends on the TWR of your craft. If it takes a long while to get to 100m/s then you will maybe need to be on 95% but if you get there very fast then you could only need 15%...

[MisterFister]

G-force IS a readout of acceleration, it has nothing to do with terminal velocity.

If you want to know what your terminal velocity is, drop something and see how fast it goes. Than remember that number.

Or just go back to your mod

Wow, such a flame comment.

How about with FAR installed?

FAR generally displays it for you with one of the in-flight readouts. I recommend spending some time on the launchpad actually reading all of the readouts word for word, and clicking on all the ? buttons as they appear. Same in the craft editor. All abbreviations and variable notations are explained and described in accessible terms. While the numbers are indeed different between FAR and stock, I can say from experience on both sides of that divide that in practice, keeping a launch TWR of ~1.8-2 will be fine. In fact, for the most part, with FAR it is quite difficult to attain tV during launch for extended periods, even at full throttle. (The forces needed for sustain upward tV would usually end up g-stressing your craft if you have DeadlyReentry installed.)

[el_coyoto]

Coming back to vanilla gameplay and missing an acceleration readout I had in modded play.

The g-force meter to the side of the navball seems to be the only tool/readout to represent acceleration. Is there a rule of thumb to read it and keep your craft on an optimal throttle for ascent?

Is there an indicator you are accelerating at terminal velocity and should reduce throttle?

You can use the g-meter to make sure that your craft doesn't have a TWR brutally over 2.0 at lift-off.

If it's the case, then your rocket probably has too much thrust at lift-off and "wastes" weight in engines.

But the g-meter won't tell you very much about your speed and its relation to terminal velocity (defined as "atmospheric efficiency" in KER).

Personally I use KER to hug the terminal velocity as much as I can during the first kilometres of the ascent, because as it has been mentioned earlier, terminal velocity raises exponentially (as air pressure decreases exponentially as well) and atmospheric efficiency stops being a problem after ~20 km.

When I play stock ("RP" career for example), which seems to your case, I use a "altitude vs terminal velocity" chart. It adds a nice 50's "charts and sliderules" feeling to the game...

Or, as Nao proposed, you can build your rocket first stages so that they are close to terminal velocity even at full throttle. It's a longer "design-tweak-test" process, but this how I used to build my flagship rockets (so they can reach orbit efficiently, even without MJ or KER).

Edited December 17, 2013 by el_coyoto

[Peppe]

Thanks all. Looks like cheat sheat / rule of thumb will have to work... let go of min/maxing performance and just fly it kerbal style.

[Superfluous J]

Thanks all. Looks like cheat sheat / rule of thumb will have to work... let go of min/maxing performance and just fly it kerbal style.

Yeah you can play KSP one of two ways: Min/Maxing with a mod like KER, or "Kerbal Style".

There is a 3rd way that involves a lot of time and coffee and pencils, but I do not consider that "playing."

[WafflesToo]

Yeah you can play KSP one of two ways: Min/Maxing with a mod like KER, or "Kerbal Style".

There is a 3rd way that involves a lot of time and coffee and pencils, but I do not consider that "playing."

Define "playing".

For every hour I spend in front of the terminal I probably have one spent with coffee and pencils. I get this huge rush seeing a plan I put together hours in advance actually work in practice at the moment of execution.

Of course there's also something to be said for Zisteaunian contraptions and shooting from the hip.

[Superfluous J]

Define "playing".

For every hour I spend in front of the terminal I probably have one spent with coffee and pencils. I get this huge rush seeing a plan I put together hours in advance actually work in practice at the moment of execution.

Of course there's also something to be said for Zisteaunian contraptions and shooting from the hip.

Hey I'm not saying it's wrong. I'm just saying I'm not going to do it.

[John FX]

Hey I'm not saying it's wrong. I'm just saying I'm not going to do it.

I take it you do not play dwarf fortress. I hear that has `fun` too...

[WafflesToo]

Hey I'm not saying it's wrong. I'm just saying I'm not going to do it.

I'm not saying it's right... though it does make me suspect I suffer from some kind of mental illness