Efficient lauch?

[Carrot]

I just think of a way to do an slightly efficient burn in my mind. Does it use less duel this way?

Do a gravity turn reaching around 30m/s slowly. It would be 45 degree around 1km. But then when the ap reaches 50 km. I cut off the engine till it is 45km (reaching the black area of the bar).

1) I don't know the terminal velocity of the rocket so this can lower my speed a bit to avoid extra fuel spending on friction and drag
2) I guess it can maximize the time of "vacuum,burn" for the maximum delta V ?

[Alshain]

No, if you are doing the most efficient burn, you won't be that far behind Ap at any time during your launch until after you are far enough into the high atmosphere it will not matter anymore.

Here is a video showing how to launch properly (good launch starts at 7:45)

 

Edited January 27, 2016 by Alshain

[Elway358]

1 hour ago, Alshain said:

No, if you are doing the most efficient burn, you won't be that far behind Ap at any time during your launch until after you are far enough into the high atmosphere it will not matter anymore.

Here is a video showing how to launch properly (good launch starts at 7:45)

 

Absolutely love that black and white navball! Can you get just that b/w ball from a mod or does it come with lots of other stuff?

[Carrot]

2 hours ago, Alshain said:

No, if you are doing the most efficient burn, you won't be that far behind Ap at any time during your launch until after you are far enough into the high atmosphere it will not matter anymore.

Here is a video showing how to launch properly (good launch starts at 7:45)

2 hours ago, Alshain said:

No, if you are doing the most efficient burn, you won't be that far behind Ap at any time during your launch until after you are far enough into the high atmosphere it will not matter anymore.

Here is a video showing how to launch properly (good launch starts at 7:45)

 

I tried this method but i couldn't even reach orbit for my launch. I can reach 100 km obit and return in stock game with my 1938+467 (atmo dV) / 2297+1894(vac dV) rocket

 

[Alshain]

47 minutes ago, Elway358 said:

Absolutely love that black and white navball! Can you get just that b/w ball from a mod or does it come with lots of other stuff?

Texture Replacer is the mod and you can find tons of options here.  Specifically this post is where I got mine.  Rename the file HUDNavBall.(extension) and place it in GameData\TextureReplacer\Default

[Alshain]

51 minutes ago, Carrot said:

I tried this method but i couldn't even reach orbit for my launch. I can reach 100 km obit and return in stock game with my 1938+467 (atmo dV) / 2297+1894(vac dV) rocket

It takes practice.  Keep trying, you will get it.

Edited January 27, 2016 by Alshain

[Nich]

That was a terrible launch TWR on that rocket was way too low and his cos losses were enormous as well as his gravity losses.  Basically the best launch possible you  turn to 10/20/30 degrees asap depending on TWR and follow the propragrade marker the entire time at full throttle and end in a 70x70 orbit.  Not easy to do.  Try to get your AP to 65 as late in the burn as you can at full power then burn full power horizontal (between blue and brown) until your AP is at 70 or 71km then circularize.

[NathanKell]

Efficient in terms of what?

Cost? You'll need to accept higher delta V to orbit and lower payload fraction.

Delta V? You'll need to accept a much lower payload fraction, and you might burn up.

Payload fraction? It may take a lot of delta V, and you'll need to use only the best equipment.

 

Generally delta V is the worst thing to optimize for, because if you get a higher payload fraction who cares if you spent 300 m/s more? (And same with cost.)

In terms of some happy medium, however, here's a few guidelines (they may conflict). Any time you're throttled down (or not burning at all, throttle 0, the worst) that's a sign you brought too much engine (payload fraction is suffering) or you're taking gravity losses you shouldn't (delta V cost is increasing). Any time you're not burning prograde, you're taking steering losses. Any time you're under terminal velocity (it can be very high, mind), again, you're taking gravity losses you could avoid.

[Majorjim!]

11 hours ago, Nich said:

That was a terrible launch TWR on that rocket was way too low and his cos losses were enormous as well as his gravity losses.  Basically the best launch possible you  turn to 10/20/30 degrees asap depending on TWR and follow the propragrade marker the entire time at full throttle and end in a 70x70 orbit.  Not easy to do.  Try to get your AP to 65 as late in the burn as you can at full power then burn full power horizontal (between blue and brown) until your AP is at 70 or 71km then circularize.

This! Gravity turn in this game is dependent on the nature of each rocket. TWR, Areo cross section ect. After thousands of hours you will be able to eyeball it depending on the TWR of each stage. Just practice, don't watch a video showing the 'correct way' as it will be incorrect for the rocket you have. Just try different launch profiles and see what is best. You cannot beat practice, practice, practice!

[Armisael]

12 hours ago, NathanKell said:

Efficient in terms of what?

Cost? You'll need to accept higher delta V to orbit and lower payload fraction.

Delta V? You'll need to accept a much lower payload fraction, and you might burn up.

Payload fraction? It may take a lot of delta V, and you'll need to use only the best equipment.

Only one of those is really amenable to change during an ascent. Your payload is what it is and needs to go wherever it needs to go - the mass fraction is irrelevant. Similarly, you've already paid for the rocket and designed some portion of it to return for recovery value.

[nobodyhasthis2]

14 hours ago, NathanKell said:

Efficient in terms of what?

Cost? You'll need to accept higher delta V to orbit and lower payload fraction.

Delta V? You'll need to accept a much lower payload fraction, and you might burn up.

Payload fraction? It may take a lot of delta V, and you'll need to use only the best equipment.

 

Generally delta V is the worst thing to optimize for, because if you get a higher payload fraction who cares if you spent 300 m/s more? (And same with cost.)

In terms of some happy medium, however, here's a few guidelines (they may conflict). Any time you're throttled down (or not burning at all, throttle 0, the worst) that's a sign you brought too much engine (payload fraction is suffering) or you're taking gravity losses you shouldn't (delta V cost is increasing). Any time you're not burning prograde, you're taking steering losses. Any time you're under terminal velocity (it can be very high, mind), again, you're taking gravity losses you could avoid.

Which is pretty much the answer in a nut shell. Variables on top of variables.  

20 hours ago, Carrot said:

Do a gravity turn reaching around 30m/s slowly. It would be 45 degree around 1km. But then when the ap reaches 50 km. I cut off the engine till it is 45km (reaching the black area of the bar).

However if we isolate the variable of rocket, control and everything else we can start to pick out specific characteristics. Here is a video showing variations of just turn by altitude with as many of the variables removed. From this foundation tailor specific missions, designs and launch profiles. To answer  @NathanKell's very good question as required. Efficient in terms of what?

It is really good advice.

Kerbal Space Program 101 - When Do You Start Your Gravity Turn?

 

  

[Alshain]

16 hours ago, Nich said:

That was a terrible launch TWR on that rocket was way too low and his cos losses were enormous as well as his gravity losses.  Basically the best launch possible you  turn to 10/20/30 degrees asap depending on TWR and follow the propragrade marker the entire time at full throttle and end in a 70x70 orbit.  Not easy to do.  Try to get your AP to 65 as late in the burn as you can at full power then burn full power horizontal (between blue and brown) until your AP is at 70 or 71km then circularize.

Well, if you are optimizing for cheap fuel instead of inexpensive engines, absolutely.  Sure, you could add boosters or a more expensive engine and raise the TWR, but I promise you my rocket would end up cheaper and accomplish the same goal.  Fuel cost is insignificant making it the same for cosign losses.  It doesn't matter if I lose more fuel if my lifter is still less expensive.  As long as fuel is so cheap, it will always be that way.

As @NathanKell said, it really depends on what you are optimizing for.  However, optimizing for Delta V is silly on a lifter, and in a career game you want to accomplish the goal using as little funds as possible, so I always assume cost.  Short of an SSTO, you aren't getting cheaper by increasing TWR higher than what my rocket had.

Edited January 27, 2016 by Alshain

[NathanKell]

@nobodyhasthis2 it's true if we isolate just to rocket control we can pick out stuff, but...that video is two years old, it's long before the aero overhaul last spring, and is very, very bad advice (and quite possibly rocket-destroyingly catastrophic due to aerodynamic forces) these days.

 

@Armisael you're not wrong that if the LV is designed then what you want to optimize for is delta V expended. However, at *that* point, minimizing delta V just to LKO is a fool's errand because, unless that's all you're going to, you might do better going for a higher orbit (i.e not thorttling down let alone coasting). You have to balance steering, gravity, and drag losses *and* you have to include subsequent burns in your calculation...

[mcirish3]

Isn't the real question, "Hey guys I have this rocket, that i have already built and is sitting on the launch pad ready to go.  Based on this particular ship's design could you give me an approximate flight profile that will minimize my DV expenditure LKO, and maybe some extra information that will guide me towards how changes in ship design should affect my chosen flight profile."
@Overengineer1  This question seems right up your alley.

I know carrot did not say all that but he is new to the game and doesn't even know what questions to ask. 

Edited January 27, 2016 by mcirish3

[nobodyhasthis2]

1 hour ago, NathanKell said:

@nobodyhasthis2 it's true if we isolate just to rocket control we can pick out stuff,

Totally correct again.

1 hour ago, NathanKell said:

but...that video is two years old, it's long before the aero overhaul last spring, and is very, very bad advice (and quite possibly rocket-destroyingly catastrophic due to aerodynamic forces) these days..

I was actually thinking about that when I posted that. The aerodynamics has changed. The numbers have changed. Which ties in with the variables changing. Which kind of illustrates that there is no perfect answer. Ironically I was thinking that not using new aerodynamics in an example also removes another variable. Aerodynamic performance. 

I didn't expect anybody to religiously follow the numbers but just to understand the concepts you explained by a visual demonstration. Besides the general principles worked for FAR users before last spring. They just had to factor in aerodynamic forces into the design and launch parameters like everyone now has to now. It also fair to say that following this verbatim is not going to help anyone using RO . However the principles of turn efficiency should still be relevant to the question. I just felt it help clear this up

16 hours ago, NathanKell said:

In terms of some happy medium, however, here's a few guidelines (they may conflict). Any time you're throttled down (or not burning at all, throttle 0, the worst) that's a sign you brought too much engine (payload fraction is suffering) or you're taking gravity losses you shouldn't (delta V cost is increasing). Any time you're not burning prograde, you're taking steering losses. 

If you think it bad to use this to show an isolation of factors such a steering loss. Or something that will show a fuel efficiency changes based on changing point of the turn whilst isolating others. I could remove the link. More that happy to do this as I really don't want to mislead someone. Or confuse them over the new aerodynamics. Really think it should be removed if it does more harm that good.

Should I do this?

Also it might be better if someone used the Gravity Turn mod to plug in the numbers to isolate the effects of gravity turn on fuel efficiency for education purposes.

 

Edited January 27, 2016 by nobodyhasthis2

[Skalou]

Really nice lauch Alshain!

It answers a bit what i'm wonder actually, does a perfect Gravity turn is the best solution? what about having some higher steering losses and drag balanced by the gravity losses by going out of prograde after the lower atmosphere in an objective of low price.

Does some real rocket goes out of prograde like these? (even if KSP is different, just to know)

Mechjeb has some nice tools to see the différents losses  and other usefull features such as real expended DV:

 

(here in action during some tests)

[Alshain]

@Skalou On non atmospheric bodies, all you have to do is clear the terrain and then go flat as possible, right on the horizon.  A best launch in the atmosphere would have you turn gradually without stopping till you are on the 10 degree mark.  In the video I turned just a little too fast and had to burn above prograde for a while to keep from reaching Ap, that is why I ended a little deeper in the atmosphere.  In the end, it didn't cost me that much more though and it's certainly better than burning to 70km at 30 degrees, as you can see if you watch the first half of the video.

[NathanKell]

@nobodyhasthis2 fair enough; might as well leave it then.

 

@Skalou in real life you go off-prograde all the time, indeed most modern LVs (and Saturn V when ascending to parking orbit) circularize after apogee. But in real life while the cosine loss for burning off prograde for 30s may be 50m/s, just as in KSP, you're talking about total steering losses of maybe 75m/s out of 9400m/s expended, vs KSP where that would be 75 out of 3200 or so.

[Nich]

I think red iron crown posted something from mech jeb or another tool from a gravity turn and in it 15% was lost to gravity 1% was drag and 2% was cos losses.  In reality loss counts for very little even in a straight up to 70m and then circularize.  Most engines max out there dv/ton when you put it under 1 of the largest tanks that fit above it.  2 and you don't get as much dv/ton but you do get more dv per engine.  Are you limited by pad weight? part count? cost? they all have different design stratagies

 

If you want cheap solid booters cant be beat.

[Skalou]

On 28/1/2016 at 2:07 AM, Alshain said:

@Skalou On non atmospheric bodies, all you have to do is clear the terrain and then go flat as possible, right on the horizon.  A best launch in the atmosphere would have you turn gradually without stopping till you are on the 10 degree mark.  In the video I turned just a little too fast and had to burn above prograde for a while to keep from reaching Ap, that is why I ended a little deeper in the atmosphere.  In the end, it didn't cost me that much more though and it's certainly better than burning to 70km at 30 degrees, as you can see if you watch the first half of the video.

That's what i was doing on the mun image, keeping the prograde on the horizon ( there with infinite full to check some other stuff for an other test), it was just to show you the mechjeb tools that are a bit hidden.

and for the atmospheric, i wonder if it's better to do a prefect gravity turn (follow the progade until we are in a circular orbite) or at one point going out of it to do the same kind of maneuvre as on airless bodies: not a perfect constant altitude burn, but a kind-of by increasing a bit the altitude too (it means the prograde will be a few degrees up). but your test seems to show it is. of course it depends of the rocket building, but it shows it is a possible way to go.

I've seen a rocket doing this but in the other way: the ariane 5, it  seems to lauch more vertical and then the trajectory make a jump out of the atmosphere (30km) to give the time to the low TWR stage to burn, but whitout the pitch and the flight path or AoA angle it's hard to understand well what it does, also It has a lot more and differents constraints than KSP.

Edited January 29, 2016 by Skalou

[Alshain]

4 hours ago, Skalou said:

That's what i was doing on the mun image, keeping the prograde on the horizon ( there with infinite full to check some other stuff for an other test), it was just to show you the mechjeb tools that are a bit hidden.

and for the atmospheric, i wonder if it's better to do a prefect gravity turn (follow the progade until we are in a circular orbite) or at one point going out of it to do the same kind of maneuvre as on airless bodies: not a perfect constant altitude burn, but a kind-of by increasing a bit the altitude too (it means the prograde will be a few degrees up). but your test seems to show it is. of course it depends of the rocket building, but it shows it is a possible way to go.

I've seen a rocket doing this but in the other way: the ariane 5, it  seems to lauch more vertical and then the trajectory make a jump out of the atmosphere (30km) to give the time to the low TWR stage to burn, but whitout the pitch and the flight path or AoA angle it's hard to understand well what it does, also It has a lot more and differents constraints than KSP.

It is absolutely possible to get a perfect launch, it just takes practice.  Most of my launches I get into a not-quite-orbit just like in the video, though I usually end up at about 60km before I'm done with the burn.  The key thing is to just get your Periapsis as high as possible before reaching your target orbit apoapsis.

Edited January 29, 2016 by Alshain

[Skalou]

On 29/1/2016 at 7:47 PM, Alshain said:

It is absolutely possible to get a perfect launch, it just takes practice.  Most of my launches I get into a not-quite-orbit just like in the video, though I usually end up at about 60km before I'm done with the burn.  The key thing is to just get your Periapsis as high as possible before reaching your target orbit apoapsis.

I think it's not,

For me the 1st step is to know what is a perfect launch ( for a given rocket even if it is a bad one) for a given goal, and i'm discovering some small tricks frequently. also this perfect could match different criteria of success  (efficiency, precision, reliabilty (margin to fail the whole mission),...).

and then do it, and there is always a small differences with the targeted trajectory and also between 2 launch.

But if you claim it is possible i would like to see it (even if it's not you).

Also i don't need a perfect lauch on my side, i take margin to success even with  small deviations.

[Alshain]

We may have a different defintion of a perfect launch.  You said to follow the prograde the whole way, that is what my launches are usually like.  You just have to find the right time to turn.  Do it too early and you will end up like the video where you have to keep high to avoid hitting Ap, too late and you will have trouble turning against aerodynamics and burn a lot of fuel in circulation.  Really late and you won't have any trouble aerodynamics but you will burn an absolute ton of fuel.

Given that it is a manual process, it's hard to do it every single time.

Edited February 3, 2016 by Alshain

[Skalou]

aha, perfect= can't do better for me.

and as you i don't look at a perfect lauch but a good enought (and there is so much things to do in KSP and so few time that i can't focus only on launch and try to improve also some other things as long as it gives me fun)

I discovered soon enought the balance between when to start the gravity turn ( with the effects you descibe), discovered too the thermal limit, after this that with a low TWR 2nd stage i can avoid a fail by giving a big enought AoA at a propper time (after the denser atmosphere) and even save some DV, but your test as you are a far more experience player than me (and play challenge with other people? ) confort me that it could also be one of the cheepest way to go to space with a rocket.

[Warzouz]

On 27/1/2016 at 8:10 AM, NathanKell said:

Generally delta V is the worst thing to optimize for, because if you get a higher payload fraction who cares if you spent 300 m/s more? (And same with cost.)

This so much true when designing a launcher.

Few days ago, I was testing a high TWR rocket (for fun). The rocket was powered by a vector and had near a TWR of 5 (so much overpowered ). It was so hard to steer that I needed to launch it already inclined with a very flat ascent (crossing 45° at around 4000m, If I'm not mistaken). The rocket speed increased so fast I had to stop engines mid course to cool down.

Finally I went to LKO for 2850m/s (VAC). But the rocket had no payload and was finally quite big (3 biggest 1.25m tanks, probe core, a nose cone, a vector and 4 fins).

Strangely the rocket was going to LKO without steering and without SAS. I was only managing the throttle mainly to handle heat.

In the end, you can go to LKO for a very low dV, but you don't pay for dV, you pay for fuel and rocket parts.