Spaceplanes: when to go supersonic?

[Laie]

I don't fancy going supersonic in the lower atmosphere. I tend to pull up as hard as the craft allows, that is, I find and hold a pitch where airspeed increases only very slowly, up to however high that takes me until I feel that either AoA is becoming too steep, or the climb rate too small, to keep going like this. Only then do I go into more-or-less level flight to break the sound barrier (emphasis on more-or-less, don't get hung up on level flight). Usually that happens around 8-12km, depending on the vessel.

However, it seems that I'm the exception. Most galleries and videos I see, the plane goes supersonic at much lower altitudes, often even as a first measure before starting a serious climb.

Now, who's doing it wrong? Is there even a right and a wrong way of going about it?

[foamyesque]

A fair number of my planes *can't* climb worth a damn unless they're supersonic. The big reason to go supercruise early is because, if you're using RAPIERs (and to a lesser extent, Whiplashes) you can use the positive feedback on the thrust curve to reach your top speeds much easier. After about the 10km mark I want to be doing around 1km/s, with approximately an additional 100m/s per 1 km of altitude up to the point where the jets can't make the thing go any faster.

Also, I often find it tricky to get the thrust and lift needed to go supersonic at the altitudes you are unless I'm already thoroughly supersonic; between the Mach wall, the reduced thrust, and needing to have a significant AoA just to maintain level flight, the engines just can't overcome the drag. Something carrying more engines or wing would have less trouble, I suppose.

[Geonovast]

The only consistently successful SSTO plane I've built hits Mach 1 about 1km above sea level, and about Mach 4 when I have to kill the air breathers.

[Laie]

51 minutes ago, foamyesque said:

Also, I often find it tricky to get the thrust and lift needed to go supersonic at the altitudes you are unless I'm already thoroughly supersonic; between the Mach wall, the reduced thrust, and needing to have a significant AoA just to maintain level flight, the engines just can't overcome the drag. Something carrying more engines or wing would have less trouble, I suppose.

Hmm, maybe it's because I tend to field a lot of wing.

Just btw, the jets' thrust follows air pressure up to 5km, then outpaces it: by 15km Whiplash and Rapier are supposed to have about twice as much thrust per drag. But of course, a high AoA can eat all that and then some.

[Rocket In My Pocket]

The way I learned to do it was like a 15 degree climb till your max engine output (like 10k for Whiplash) then a 5 degree climb at Mach 3-4 till you run out of air, then vacuum engines.

If your plane has the TWR, I see no reason to loiter about though, just get right up to altitude and make your speed run.

Most of my designs don't have the TWR for that though. I think this is one of those things that just comes down to personal preferences.

[foamyesque]

1 hour ago, Laie said:

Hmm, maybe it's because I tend to field a lot of wing.

Just btw, the jets' thrust follows air pressure up to 5km, then outpaces it: by 15km Whiplash and Rapier are supposed to have about twice as much thrust per drag. But of course, a high AoA can eat all that and then some.

 

High AoA certainly can, but the other factor here is that to really beat thrust-per-drag you need to *already be going fast*.

[GoSlash27]

Laie,
 I do it the same way you do. 5° AoA on the climbout (which translates to nose at prograde with the static incidence), and decrease pitch as necessary until I clear the wall. My SSTO spaceplanes are generally too underpowered to exceed Mach 1 at sea level, they generally jump into the supersonic regime around 6km altitude in level flight. I think others tend to go supersonic at sea level because they have the thrust to pull it off.

Best,
-Slashy

Edited January 14, 2018 by GoSlash27

[bewing]

OK, some points from where I sit:

1) Airplanes are supposed to have wings for a reason. And that reason is to provide lift to make you go up. If you crank your nose up to some high AoA, then you might as well be flying a rocket and not a plane.

2) Jet engines burn a heck of a lot of fuel to go fast at low altitudes (compared to other jet engines at other speeds and altitudes -- rockets are another story).

3) Drag gets nasty once you go over 240 m/s or so.

So, as far as I'm concerned, it's mostly a question of "how high is good enough for going supersonic in this plane?" So once I get off the ground, I adjust my throttle for slow acceleration, and my nose for a 10 m/s climb rate. I take the first part of the climb through the thick air slowly. Once I get to about 5km, my planes are usually getting close to going supersonic -- and it's best to break through that sonic wall quickly. So I go to full throttle then. By 5km altitude, drag is down by quite a bit, and it drops quickly above that.

Some planes have a much harder time going supersonic, as said above, and you have to do it low when you have megatons of thrust -- even if it costs a lot of fuel.

 

[Guest]

Every plane is different.

For rocket-powered planes, point it up at 45 degrees as soon as you're able, then when you hit 500 m/s (should be a bit over 10 km), follow prograde.

For RAPIER-powered ones, it all depends on the aerodynamics, the size, and the TWR.

I know lots of people are recommending an acceleration phase at a shallow climb once over 10k but I have only rarely found that to work well. There are a number of problems I tend to hit if I try it. If I'm flying near the limits of my design, climbing subsonically to 10k doesn't work because I won't be able to get into the ramjet cycle while up there, whereas climbing supersonically at a steeper angle and then nosing down will take a significant bite out of my speed due to the additional drag caused by the manoeuvre. And if I'm not flying near the limits of my design, accelerating at a shallow angle tends to get me to go too fast and parts start burning up. So basically I try to make it so that I need to manoeuvre as little as possible once I've set up my trajectory.

I.e., my usual launch profile is something like this --

1. Take off.
2. Climb at a shallow pitch -- say, 10 degrees pitch, tops, for some designs even starting straight-and-level -- until I'm in the ramjet feedback loop (TWR starts climbing briskly).
3. Once TWR is close to 2 and climbing, nose up to anywhere between 15 and 25 degrees, depending on the design.
4. At 12-17k or so, depending on the design, start following prograde. 

This should leave me flying near the thermal limits of my plane while climbing briskly to Ap, leaving me a rocket-powered circularisation burn of maybe 200-300 m/s, depending.

As to aerodynamics, what @bewing says about wings is oh so very true. Having enough wing makes a huge difference -- if your attitude and prograde markers don't line up, you're producing massive amounts of drag, so make sure you have enough wing to do that (without having to angle up the wings themselves more than a little bit). It's amazing what a difference even a slight adjustment to the wings can make!

[FancyMouse]

Really depends on my patience. When I do have patience, I will design a really low TWR plane for maximum efficiency. That usually ends up having to dive from 8km to 7km for breaking transonic drag. When I don't have patience, then I just slam more engines and keep constant 10 degree AoA to send it to space.

[Laie]

18 hours ago, foamyesque said:

High AoA certainly can, but the other factor here is that to really beat thrust-per-drag you need to *already be going fast*.

I sense a misunderstanding: all the airbreathing engines have two thrust curves, one for air pressure, one for airspeed, that are combined to compute your thrust. The airspeed curve is what Brikoleur describes as "ramjet feedback loop". The pressure curve is another beast.

As you go up, thrust decreases not as quickly as air pressure (up to a point at any rate). That's true even of the lowly Juno, though in that case it doesn't amount to much. And I've been cutting several corners when I simply described it as "thrust per drag". Overall thrust still goes down, of course, and if you have a low TWR to begin with there'll be little oomph left to accelerate with at high altitudes.

9 hours ago, Brikoleur said:

Once TWR is close to 2 and climbing, nose up to anywhere between 15 and 25 degrees, depending on the design.

Well, that sounds more like a rocket launch. And an overpowered rocket to boot.

[foamyesque]

3 hours ago, Laie said:

I sense a misunderstanding: all the airbreathing engines have two thrust curves, one for air pressure, one for airspeed, that are combined to compute your thrust. The airspeed curve is what Brikoleur describes as "ramjet feedback loop". The pressure curve is another beast.

 

That's true, but it is, in practical terms, often much easier to get the airspeed feedback loop going first, when you're low down, because even if the engine's pressure curve says it can operate more efficiently higher up (on a thrust/drag basis), as you get higher up it has to compensate for the loss of lift (hence flying at higher AoAs and/or requiring higher control deflections, both of which are murder on your drag profile, and can cause cosine losses and impact your intakes as well). This is in my experience particularly noticeable with planes with relatively low liftoff TWRs, i.e ~0.3.

[Spricigo]

So, there is the method you use  and there is the method you more often see people using. And you want to know which one is "better".

So you just nees to try that method you see so often and see how it goes. 

 

Certainly there is some reasons to go supersonic while still low in the atmosphere. But if those are good enough reasons for you, that is another story. 

[OHara]

I also expected, maybe based on common real-world airplanes with prop and turbofan engines, that a subsonic climb would save fuel.  Experimentally I find that, for spaceplanes, I use less fuel to get to orbit by accelerating at low altitude more than feels right to me.

13 hours ago, Laie said:

all the airbreathing engines have two thrust curves, one for air pressure, one for airspeed, that are combined to compute your thrust.

I think of it in a similar way, and can make some rough conclusions if I use some rough approximations.  In the climb, more energy goes into gaining height than into kinetic energy, so roughly we want the 'best rate of climb' airspeed as we climb out of the lower atmosphere.  That would be the speed with maximum excess power, the maximum value of (thrust − drag )·velocity.   Ignoring lift-induced drag for now, that excess power would be
  (thrust − A Cd density v² )·v 
and the drag coefficient Cd peaks a little at mach-1 but doesn't change much through the climb compared to the other factors.   As you say, the thrust decreases with higher altitude a bit less than density does, giving an advantage in thin air, but then the question is how best to get to the thin air.  

For every engine except the turbo-fan, thrust remains constant or increases with airspeed up to mach-1 or higher. In these cases, that factor of velocity in the excess power really favors higher speeds. If I oversimplify to thrust being independent of airspeed, the excess power is some_constant × (w² − v² )·v  where w is the max maintainable speed at level flight, and in this case excess power is highest at v= w/√3 = 58% w --- best rate of climb is at 58% of the max speed in level flight.   KSP's Whiplash and Rapier engines have thrust versus airspeed peaking around mach-3-to-4, so their best-rate-of-climb speed tends to be around mach 3 wherever the air is thin enough for thrust to overcome the peak drag around mach-1.  

[foamyesque]

Just as an example of the kind of plane I'm talking about:

 

 

This needs to get up to around 700m/s before a climb of more than 5 degrees is feasible, at which point it's already at the altitude where you'd want to be flattening out for a speed run anyway. Delivers 180t pure payload to orbit (the piggy-backed modules), and another 80t or so in the carrier's dry mass + leftover fuel.

[xlm]

On 1/14/2018 at 11:59 PM, bewing said:

1) Airplanes are supposed to have wings for a reason. And that reason is to provide lift to make you go up. If you crank your nose up to some high AoA, then you might as well be flying a rocket and not a plane.

Not necessarily. I have two rapiers on my measly SSTO that was meant for low orbit carrying a payload of about 1t. The TWR is so big you can pretty much fly it like a rocket (at least at low altitude). The wings are there for reentry only.

Of course, that's just my over-powered poor design. A more sensible design would have a more sensible TWR, and thus needs to fly horizontally.

[Spricigo]

8 minutes ago, xlm said:

Not necessarily. I have two rapiers on my measly SSTO that was meant for low orbit carrying a payload of about 1t. The TWR is so big you can pretty much fly it like a rocket (at least at low altitude). The wings are there for reentry only.

Of course, that's just my over-powered poor design. A more sensible design would have a more sensible TWR, and thus needs to fly horizontally.

So, you could as well be flying a rocket in the way up if you didn't wanted the wings for reentry. Which seems to be his point.

BTW don't confuse 'sensible TWR' with 'low TWR'. A more powerful design will have their own advantage, such as a shorter time to achieve orbit and less trouble with drag. Rather than being paranoid with efficiency, go for effective designs that got the job done.

 

Personally, I'm more of a rocket guy and even my eventual spaceplane tend to have a healthy TWR and not much wing. So I indeed switch from "flying a plane" to "flying a rocket" at quite a low altitude.

 

[Daveroski]

If I have built it right, I lift off at about 5-10 deg. Almost immediately level off at full thrust.

Then I shouldn't need to touch it again until the air is too thin to feed the engines and it's time to switch engine modes.

After my initial level off, It will gradually climb by itself. On prototypes I may have to nose down a little if it wants to climb too fast.

 

[GoSlash27]

4 hours ago, Spricigo said:

BTW don't confuse 'sensible TWR' with 'low TWR'. A more powerful design will have their own advantage, such as a shorter time to achieve orbit and less trouble with drag. Rather than being paranoid with efficiency, go for effective designs that got the job done.

^ Agreed. The only practical application in KSP for an SSTO spaceplane is to get your payload to orbit as cheaply as possible. Since the only expenditure is fuel, efficiency is measured by how much fuel is expended per tonne of payload. Going with a higher t/w may reduce your fuel expenditure faster than it reduces your payload capacity, saving you both time and money. The most cost effective SSTO spaceplanes I've seen had roughly twice as much engine as they actually needed.

Best,
-Slashy

[Wanderfound]

My general pattern:

1) Moderate climb (ten degrees or so) until I hit the transonic drag.

2) Drop the nose however much is necessary to maintain acceleration until I'm through the drag (400m/s +).

3) Back to the moderate climb until 10,000m.

4) Drop the nose again and crank it to as close as possible to 1,500m/s. Temperature is usually the limiting factor.

5) Pull up into a moderate zoom climb (about 3g, held until the apoapsis is over 20,000m) while flying at around 13,000m. You'll lose a bit of speed initiating the climb, but if you do it low enough you can regain that speed before the jets lose power.

6) Keep the nose on prograde and the engines in air-breathing mode until 29,000m or so. Do it right and you should hit that altitude at about 1,400m/s.

Edited January 17, 2018 by Wanderfound

[Freds]

I never minded when to break sound barrier....

My cargo space planes are designed to take off at mach 0.35 with full cargo, and then goes into a step climb until 12000m . At 12000 m its just a small climb trying to get the maximum horizontal speed possible, until reach 18000m. At 18000m, another step climb until 150000m. The droptanks (I never do cargo SSTO - IMHO droptanks give extra deltaV and cargo capacity without add extra complexity)  usually are enough to raise apoapsis to 120000m with full cargo.

My crew transfer spaceplanes are SSTO, i just go subsonic  until 10000m to save fuel, and then goes fullspeed until the LF only tanks are near depletion. After that, it is the same as cargo planes - climb to 150000m, circularization, etc.

[Guest]

13 hours ago, GoSlash27 said:

The only practical application in KSP for an SSTO spaceplane is to get your payload to orbit as cheaply as possible. Since the only expenditure is fuel, efficiency is measured by how much fuel is expended per tonne of payload. Going with a higher t/w may reduce your fuel expenditure faster than it reduces your payload capacity, saving you both time and money. The most cost effective SSTO spaceplanes I've seen had roughly twice as much engine as they actually needed.

Hear hear.

I would add a less tangible measure of efficiency -- how easy/quick/fun they are to fly. A low-TWR plane just isn't all that much fun, you have to spend a long time nursing it to the point where it can actually do its job. Since fuel is really cheap and the craft is fully recoverable, if you have the capability, why not overbuild so you can get on with it? That is, unless your specific personal objective is to go for maximal efficiency every time.

(This doesn't apply to truly extreme designs, but I've found it relatively straightforward to build planes that loft up to 200 tons or so while remaining easy to fly.)

----> tangent

It is fun to chase efficiency too, but for me personally only when I'm doing it because there's something I want to do that's at the limit of my tech tree and my design abilities. This does apply to SSTOs earlier in the career. But later on? Yeah I'll happily build an SSTO that can loft 200 tons if my payload weighs in at 130, or use my 70-ton lifter for a 20-ton load. 

-----> another tangent

Inspired by this thread, I tried again experimenting with launch profiles. Specifically, I attempted a launch with my heavy lifter (design capability over 200 tons, actual payload mass about 150) where I did the shallow acceleration run from 10k up. Guess what? It didn't work. I did manage to keep my jets burning longer and hit 1500 m/s without any bits falling off, but I was climbing much too slowly. When it was time to switch to closed-cycle, I had to point my craft up and ended up burning more rocket fuel lifting my Ap than with my usual launch profile for this craft, which is a constant 17 degrees pitch once it reaches 5k@400 m/s, then prograde from 25k altitude up. That way my jets flamed out at "only" about 1300 m/s but I had a high-enough Ap that I had no trouble making orbit and ended up burning a great deal less fuel. 

OTOH when I tried it with a smaller SSTO it did work as described: it was a bit cheaper to climb subsonically to 10k, then level out to 5 degrees to accelerate, then nose up and be on my way. So it really does depend on the specifics of the craft and the launch.

[GoSlash27]

Brikoleur,

 In my experience, 10km is really too high. I've gotten my best fuel efficiency at 3-6km for low thrust spaceplanes. You might want to try various altitudes if you're inclined to experiment.

But yeah, it really does depend on the design of the spaceplane.

Best,
-Slashy

[Spricigo]

58 minutes ago, Brikoleur said:

Hear hear.

I would add a less tangible measure of efficiency -- how easy/quick/fun they are to fly. A low-TWR plane just isn't all that much fun, you have to spend a long time nursing it to the point where it can actually do its job. Since fuel is really cheap and the craft is fully recoverable, if you have the capability, why not overbuild so you can get on with it? That is, unless your specific personal objective is to go for maximal efficiency every time.

 

 

Yes, that is close to the point I was trying to raise. 

Trying to design a "perfect" spaceplane can take a good amount of time from actually flying it. It often is uncessary and don't bring  significatively better performance.

Of course, improving your designs is a good thing. But it can be done with patience and fun, instead of with frustation.

 

 

[foamyesque]

@Brikoleur:I'd be interested in seeing those heavy-duty spaceplane designs. Also, a question: When you say 'payload', is that counting all the miscellanceous hardware in the launcher (wing, engines, etc) or is it mission gear -- passenger modules, deployable craft (satellites, spaceport modules, etc), left-over-fuel if you're doing refuel runs...?

Edited January 17, 2018 by foamyesque