SSTO what to do to concerve fuel?

[endl]

so im trying to get my SSTO space plane to reach my station 150km in orbit but im having problems conversing fuel across all my design attempts.

most of my planes only have enough air intake until 21km regardless of intake spam. ive tried different engine combinations for 21-100km stage and nothing seems to work. i know turbo jets are best air breathers and 909 are best for orbital translations unless were talking nukes for longer distances. but whats a good "work horse" for the gap in between. right now im using rapiers because they save on weight but they eat fuel like crazy so that leads to my plane getting heavier since i just cant seem to strike a balance i feel like ive hit a brick wall.

Edited September 20, 2014 by endl

[Carsogen]

Hi there, this depends alot on the mass and configuration of your craft. For example nukes are really heavy (compared to other engines) but if your ssto is really heavy anyway this might not affect you too much. On the other hand if your craft is heavy then nukes may be too weak, or the layout of your ship may limit your engine configuration. I generally find that rapiers are a great option unless you're going on a long trip, because the weight saving you get by combining your engines (engines are heavy) can outweigh the extra fuel they need.

You're right though in that there's a balance, half my ssto's never even get through design stage , you just have to keep on trying different things and learning along the way!

Hope this helped

[Pecan]

"There is no in between. You either space or you don't."

Now I've got that out of the way ... Stop trying to go 'up' and read all the tutorials about spaceplanes that tell you to go "FAST". It sounds like you're trying to get a plane to space using a normal rocket ascent-path, which is possible but not efficient.

If you want to take all those wings and things to space the best way to do it is to be flying almost horizontal - less than 20m/s or so vertical speed - until you get near orbital velocity (~2,000m/s). What that means is spending a lot of time go sideways faster and faster until gravity just can't keep you down. As a rule - get above thick, draggy, atmosphere by going up to 20km as fast as you can. Then adjust your pitch so that you get 1km/s at 20km, 1.1m/s at 21km, etc. That should mean 2km/s at 30km and you're basically at orbital speed.

ETA: Oh - and 909s are almost never the right answer to anything

[endl]

so its the ascent path by the sound of it then, im using mechjeb is there a way to automate a none rocket ascent?

[Wanderfound]

"There is no in between. You either space or you don't."

Now I've got that out of the way ... Stop trying to go 'up' and read all the tutorials about spaceplanes that tell you to go "FAST". It sounds like you're trying to get a plane to space using a normal rocket ascent-path, which is possible but not efficient.

If you want to take all those wings and things to space the best way to do it is to be flying almost horizontal - less than 20m/s or so vertical speed - until you get near orbital velocity (~2,000m/s). What that means is spending a lot of time go sideways faster and faster until gravity just can't keep you down. As a rule - get above thick, draggy, atmosphere by going up to 20km as fast as you can. Then adjust your pitch so that you get 1km/s at 20km, 1.1m/s at 21km, etc. That should mean 2km/s at 30km and you're basically at orbital speed.

Listen to Pecan. If you try to fly a spaceplane as if it were a vertical rocket, all you get is an inefficient overweight rocket. Flown like a spaceplane, however...

How to fly a spaceplane to orbit

1) Get to 20,000m however you like. Around a 45 degree climb is probably most fuel efficient, but jet engines use so little fuel that it doesn't matter much. If the plane has enough power, I usually climb at 75 degrees or so just to get it done quickly.

2) When you get to 20,000m, level off and build some speed. You want to pile on as much horizontal velocity as possible while you make a slow ascent to 30,000m. Keep your angle of attack (the angle between where your nose is pointing and the direction in which the plane is actually moving, shown by the prograde marker when in surface mode) and climb rate low; by the time you hit 30,000m, they should both be around 10 or so. A low angle of attack reduces drag and helps your intakes work better. The low angle makes you climb slower, but that's okay; you need that time to get up to speed. As you go faster, the angle of attack required to maintain a given climb rate reduces, but as you go higher, the thinner air means that the angle of attack required to maintain a given climb rate increases. If you do it right, these two factors will roughly balance each other out and you should gain the necessary speed and altitude in a single smooth climb. However, a plane with some aerodynamic or piloting flaws may need to bounce up and down between 20,000 and 30,000m a couple of times while building speed before the final push.

3) Somewhere between 20,000m and 35,000m (exactly when depends on both plane and piloting), you'll start to run short of air. Don't switch to rockets immediately. If you've got multiple engines going, shut some down to concentrate the available oxygen into the ones you keep running. If you've already shut down as many as you can, throttle back a bit. You can dramatically increase your jet-only altitude by doing this, and once you get up to serious height the thin atmosphere means that you only need a tiny amount of thrust to accelerate.

4) Keep this going for as long as your plane and your patience can tolerate. A well-built and -flown plane should be able to get over Mach 4.5 and 30,000m in a single attempt on jets alone. Once you've wrung as much speed and altitude out of the jets as possible (you want at least Mach 4 and 30,000m), force the nose up to 45 degrees and light the rockets. If you have both jets and rockets, don't shut down the jets immediately; the thrust of the rockets will drive a ram-air effect that kicks the jets back into life for a while. Keep the rockets burning until your apoapsis exceeds 70,000m, then shut off and coast until it's time to circularise. Point prograde and close your intakes while coasting to minimise drag.

A good plane and pilot should be able to get the apoapsis to 70,000m with less than a minute of rocket power. Done properly, it requires very little fuel. But if you try to brute-force it from lower speeds and altitudes, the atmospheric drag is going to drain your oxidiser tanks before you get anywhere near orbit.

[Wanderfound]

so its the ascent path by the sound of it then, im using mechjeb is there a way to automate a none rocket ascent?

Mechjeb is nowhere near smart enough to fly a spaceplane on its own. If you want to fly, you need to learn to fly.

Have a look at the design contest in my sig; you should find some good stock-aero trainers there. O-Doc's stuff is particularly nice.

[mhoram]

ETA: Oh - and 909s are almost never the right answer to anything

That is true, but Tavert is: Mass-optimal engine type vs delta-V, payload, and min TWR

[Pecan]

Tavert's analysis is the only stuff that could beat your own and Kashua's, mhoram, that's why I keep it bookmarked ;-0

He made a particular comparison of LV-Ns, 48-7Ss and 909s and the 909 is only ever slightly better than the worst of those other two and only in very specific circumstances the best choice of the three. In other words - if you're thinking of 909s you almost certainly want either LV-Ns or 48-7Ss (science-tier considerations excluded).

Edit: put 'could' into the first sentence.

Edited September 17, 2014 by Pecan

[FleshJeb]

so its the ascent path by the sound of it then, im using mechjeb is there a way to automate a none rocket ascent?

(NOTE: I play stock aero)

Not that I've found. I have plans for an an adaptive KOS script.

SMART A.S.S. for pitch and heading control. 45-90 pitch for climbout; reduce pitch by 15 deg increments until speed run altitude, then hold 7-15 pitch.

Ascent Guidance can take over after that, you just need to fiddle with the ascent curve parameter.

FYI, Mechjeb's flameout protection is very conservative--if you throttle manually, you can get substantially higher and faster. Typically, I don't need to, but it's great for a marginal design.

One method for improving your speed run that I never see discussed:

Once you get to where the plane won't go any faster or higher, set pitch to 0-5 to build horizontal speed and let it fall. Use pitch to prevent more than -200 m/s vertical speed. Try not to lose more than 50 m/s horizontal speed. Let it fall to where you're almost back to full power on the engines, and resume your climb to speed-run height. Repeat 3 or 4 times, and each bounce will be higher and faster than the last (not just due to burning fuel, either). You should be able to squeeze another 200-300 m/s horizontal speed and a couple thousand meters of altitude out of this. It doesn't take a lot of extra liquid fuel, and it saves you a lot of rocket fuel.

Edited September 18, 2014 by FleshJeb

[Pecan]

Mechjeb is nowhere near smart enough to fly a spaceplane on its own. If you want to fly, you need to learn to fly...

While that is true, I know both Wanderfound and I use MJ's Smart A.S.S (SAS controller). For my flights - bearing in mind that a lot depends on the individual plane designs - I typically climb-out at 50-degrees, reducing that by 10-degree increments so that I'm at 20-degrees at 20km altitude. My horizontal speed will then be well below the 1km/s I recommend but building fast. From there up it's into speed-matching - the higher you go the faster you can go because there's less drag but the higher you go the faster you HAVE to go because there's less air to feed the engines. It's a balancing game and takes a bit of practice but once you've more or less "got it" it doesn't take much to adjust to any new design you come up with.

If you enjoy flying planes then - as Wanderfound would definitely recommend :-) - FAR or NEAR are worthwhile mods, because they make atmospheric conditions much more realistic. Personally, I tend to ignore the atmosphere except to get out of it as quickly as possible, which is why I prefer rockets, even if they do use more fuel.

@ FleshJeb - nice post :-)

Edited September 18, 2014 by Pecan

[SRV Ron]

Using the turbojet and the right design, you should be able to get your speed up to 2200+ and climbing by burnout. That will take a precise flight profile to do that as well as give you an apogee of 100+K. From there, a small rocket burn will place your ship into orbit.

This hybrid two stage will reach Mun and back. As a SSTO, it will easily go into orbit, but, only if you fly the high speed run carefully. Otherwise, you get excessive overheating in the air, or flame out going too slow to orbit efficiently.

Edited September 18, 2014 by SRV Ron

[Wanderfound]

While that is true, I know both Wanderfound and I use MJ's Smart A.S.S (SAS controller).

The lack of accounting for gimballing roll has made Smart A.S.S. useless for spaceplanes under power lately; you get an increasing oscillating roll that rapidly becomes lethal. It seems to have been fixed in the dev build, though.

Very handy for long-distance cruise control when you misjudge your reentry and have to fly halfway around Kerbin to get back to KSC.

[KandoKris]

I think what these guys have said is great but maybe to technical.

To keep it simple just take note of what altitude ur jet cuts out at and next time level out before u get there.

Then try to get your speed above 1350m/s, more is better but u can get to space from 1350.

Now ur up and flying marc 3 and just before jets cut out turn on rockets and aim at least above 40 deg up, more for heavy and slow designs.

keep going till your ap ( blue line) is above 70 000 then wait till your approx 30 sec to 1 min before the ap point (blue triangle on line) and burn again till orbit is reached.

I like to reach 75 to be on the safe side but thats me.

once in orbit making it bigger is easy to do.

the Basic ssto flight plan is to use jets as much as possible to get as fast as possible first as they use alot less juice.

Oh And yes YOU have to fly not mechjeb.

[Laie]

FYI, Mechjeb's flameout protection is very conservative--if you throttle manually, you can get substantially higher and faster. Typically, I don't need to, but it's great for a marginal design.

If you even need it, you're not aware of a very important quirk regarding the distribution of intake air:

http://forum.kerbalspaceprogram.com/threads/84217-Explaining-burnout-asymetry?p=1234436&viewfull=1#post1234436

(also check page 2 of that thread)

One method for improving your speed run that I never see discussed:

I'm not convinced that it's better in *any* way. If you get your initial approach right, you'll be pretty close to max; so close that falling back and taking another go will ultimately waste fuel (not to mention time).

As to the ascent, I learned not to look at the pitch all that much. What really matters is climb rate; just pitch to set your climb rate to whatever you want it to be. Things start to get interesting at about 18-22km; you don't want to climb too fast by the time you get there. This probably means that you need to start clobbering down on your climb rate at 12-15km and maybe even much earlier.

There's a brief phase when a lot happens all at once. Turbojets and Rapiers have their maximum thrust at an airspeed of 1000m/s, and half their rated thrust at zero and 2000m/s; when plotted, that function looks somewhat bell-curvish. This means that during most of the flight you have only 50-60% of their nominal thrust at your hands. But while you approach (and quickly pass) an airspeed of 1000m/s, your craft will be comparatively overpowered. Steer that power in the right direction, or at least keep it reined in.

If you're playing in the stock atmosphere, you're ruled by terminal velocity. Not THE terminal velocity, but that of your plane: at any given altitude, there's an airspeed where your engines' thrust and your vessel's drag would even out. If you don't climb, you will slowly approach (but never quite reach) that speed. You want your climb rate to be "just right" so that your vessel can accelerate at about the same rate as the terminal velocity increases. For most of my planes, this means about 100m/s climb when the engines begin starving, and ~40m/s by the time I get my first flameout.

[Wanderfound]

I'm not convinced that it's better in *any* way. If you get your initial approach right, you'll be pretty close to max; so close that falling back and taking another go will ultimately waste fuel (not to mention time.

There is one situation that leads me to use the "up and bounce" approach. With some of my planes, if I use a flat enough approach from launch to be hitting max speed and altitude at the same time, I end up losing my canards to Deadly Reentry on the way up. So with those ones, I deliberately use an overly-steep approach to start with, flatten out at 30,000m, "bounce" off the thicker stuff at 25,000m and then climb again once I'm up to speed.

At 30,000m/Mach1, trying to pitch up enough to hold altitude while accelerating often puts the plane into a stall, increasing drag; better to let it drop and build speed and resume climbing once I'm fast enough to do it with minimal angle of attack. Instead of trying to hold climb rate or pitch constant, I try to hold angle of attack as constant (and low) as possible. So long as your plane is right and you aren't actively diving, you'll start climbing again eventually once you get fast enough without having to touch the controls. Drag is the enemy, and angle of attack and control surface deflection are its allies.

I could probably avoid the "bounce" by gradually reducing pitch at exactly the right rate on the initial climb, but it doesn't seem worth the effort.

[Wanderfound]

Incidentally, there's an easy way to settle this question: http://forum.kerbalspaceprogram.com/threads/91544-Spaceplane-economy-challenge

[Kasuha]

How to fly a spaceplane to orbit

Impressive instructions but let me add my own experience. I am assuming stock physics and aerodynamics here so I'm not sure if it applies to OP.

A/ build your plane right. Rules for intake air distribution have already been linked so I don't have to do that. Use enough engines for the plane mass, the more engines the less time and fuel will you need. Optimal engine numbers for planes, minimizing problems with asymmetric flameout are 1 and 4. 1 engine is obvious, 4 engines need to be interleaved with intakes properly and placed in order outer-inner-outer-inner. For instance: (1)-(2)-(4)-(3). Why? Because when first stage flameout occurs, both your outer engines will flameout at once and inner engines will keep running. No asymmetry, and giving you enough time to lower the throttle. One RAM intake per engine is enough to reach orbital speeds, but two or three make it more comfortable. Don't make the plane too heavy, have 2 intakes per 10 tons at least (some will call that airhogging)

B/ flying

1/ reach 10 km as fast as possible. Use maximum pitch that allows you to keep your speed. 70 m/s is fine as long as you're getting out of the soup fast.

2/ above 10 km concentrate on building your speed. Switch to map view and watch your apoapsis. Switch navball to orbital mode. Raise your apoapsis to 20-25 km, not higher until you reach at least 1000 m/s. Regulate the apoapsis with pitch, don't be afraid to burn below prograde if your apoapsis is too high. Then raise it to 30 km and build your speed to about 1800 m/s. Then raise the apoapsis to 36 km and burn strictly prograde. If you have enough power, your orbit will grow out of atmosphere at this point. If you don't have enough power and reach the apoapsis, pitch up and keep the plane at the apoapsis (or rather the apoapsis below the plane) by changing pitch slightly up and down, continue burning until the apoapsis jumps to the other side of planet. After that, burn prograde.

3/ build orbital speed by burning prograde. As you lose air, your engines will start decreasing their thrust. When you get a flameout, decrease thrust to 2/3, then 1/3, and then go to just one tick above zero thrust. You can use physics time warp at this phase to wait till your plane gets out of atmosphere.

Edited September 18, 2014 by Kasuha

[O-Doc]

It might be a case of your ascent profile being wrong, it also might be a design issue. It's probably both. Given that everyone has made a good fist of discussing ascent profiles and we haven't seen your design yet, I'll do a quick checklist of design considerations.

1. TWR. On the runway you turbos jets should have a TWR of at least 0.75. So, for every 15 tons of aircraft you should have 1 turbojet.

2. Air intakes. You should have around 4 RAM intakes per turbojet to get into that efficiency zone. You can do it with less but you need a higher TWR. TWR and intakes have a direct relationship.

3. Orbital engine. 909s are fine and will serve you well if your craft is small. Beyond about 15 tons you will want to consider a NERVA powerplant to make the most out of your rocketfuel.

4. No useless gear. SSTOs are like rockets in that your really want to take up only what you need. Leave Bob's snack machine back at the KSC.

[Kasuha]

2. Air intakes. You should have around 4 RAM intakes per turbojet to get into that efficiency zone. You can do it with less but you need a higher TWR. TWR and intakes have a direct relationship.

Intakes per mass are IMO much better measure - at least in stock aero.

1 RAM per 10 tons won't get you to orbital speeds. 2 are enough, more are just for comfort.

[magnemoe]

Mechjeb is nowhere near smart enough to fly a spaceplane on its own. If you want to fly, you need to learn to fly.

Have a look at the design contest in my sig; you should find some good stock-aero trainers there. O-Doc's stuff is particularly nice.

You can use mechjeb smartass to keep direction and climb rate, you can also use the accent autopilot to maintain Ap after switching to rocket.

Mechjeb also have two other smart function, prevent jet flame out and manage intakes.

It also have an plane landing autopilot who is nice for crashing spaceplanes, it work for one plane I have but none of the others, better to land at the side of runway and drive up on it for recovery.

The normal landing autopilot can be used for the deorbit burn but as planes glides it will get confused so you have to put target around the mountains.

[O-Doc]

1 RAM per 10 tons won't get you to orbital speeds.

Is that a fact?

[Stratzenblitz75]

Is that a fact?

...1917 m/s is not orbital velocity since you do not have a periapsis above the lithosphere... ~2300 m/s is more like orbital velocity

So yes, it is a fact.

[Kasuha]

Is that a fact?

http://i.imgur.com/TlXpeG0.jpg

http://i.imgur.com/22pn59l.jpg

That depends on what you consider orbital speed. From my point of view it is suborbital speed/trajectory. But it can be called near-orbital, too.

I'm not saying you cannot fly an SSTO with one RAM per 10 tons, you just can't get it out of the atmosphere on jets. And since the topic is how to use least fuel to orbit, that's what you want - use jets for as much dv as possible, leaving only a little to circularize with rockets.

[O-Doc]

...1917 m/s is not orbital velocity since you do not have a periapsis above the lithosphere... ~2300 m/s is more like orbital velocity

So yes, it is a fact.

You can't get orbital velocity from any amount of intakes and jet engines if you want to take this argument to it's logical conclusion. So, you are engaging is sophestry.

That depends on what you consider orbital speed. From my point of view it is suborbital speed/trajectory. But it can be called near-orbital, too.

I'm not saying you cannot fly an SSTO with one RAM per 10 tons, you just can't get it out of the atmosphere on jets. And since the topic is how to use least fuel to orbit, that's what you want - use jets for as much dv as possible, leaving only a little to circularize with rockets.

Is that not the point? The craft I posted will probably take one ton of rocket fuel to get it to orbit leaving it at around 9T total mass, a third of which is fuel. That's quite efficient. And yes, with one RAM intake this craft will reach orbital speeds, just not on turbos alone. An obvious point I responded to Stratzenblitz75 which, you already know is implied in the statement you made.

Edited September 18, 2014 by O-Doc

[Stratzenblitz75]

You can't get orbital velocity from any amount of intakes and jet engines if you want to take this argument to it's logical conclusion. So, you are engaging is sophestry.

Yes you can reach orbital velocity with only jets, quite easily in fact:

Is that not the point? The craft I posted will probably take one ton of rocket fuel to get it to orbit leaving it at around 9T total mass, a third of which is fuel. That's quite efficient. And yes, with one RAM intake this craft will reach orbital speeds, just not on turbos alone. An obvious point I responded to Stratzenblitz75 which, you already know is implied in the statement you made.

Burning with turbojets will always be more efficient than burning with rockets. If you put more intakes on, you could have easily reached orbit with less fuel.