How to optimize a space plane - how to know where it is lacking?

[Pappus]

I am currently working on my first SSTO Spaceplane with Mk 2 parts. I already tried it with the basic jet engines but those didn't have enough power so I postponed my try until I unlocked the turbo engine. However my craft doesn't make orbit and while I technically know that more fuel for my rocket thruster attached to it would eventually lead for it to orbit it would also be wrong for me to design it that way, because it would go more towards rocket than to spaceplane.

So I try different stuff, more turbo engines, more wings, less wings, more intakes and switching it up, but the results stay the same and I have no clue as to where to begin to push the craft in the right direction and if I don't know that (since obviously how am I supposed to know what I am really doing I don't have a spaceplane degree after all) I can't finish the plane. I also don't want to see too many different finished planes as that defeats the purpose of building your own craft - I just want to know what the key problems are to solve them.

So here is an orbit try with the little craft and obviously I lose control because of overeager steering, but it doesn't really make a difference that 215 thrust engine doesn't get me into orbit from that position since I tried it numerous times - at least not with a 45° angle I run out of fuel too early. Also remember since this is career mode and I don't have everything unlocked more intakes is a problem because I manually need to close them all.

Personally I think I need to reach ~30k AP on the turbo engines alone to have a shot, but maybe I am wrong.

http://youtu.be/q6FNpw0cG-o

[KerikBalm]

#1) I don't like to pitch down so suddenly, I would suggest following prograde from 8km until your pitch is at 20 degrees

#2) once at 20 degrees, keep it there, maybe decrease to 15 degrees with rapiers... Pitching down as rapidly as you did just kills your vertical velocity, and doesn't get you much more horizontal speed.... its betterto get a bit higher up, where there is less drag.

[eddiew]

I think you're trying to fly a 0.90 ascent, which doesn't work in 1.0.2. Turbos will quit by 20km, we just have to deal with it now. The pitch down you do at that altitude is not gaining you any speed from the turbos, and instead is being penalised by the drag model and costing you some. Avoid adjusting your pitch below 35km, just stick at whatever you're aimed already.

With a turbo/rocket plane, I'd be inclined to go up steeper; you'd pretty much maxed out your speed already, so see if you can find the steepest angle that results in 970m/s on air at 20km. It should be reasonable to get a 35-45km AP out of them. (With rapiers I'd say go shallow, but turbos don't have the same top end speed and deliver a lot more low-altitude grunt, so they're quite good at giving you a massive shove into a ballistic trajectory.)

How heavy is your cargo? Without rapiers, you may be a bit optimistic if you're trying to get both a crew cabin and a significant cargo to orbit; your plane isn't very big and you probably need that bay space for fuel and oxidiser.

Also, the open cargo bay on back end of your plane likely isn't doing you any favours... reality would rip it to shreds, and nuStock aero does not approve of flat back ends either (to the point people are starting to put nosecones on the backs of rapiers to smooth them out!). Use a mk2-mk1 adapter for a smoother transition and less rear-end drag

Check out the sticky SSTO Showcase thread for many examples of MK2 spaceplanes that work, and the various design tips and traps that the community is coming up with ^^

[Pappus]

For testing purposes I put a 4.5 T fuel container into the cargo bay to mimic any form of additional weight. I also wonder how am I supposed to avoid adjusting my pitch below 35km when you say that the turbos will give out at 20km. Do you mean go steadily and then just start the rocket? The problem I have is that the swivel or whatever the correct name for the 215 thrust engine is, that it is a bit too weak I will lose out huge margins of speed, which I think is not the way to go so maybe I should double the rocketry power to not lose too much to gravity?

Kerik so you think the plane itself should be able to orbit as is? I feel the transition to the rocket part is definitely the weakest about it. It doesn't feel like having real power

[Jaeleth]

1. Use sandbox mode to try. When ready, get back to career mode and proceed.

2. Enough wings to take off ground below 100m/s

3. Enough thrust to reach mach 1 below or around 10000m with a constant rate of climb.

4. Enough fuel to take it to orbit

5. Take off, pitch up 25o to 30o, keep it steady. Speed should build up, slowly at first, faster when passing mach 1 and 10000m. do not climb to steeply or you'll starve the air intakes, do not climb to shallow or you won't get to orbit. Increasing speed as it climbs is essencial since at higher altitude if you have no speed, you'll starve the engines.

6. Best results is a mix of whiplash and rapier, whiplash is ver efficient until around 20000m

7. Make sure you program a custom key to turn off whiplash and turn on rapier on closed circuit mode whenever you wish, which should occur at around 20km - 24km, do not let that decision to the computer.

[georgTF]

Am i the only one who begins his TurboJet ascents at 6-8km alt, airspeed 250-300 with angle of attack usually between 5-20 degrees? The TurboJet thrust is pretty insane and gives a lot of push before rocket engines are ignited. That said, i allways refuel at LKO and the planes i build aren't too heavy and barely have wings. But it seems strange that most people go higher up and pitch the plane up at a certain point (which i never do).

In my mind, an LKO SSTO needs to be easy to control during acent, not overly burdened with fuel and engines, be stable during re-entry and easy to land. Mass and drag are the enemy as are control surfaces. They should be placed further back. Being able to takeoff the runway is a bonus but there's nothing really wrong with gliding off the runway then pulling up. I've made too many failures to count and i'd advise building the thing as simple as humanly possible. Better to realise you'll need a tiny bit more grunt to achieve a stable orbit, then to carry too much mass from the beginning and never realise it.

Engines that are great in my experience: TurboJet, Aerospike, T-30, radially attached Twitch pairs and 909s. Thuds are not so good, T-45 is a bit wimpy in athmosphere compared to T-30 and features gimbals which can do more harm than good. I don't use Nukes because my planes don't fly very far.

Wings are allways a problem. Best to use them to get the center of lift where it needs to be. Control surfaces should be close enough to the center of mass to exert control over it, yet at the same time they should never bee too far up front because they tend to cause the same problems fins do when placed too high up on rockets. For that reason placing the CoM as far back as possible is helpfull. Placing tanks that drain and change mass and other heavier bits (cockpits and things) that don't goes a long way towards making a plane a good one. If the CoM moves too much during flight, unexpected things can happen and you cannot monitor it during flight. A general rule is to place the CoL behind the CoM so that both markers touch in the SPH. An empty plane can fly and land well enough if the CoM moves further forward as tanks empty. If it moves back, the plane is in trouble.

It's pretty much a constant challenge. It's probably easier to herd cats than to create SSTOs... Best give it a single, specific purpose (rescuing kerbals from LKO is a good one) and use only the bare essentials for the initial build. Then as things go wrong or you discover you need more stuff on the plane, modify, check where your centers of frustration are, occasionally rebuild and keep going until ragequit or succes.

Resist the temptation to overbuild. Less truly is more. Until you're 100% sure it isn't.

[KerikBalm]

Kerik so you think the plane itself should be able to orbit as is? I feel the transition to the rocket part is definitely the weakest about it. It doesn't feel like having real power

Post a craft file, and I'll give it a whirl.

The transition to rockets is easier if your turbos give out as you are in a 20 degree climb at 25,000 meters, going >1,000 m/s

[georgTF]

The transition to rockets is easier if your turbos give out as you are in a 20 degree climb at 25,000 meters, going >1,000 m/s

I engage rockets when Turbos hit their peak thrust. Thinking this gives them a bit more air as well as lightening the load by burning oxidiser. Usually they die out when the craft is moving at about 1300-1400 m/s and the rockets push for a 72km Ap. I then coast and circularise. Am i doing it wrong?

[KerikBalm]

When turbos are at their peak thrust, let them be....

They shouldn't be starved for air (airhogging is dead, you can spam intakes to provide the intake resource, but they have an altitude curve that means they won't produce thrust over a certain altitude) - so more speed at a given altitude doesn't really help.

What it will do is accelerate you faster, putting them out of their power curve faster. Thus more of the acceleration comes from the oxidizer consuming, ~320 Isp rockets, instead of the 8,000 Isp turbos.

It increases climb rate, and decreases the thrust they provide (because when they go faster, after their peak, their thrust declines).

If you are really igniting rockets at peak thrust, you're probably severely hurting your efficiency.

I wouldn't ignite them until your velocity starts to decrease.

As long as the turbos are still acclerating you... increasing velocity with an increasing apoapsis... do not fire your rockets.

Also, use the LV-t45, not the LV-T30 (which I assume you are, as it is 215 thrust), the LV-T45 gives you better control, and more Isp.

0.25 ton dry mass difference is miniscule compared to the dry mass of the rest of the SSTO.

[georgTF]

Thanks for the explanation. I guess pushing faster after peak thrust doesn't help at all. In fact it seems to hurt. It does seem to accelerate the craft faster overall, and it may be worth doing it anyway because it would spend less time fighting heavy drag. It is not the best way to get 100% out of the turbojets, but even at <100% they sitll push very hard and together with a rocket engine that is a lot of thrust used at a time where gravity and drag are doing their best to spoil the fun.

I wasn't trying to airhog btw, i mistakenly believed that pushing the engine through air as fast as possible is the best way to use it.

As for the T-30. It has better athmospheric Isp and i often engage below 20km so that's why i chose it over the T-45 which has better Isp in space, but worse in athmosphere. It's also not as punchy and i don't like the gimbal at all. Without aerospikes or rapiers, the T-30 is probably the best choice for LKO bound SSTOs. Since the worse Isp in space won't come into play as much.

[cybersol]

In a similar design from my career save, I ended up using a Poodle instead of the Reliant.

It could do 2 LKO rescues and launch a satellite that could fulfill 2-3 sat contracts all in one launch with full recovery.

Edit: Turn on your rockets when you stop gaining speed, usually around 20km with turbojets. Turn off the jets when they flame out around 25km.

Edited May 30, 2015 by cybersol

[KerikBalm]

Thanks for the explanation. I guess pushing faster after peak thrust doesn't help at all. In fact it seems to hurt. It does seem to accelerate the craft faster overall, and it may be worth doing it anyway because it would spend less time fighting heavy drag.

Its not about time fighting the drag, its about the integrated force of the drag. Going faster just means more drag per unit time.

And again... its better to fight that drag with 8000 Isp, than 320 Isp.

It is not the best way to get 100% out of the turbojets, but even at <100% they sitll push very hard and together with a rocket engine that is a lot of thrust used at a time where gravity and drag are doing their best to spoil the fun.

Waiting to fire your rockets lets the jets push harder, for longer. Don't worry so mych about gravity and drag, you've got 8,000 Isp!

As for the T-30. It has better athmospheric Isp and i often engage below 20km so that's why i chose it over the T-45 which has better Isp in space, but worse in athmosphere. It's also not as punchy and i don't like the gimbal at all. Without aerospikes or rapiers, the T-30 is probably the best choice for LKO bound SSTOs. Since the worse Isp in space won't come into play as much.

It very much does come into play. The Isp is not binary, but varies continuously between the value at 1 atmosphere, and a vacuum.

At 20km, engines will be getting very close to their vacuum values.

In an SSTO, you want to pick engines for their vacuum performance, because the air is so thin when you use them, that you get very close to the vacuum Isp.

This is why pure LF fueled SSTOs work... the LV-Ns are getting close to 60kn of thrust.

Of course, LV-Ns aren't great, because of low thrust, just like LV-909s.

Aerospikes and LV-T45s provide a good amount of thrust per attachment node, and have good vacuum performance (the aerospike does better than the 45, of course).

Rapiers? meh, their Isp is not good at all, but they save an attachment node that needs to be dedicated to a rocket (and allows for a lower drag design)

[georgTF]

I'm almost at the point of continuing to argue because i'm having difficulties accepting someone else is right. But humor me one more time on the subject of T-30 vs. T-45.

T-30 has better thrust and less mass. Gimballing is unimportant in this one case (at least for me). Worse Isp is what's confusing me. Does it mean it will suck up fuel faster as it provides it's thrust or that it will produce less thrust/acceleration overall due to using more fuel compared to it's cousin, the T-45? And would it not technically be an advantage to burn up fuel/mass more quickly and get more thrust during ascent to LKO only. If the T-45 can get the job done carrying less fuel then the anwser is no i guess. But if they're close to equal would the T-30 not provide a stronger push into orbit? I mean if it gets a higher Ap or raises it faster, then whatever it loses during circularisation is what the T-45 loses during ascent, so they come close. Esp since they will be almost equal in Isp during a portion of the ascent. Or am i off?

[edit]

Regarding when to best engage rockets: peak thrust is far less efficient than peak airspeed. KerikBalm has suggested correctly that loses to drag are much higher due to higher top airspeed in athmosphere with jets + rockets burning. Shedding mass wasn't enough to offset this higher drag. When activating at peak TurboJet thrust, the jets cut out one Km lower (23.8km vs 24.8) and the difference in airspeed at this point was 1416 vs. 1356 m/s. Though slightly faster, the less efficient approach used up more LFO overall. Both craft were piloted to a 72-80 orbit using an identical plane. Ascent profile was 7 degrees AoA at 7km with roughly 300 m/s airspeed (probably not the most efficient choice but that's what i did in the first run so i replicated it for the second attempt).

Even if you factor in pilotin errors/inconsistencies the end result were two craft in near idential LKO orbits except one still had 47.67 oxidiser left in it, while the other barely made it with only 1.71 remaining.

Thank you for your posts KerikBalm and setting me straight regarding this issue. I was succeding with the "engage at peak thrust" approach and thought of no reason to change it.

Edited May 30, 2015 by georgTF

[Bluntie]

As someone currently trying to figure out SSTO's, this thread is highly appreciated.

[Pappus]

How do you guys measure peak airspeed in stock? With the first comment Kerik made I assumed he meant to engage rockets when my AP isn't increasing anymore, which pretty much coincides with their burn out for me or do you click your intakes and start the rockets when their airspeed intake starts decreasing?

[cybersol]

Just use orbital speed at the top of the navball. As the jets lose thrust the speed there will stop accelerating rapidly and then quickly slow down. Right as it stops increasing is when I punch the rockets.

[Pappus]

In a similar design from my career save, I ended up using a Poodle instead of the Reliant.

It could do 2 LKO rescues and launch a satellite that could fulfill 2-3 sat contracts all in one launch with full recovery.

Edit: Turn on your rockets when you stop gaining speed, usually around 20km with turbojets. Turn off the jets when they flame out around 25km.

I simply dont understand why your craft makes LKO while mine looks pretty similar through the interations and doesn't make it. I tried 215 thrust, 200 and dual terrier. Nothing breaks orbit. More rocket fuel & wings was tested too.

Edited May 30, 2015 by Pappus

[GoSlash27]

Pappus,

It looks to me like you're about 10-15 tonnes short on fuel & O2, and you're underpowered on rocket thrust.

Your turbojets are easily strong enough to push a 30-35 tonne spaceplane up to 20km at Mach 2. Your rockets can take over from there, but if you're under .5 t/w ratio, you will lose too much to drag and may not be able to escape.

Finally, don't expect to exceed 14% of your total vehicle mass as payload if you're running a turbojet hybrid. You need to be pretty efficient to hit that mark.

Here's an example of a working 2 jet SSTO:

The wings are half- filled with fuel. All other tanks are full.

It makes 72x72 LKO with over 350 m/sec DV and still some reserve fuel for powered landing.

Even at this size, you can see how little payload fraction a turbojet hybrid is capable of. You can probably get 2 tonnes or so out of it if you omit the passenger compartment and docking port. You can push that up to 4 if you go with remote guidance instead of a flight crew.

Good luck!

-Slashy

Edited May 30, 2015 by GoSlash27

[Pappus]

Pappus,

It looks to me like you're about 10-15 tonnes short on fuel & O2, and you're underpowered on rocket thrust.

Your turbojets are easily strong enough to push a 30-35 tonne spaceplane up to 20km at Mach 2. Your rockets can take over from there, but if you're under .5 t/w ratio, you will lose too much to drag and may not be able to escape.

Finally, don't expect to exceed 14% of your total vehicle mass as payload if you're running a turbojet hybrid. You need to be pretty efficient to hit that mark.

Here's an example of a working 2 jet SSTO:

http://i52.photobucket.com/albums/g13/GoSlash27/TwinJet_zpsbopzxjt0.jpg

The wings are half- filled with fuel. All other tanks are full.

It makes 72x72 LKO with over 350 m/sec DV and still some reserve fuel for powered landing.

Even at this size, you can see how little payload fraction a turbojet hybrid is capable of. You can probably get 2 tonnes or so out of it if you omit the passenger compartment and docking port. You can push that up to 4 if you go with remote guidance instead of a flight crew.

Good luck!

-Slashy

The problem is if I add that fuel, then the plane will not make mach 2 anymore. With 1000 units more I capped at not even mach 2 but more like 700m/s. For example I tried incorporating your advice and double my rocket thrust and added another fuel tank mk2 long and it is definitely worse than 1x 215 thrust and 1x less fuel.

[cybersol]

SSTO are definitely not easy, and you can add too much fuel and not be able to make it to space when less would do. At this point its time to either play around with the design until it works or start to apply some calculations. For the latter you want 1700-1900 m/s dV at a TWR of 0.7-0.9. Using Methian's engine calculator you want a Poodle, or 2x aerospikes, or 5 terriers for around your current weight.

[Pappus]

Also a quick question: How to fix that the turbojets eat fuel of the rockets, but the rockets can't do the same with the fuel attached to the turbos.

- - - Updated - - -

SSTO are definitely not easy, and you can add too much fuel and not be able to make it to space when less would do. At this point its time to either play around with the design until it works or start to apply some calculations. For the latter you want 1700-1900 m/s dV at a TWR of 0.7-0.9. Using Methian's engine calculator you want a Poodle, or 2x aerospikes, or 5 terriers for around your current weight.

Just a question on those calculations: If I want 5 terries, why not 2x LV-T30? Those could deliver more thrust for the same weight or the same thrust if I don't full throttle them. 5 terries weight 2.5 afterall, b ut mainly when I put in the data you gave me I am getting mostly reliant and skipper

[GoSlash27]

Also a quick question: How to fix that the turbojets eat fuel of the rockets, but the rockets can't do the same with the fuel attached to the turbos.

- - - Updated - - -

Just a question on those calculations: If I want 5 terries, why not 2x LV-T30? Those could deliver more thrust for the same weight or the same thrust if I don't full throttle them. 5 terries weight 2.5 afterall, b ut mainly when I put in the data you gave me I am getting mostly reliant and skipper

Pappus,

Q1: the turbojets will take fuel from anywhere on the craft, while the Rockets will only drain fuel that is connected to them. You can use plumbing to feed fuel from unconnected tanks.

Q2: you could use less than optimal engines, but their lower Isp means they will require more fuel to make orbit.

More fuel means more structure and wings, which compounds the problem. I really recommend going with the poodle for this job.

Best,

-Slashy

[cybersol]

The skipper is definitely not right here. To use the calculator with a spaceplane is a little tricky, as the "payload" is the weight of everything minus the rocket engines and all the LFO supplies and tanks to get from the end of airbreathing to orbit. The calculator will then give number and type of engines and the total weight of the plane you will need to build to have enough LFO and thrust.

GoSlashy has good advice, as I also found a Poodle to be perfect for a similar plane.

[GoSlash27]

The problem is if I add that fuel, then the plane will not make mach 2 anymore. With 1000 units more I capped at not even mach 2 but more like 700m/s. For example I tried incorporating your advice and double my rocket thrust and added another fuel tank mk2 long and it is definitely worse than 1x 215 thrust and 1x less fuel.

Pappus,

Sorry I missed this earlier.

Mach 2 is around 650 m/sec, so you're still over Mach 2 there.

Turbojet spaceplanes have a tighter margin than RAPIERs, so the balance point where your spaceplane will work is also tighter. This means that a TJ design is harder to pull off than a RAPIER design.

What I've found for TJ designs is that they don't respond well to overloading. A RAPIER can be loaded to where it needs to dive to break Mach 1 and still get plenty of speed and altitude for the closed cycle portion of the ascent.

Turbojets are different. They will lose top end speed and altitude long before they have any issues around Mach 1. This is crippling for turbojets because they run out of steam at a lower speed and altitude and it takes a lot of fuel to muscle out of the hole on rockets. 100 m/sec or 1 km altitude is a lot to make up in the region where turbojets crap out. But OTOH you *have* to have enough fuel on board and engine thrust to get you out of the hole and into orbit. If you don't have that, it doesn't matter how high and fast you're going when you kick in the rockets. You ain't making orbit, so there's no such thing as "better" or "worse".

So try this:

Start over on your spaceplane. Keep it around 30 tonnes total mass. 300 units of fuel for the jets, the rest LF&O for the Poodle. And here's the important part: Make the whole darn thing fuel tanks.

Get that into orbit and see how much LF&O you had left over.

The leftover fuel and oxidizer (minus the reserve for maneuvers in orbit) is how much of your design can be dedicated to payload instead of fuel.

This is where I always start when I'm designing an SSTO spaceplane, whatever the type.

Good luck!

-Slashy

[KerikBalm]

T-30 has better thrust and less mass. Gimballing is unimportant in this one case (at least for me). Worse Isp is what's confusing me. Does it mean it will suck up fuel faster as it provides it's thrust or that it will produce less thrust/acceleration overall due to using more fuel compared to it's cousin, the T-45?

It means that it uses more fuel per unit thrust.

An engine producing 200 kN at 400 Isp will suck up fuel at half the rate(fuel/time) of an engine producing 200kn at 200 Isp.

An engine producing 100 kN at 400 Isp will suck up fuel at half the rate(fuel/time) of an engine producing 200kn at 400 Isp.

And would it not technically be an advantage to burn up fuel/mass more quickly and get more thrust during ascent to LKO only.

It can be... if you don't have enough thrust to get to orbit, it doesn't matter how efficient your engine is... you won't get to orbit.

Like a lander with a peak TWR of <1.0... I don't care how efficient its engines are... its not going to lift off an achieve orbit.

Likewise, when you need to raise your apopasis by ~20km, and your horizontal speed by ~1,000 m/s.. a tiny LV-1 is not going to cut it, evne if it is lighter and has a higher Isp.

However... that is a pretty extreme difference.

The LV-T30 vs LV-T 45 is a 0.25 ton difference, and a 15 kN difference... There are marginal cases where the LV-T30 is better than the LV-T45... but *generally speaking*, the 45 will do better.

If your SSTO is getting a bit too big for 1x LV-T45 to push it into orbit, but you don't want to add another attachment node (and associated drag and weight), switching to 1x LV-T30 may be a good idea.

If the T-45 can get the job done carrying less fuel then the anwser is no i guess.

Generally, yes.

Assuming payload + fuel mass makes engine mass negligible... a LV-T45 will achieve a dV of 320 m/s for the same amount of fuel that a LV-T30 uses to achieve a dV of 300 m/s

But if they're close to equal would the T-30 not provide a stronger push into orbit? I mean if it gets a higher Ap or raises it faster, then whatever it loses during circularisation is what the T-45 loses during ascent, so they come close.

If the LV-T45 is really struggling to get a payload to orbit, it will start to lose a lot due to excessive gravity and aerodynamic drag, and a LV-T30 would be better.

This has not been the case for most of my designs, and I go even 1 step further, and switch a LV-T45 for an even lower thrust Aerospike (with an eve higher Isp).

Going all the way to Nukes though... is a bad idea... the same with LV-909s I think.... then the thrust is really too weak.

As for the "stronger push to orbit"... that is valuable, and in some of my designs I use rapiers in closed cycle (which aren't that great) to intially raise my apoapsis higher, and then circularize with a more efficient engine.

For smaller designs... I just go with pure rapiers.... adding another node when you only have 2-3 engine attachment nodes will significantly increase drag...

Esp since they will be almost equal in Isp during a portion of the ascent. Or am i off?.

Assuming you're lighting rockets at ~25 km, the Isps will have a difference of nearly 20s... they will be very close to their vacuum Isp at that point.