Optimal intake layout for SSTO/SpacePlanes?

[panzer1b]

Well as the title says, im looking for the most optimal intake layout, primarily for my SSTO designs. While they have been getting better and better every day, one thing id like to know is the optimal number of intakes, and what types of intakes are the bets to use. This assumes stock game, no edits to part properties, no FAR, ect. I also airhog somewhat, although i tend to cut down on intakes, since they are dead weight in space and cut dV alot, although i like to be able to get out of atmo entirely before kicking in the rockets.

Now ive heard that 8 RAMs/engine are the best in general, but im wondering whether the other types (or perhaps a mix of intake types) can be better, or should i just stick to rams? Also, i dont care about looks/size of intakes as i clip them inside my hull anyways, so its not like that makes any difference what ones i choose. Id just want to know from a SSTO perspective, what is the best layout for getting into orbit, as well as minimizing weight one i get out of atmosphere (obviously the cones have best surface area, but they weight so much more that it feels they do more harm then good at least in the long term, once i am using my nukes in space). Ive made a very good SSTO that used 16 rams for 3 RAPIERS, so it works, although it is a little underpowered at altitudes above 45km (its dragging 20 tons that doesnt exactly have too much lift).

Is there a benefit to storage capacity for intake air as well, or does the only thing that matters is the raw surface area to mass ratio?

Also, id like to know if there are some differences in intake optimization between the rapier and the turbojet, as im guessing the turbojet can go a little faster and perhaps can squeeze more out of less intakes?

Edited December 6, 2014 by panzer1b

[LABHOUSE]

It is more efficient to use a turbojet or two to get out of the atmosphere then switch to a nuke engine or ion engine to get into fuel orbit. The optimal intake layout is any layout just not too many.

[GoSlash27]

I've been researching this subject this week. Looks to me like the optimal number/ type of intakes is 4 of the XM-G50s per engine. Any more than that, and the gains from air are erased by the drag losses. Put 'em as far aft as you can.

One engine per 13 tons of spaceplane and roughly 1.0 of lift coefficient per ton of spaceplane. Porkjet strakes are the best wings to use.

This is all a rough preliminary estimate and is therefore subject to change.

Best,

-Slashy

Edited November 26, 2014 by GoSlash27

[Claw]

What Slashy states is very close to my stock rules of thumb.

- 1 TurboJet per 12 tons, (no more than 15 tons)

- 3 Ram intakes per TurboJet (no less than 1 per TurboJet)

- 150 liquid fuel per TurboJet (no less than 100)

- 1.0 lift rating per ton (No more than 1.0, no less than 0.5 per ton)

These criteria are interchangeable. So if you have a craft that's lighter (i.e. 1 TurboJet for less than 12 tons), then you can ease up on the fuel requirements, or intakes, or lift rating. I can give you more reasoning behind why I chose these, but it was a bit more in depth than "because it feels right." These rules of thumb also utilize a climb profile that is very close to terminal velocity on the way up.

Also, if you're interested in reducing flameout chances, then you want to place your intakes/engines in an alternating pattern. Split up your intakes equally between your engines, then place "Intakes", "Engines", "Intakes", "Engines", and so on.

Cheers,

~Claw

[Jouni]

RAPIERs and turbojets are very similar in the stock game. Here are some rules of thumb:

• If you plan on carrying heavy rocket engines such as nuclear engines, use them for circularization. RAPIERs offer no benefits in this case.
  
• The turbojet is rated for 225 kN (112.5 kN at 0 m/s), while the RAPIER is nominally 190 kN (95 kN at 0 m/s). If you can fly with the same number of turbojets and RAPIERs, choose the RAPIER, unless there are other reasons for selecting the turbojet. If you would need more RAPIERs than turbojets, select the turbojet.
  
• If you plan on using enough intakes to get your surface speeds above 2000 m/s, use turbojets. RAPIERs lose thrust faster at those speeds.
  

Ram intakes are the most mass-efficient option under most circumstances. When you're flying at near-orbital velocities at high altitudes, with engines throttled down to prevent flameouts, your total thrust is determined by the number of intakes you have. As thrust is primarily needed to overcome drag, which is proportional to mass, the relevant metric is intakes/mass, not intakes/engine.

In a typical ascent, you might be flying at 1800 m/s at 32 km. Because that's pretty close to the terminal velocity at the altitude, you'll need 4-6 kN of thrust per tonne to overcome the drag, depending on the average drag coefficient. A single turbojet can produce 135 kN of thrust at that speed, while a RAPIER can produce 114 kN, assuming that you can provide them with enough intake air. As a rough approximation, any thrust exceeding the drag will make you accelerate. If you can accelerate at 32 km, you can keep accelerating all the way to orbit, because both intake air and drag decrease at the same rate.

Unfortunately, intake air (and therefore acceleration) decreases exponentially with altitude, halving once every 3.5 km. With 1 intake/10 tonnes, you probably can't get any meaningful acceleration beyond that 1800 m/s at 32 km. Every time (within reason) you double the number of intakes, you can keep producing the same thrust 3.5 km higher, where the effective maximum velocity is 40% higher.

Hence, assuming that 1 intake/10 tonnes will allow you to accelerate to 1800 m/s in a reasonable time, 2 intakes/10 tonnes will be enough to reach the point where the number of engines becomes the limiting factor again. Turbojets drop linearly from 112.5 kN at 2000 m/s to 0 kN at 2400 m/s, while RAPIERs drom from 95 kN at 2000 m/s to 0 kN at 2200 m/s. While doubling the number of intakes to 4 intakes/10 tonnes will make you accelerate faster, it won't affect the maximum speed significantly.

Some of the numbers above are rough estimates, because I haven't used airbreathing engines that much in the stock game for a long time. As a rough guideline, I would say that:

• 1 intake/10 tonnes will get you to orbit.
  
• 2 intakes/10 tonnes will make the ascent much more efficient.
  
• 4 intakes/10 tonnes will make the ascent faster, with minor efficiency benefits.
  
• Doubling the number to 8 intakes/10 tonnes will yield only very minor benefits, if any.
  
• Doubling the number again to 16 intakes/10 tonnes will probably reduce overall performance due to increased drag.
  

[GoSlash27]

An example I just knocked together to illustrate:

This one just transfers 4 kerbals to LKO with a 2 kerbal flight crew and a docking port. It's close enough to perfectly balanced that no control surfaces are necessary.

It uses just 80 KG of fuel and 35 KG of oxidizer to do the job with plenty of DV for rendezvous and docking and a powered landing back at KSP.

Best,

-Slashy

[LordFjord]

Here is some research I recently did on intakes.

Might be useful. Most important insight I got from it: it does not matter what type of intakes you use, you just have to use the right number of them. Also, balancing the intakes helps a lot in order to be able to let the jets run higher and higher.

Regarding jetpowered SSTOs, intakes and stuff, you might be interested in the mod I am working on, you can find it over here

Its still in development but works fine for the most common intake/airbreather engine setups. More details in the linked thread

[/shameless self-advertising]

Edit: after dealing with the flameout issues a bit more in detail, I think that you do not need that many intakes as some propose (16 rams on 3 jets). What you want is to let them run at max. altitude without flameout. The number of intakes for that are not that important, their distribution is. By the way, 15 intakes ordered correctly would be a lot more efficient for that craft

What a higher number of intakes without any thought of placing order is:

- the 1st engine will be maxed on air, the next one only gets the "leftover" air

You can save a lot of intakes and part count if you do the placing order right. See the pic in my "research on intakes" link. That's a single RAM for a jetengine, and it flamed out at 32km altitude. That is sufficient for SSTOs and it should be viable for most designs.

I will try to build a midsized SSTO with a relatively small amount of intakes and see how it behaves (after ofc applying the magic of my mod).

Edited November 26, 2014 by LordFjord

[GoSlash27]

Doubling the number again to 16 intakes/10 tonnes will probably reduce overall performance due to increased drag.

Something to keep in mind is that the intakes are not only a source of additional drag in this regime, but are the *primary* source of drag.

This is why adding intakes quickly becomes detrimental; You get a small increase in thrust with a huge penalty in drag.

Best,

-Slashy

[Pecan]

...although i tend to cut down on intakes, since they are dead weight in space and cut dV alot...

A SSTO should be performing exactly ONE manoeuvre once in orbit - a de-orbit burn.

All the structure of a plane is deadweight in space. Most of the structure of a plane (eg; wings) is deadweight even during launch from Kerbin, wings are only really useful for landing accurately.

Detach the useful bits once in orbit and re-dock with the landy bits again when the mission is complete.

And for the record - I use 8 RAMs per engine and don't have any problem arranging them aesthetically:

Crew Shuttle MK2 In SPH

But then I'm not trying to make something that looks exactly like everyone else's, nor replicate something that didn't/wouldn't even work on Earth anyway.

With 8 RAMs I do the 'speed burn' around 28-30km, with the engines throttling-back above 33km. Apoapsis goes to space around there and I keep my ascent flatish to bring the periapsis to around 50km by the time the jets fameout and I'm in space. That leaves <50m/s deltaV for circularisation at 75km, which is so easy 48-7Ss are usually all the rocket-power I have. Detach payload and send it off on its mission. If it has to land again later re-dock to the lander and bring it all down (but mostly what goes into space gets reused in space, so my SSTOs are almost purely launch-vehicles).

[GoSlash27]

A SSTO should be performing exactly ONE manoeuvre once in orbit - a de-orbit burn.

"Should" is a subjective word. That statement is true for anyone who uses SSTOs to put payloads in orbit without regard to any other consideration, but is false for everyone else.

Any SSTO *I* use will not only need to deorbit, but also match inclination, intercept, rendezvous, and dock with something else while in orbit. It's also a plus if it has some fuel reserve to land under power.

Others often use SSTOs as an interplanetary vehicle. I don't personally, but for those that do the DV is pretty critical.

Having said that, what people use SSTOs for *after* they're in orbit is immaterial. What matters is how we go about getting our SSTO into orbit with minimal fuel, expense, and unnecessary mass.

Best,

-Slashy

Edited November 26, 2014 by GoSlash27

[Red Iron Crown]

I've come around to Jouni's line of thought, it's intakes/ton of mass that is the useful metric. In the upper atmosphere it won't matter how many engines you have because more engines will just require a lower throttle setting to prevent flameout. I prefer the structural intake for aesthetics and convenience of placement, I've had good results with 6 or so per 10 tons of mass. With engines I just use enough to get the initial TWR to about 1.2 (I know a lower TWR can work but I find it a bit tedious).

[Pecan]

"Should" is a subjective word. That statement is true for anyone who uses SSTOs to put payloads in orbit without regard to any other consideration, but is false for everyone else.

Cheerfully granted - I am being subjectively lenient and avoided using the word "Must". Which is just as well because ...

Any SSTO *I* use will not only need to deorbit, but also match inclination, intercept, rendezvous, and dock with something else while in orbit. It's also a plus if it has some fuel reserve to land under power.

Drives a truck through the hole in either statement (you can tell I'm not desperately concerned about the idea). I'd definitely have to grant 'limited orbital adjustments' and landing so maybe it should read "A SSTO, by definition, must only be required to perform ONE significant manoeuvre while in orbit - a de-orbit burn."

Specifically, however, I am arguing against:

Others often use SSTOs as an interplanetary vehicle

Since the TO stands for To Orbit. If something's designed to go to Mun or somewhere in a single stage then all well and good but it is only briefly and incidentally a SSTO; a significant and necessary part of the design and construction being specifically NOT To or even In Orbit.

Elsewhere I have argued for the generic term "Reusable launch-vehicle" if Single Stage To Orbit And Back is meant, but I realise the futility of even getting the message To People, let alone getting them to appreciate it. (Most people will continue to just use SSTO to mean 'spaceplane', overlooking everything else a spaceplane might do and everything else that can be a SSTO).

[GoSlash27]

Since the TO stands for To Orbit. If something's designed to go to Mun or somewhere in a single stage then all well and good but it is only briefly and incidentally a SSTO; a significant and necessary part of the design and construction being specifically NOT To or even In Orbit.

Not that this is an invalid argument, but it *is* a minor point of contention which has no bearing on this discussion. An SSTO doesn't have to perform any act whatsoever in order to count as an SSTO other than 1) achieve orbit and 2) not drop any parts on the way. It matters not a whit what individuals choose to do with them once they're up there (at least from an engineering standpoint).

This particular discussion is simply about intakes for jet/hybrid SSTOs.

Best,

-Slashy

Edited November 26, 2014 by GoSlash27

[numerobis]

If you want to go as fast as possible in level flight, then the theoretical optimum is to have full throttle at the top of the atmosphere, which requires IIRC about a quarter-million ram air intakes per engine, and after that to have as little dead mass as possible. As air pressure falls, both drag and intake air fall at the same rate, so intake drag and intake air cancel out. Say you're flying at an altitude where you're starved for air. Every time you add an intake, you have added an equal amount of extra thrust and extra drag, so you did nothing. Except that now you can climb a little bit and be just as starved for air as before -- but at that higher altitude, the rest of your plane will have lower drag, so you speed up.

You need to lift those intakes to high altitude. The mass will eventually come to trouble you, but only once you have a significant fraction of your airplane being intakes. Intakes are very, very light: 100 of them (for the good intake types) only add a single tonne to your spacecraft. You'll run out of CPU cycles before you have trouble lifting the intakes.

If you want to go somewhere, you're going to pay for the mass of those intakes. So then there's a tradeoff, but to know the optimal point we need to know how far you're going. Of course, if you're optimizing, then you should undock your interplanetary vehicle from the SSTO and leave all the intakes in LKO.

tl;dr: MOAR INTAKES

[Claw]

Kerbonauts,

The discussion of "what's an SSTO" is always a little edgy. Before dragging too far off in the "how do you define SSTO" direction, please remember that the OP is asking about intakes. Feel free to offer your specific play preferences and styles, but try not to hijack the thread too far.

Besides, a quarter-million ram intakes sounds like a lot.

Cheers,

~Claw

[Kulebron]

Moar intakes than 7-8/engine does not improve performance at high altitudes. Higher TWR does.

I had TWR at sea level about 2, then added 1 or 2 engines, and the ship had about 3.5 at sea level. Fuel consumption to LKO actually decreased, because it could shoot itself out of the atmosphere and not catch too much drag at 30+Km altitude.

[edit] I mean not just shooting to 80km apoapse, that's simple. I mean achieving orbital speed in the atmo and still climbing quickly, at 100m/s.

[Jouni]

You need to lift those intakes to high altitude. The mass will eventually come to trouble you, but only once you have a significant fraction of your airplane being intakes. Intakes are very, very light: 100 of them (for the good intake types) only add a single tonne to your spacecraft. You'll run out of CPU cycles before you have trouble lifting the intakes.

Intakes are very light, but their drag coefficients can be very high. A ram intake weights 0.01 tonnes, but it generates as much drag as a 0.1-tonne part at airspeeds above 333 m/s. If you have 16 open ram intakes in a 10-tonne craft, the drag will be 14.4% higher than with an average drag coefficient. With 100 intakes, the drag will be 90% higher than the average. Unless you want to manually manage the intakes or use a large number of action groups for that purpose, getting the intakes to the upper atmosphere will be quite expensive.

[numerobis]

Indeed; that's what MechJeb's "manage intakes" is for.

[GoSlash27]

Intakes are very light, but their drag coefficients can be very high. A ram intake weights 0.01 tonnes, but it generates as much drag as a 0.1-tonne part at airspeeds above 333 m/s. If you have 16 open ram intakes in a 10-tonne craft, the drag will be 14.4% higher than with an average drag coefficient. With 100 intakes, the drag will be 90% higher than the average. Unless you want to manually manage the intakes or use a large number of action groups for that purpose, getting the intakes to the upper atmosphere will be quite expensive.

[numerobis]

You have to scale thrust to mass.

At low altitude, engines provide thrust. You need about one intake every 15-20t if you have enough lift to get you to 20km with TWR < 1.

At high altitude, the number of engines you have typically has no bearing on thrust, because they will flame out. The way you increase thrust is by increasing airflow, which is done by increasing the number of intakes. You need about one ram intake every tonne to establish an orbit at 40km. You can do with less by targeting an orbit at higher altitude, because that requires less speed, but it's trickier.

Since you need to handle both low and high altitude, both the number of engines and the number of intakes scale the same way, so it doesn't really matter whether you think of it as intakes per engine or intakes per tonne.

Closed intakes are almost free. In between low and high altitude, you can just keep opening more intakes to stay at full throttle. You can do it by hand fairly effectively; have an action group to close all intakes except, say, one per engine, then a series of action groups, each one of which triples the number of open intakes (or pick your favourite number, just be sure to grow exponentially). Then you fire an action group about every 5 km (or 5km * log(favourite number)) once you get beyond the reach of your first few intakes, around 15km-20km. Even simpler, just have two groups: one gets you to 25km, the other opens *all* the remaining intakes. You feel a jolt then, but you're going about 1km/s at that point anyway, so you're at max thrust. Or, use the MechJeb intake management, which I wrote for precisely this purpose.

[Jouni]

Now on the subject of scaling intake area to total mass... I confess I'm not sold on that concept. I'm thinking that since the intakes generate the lion's share of drag in the injection phase, the mass of the vehicle itself is of little importance so long as you've scaled everything else appropriately and gotten it "over the hump". An engine will require the same number of intakes regardless of the mass it's pushing at the time (and the drag generated from that mass) in order to stay lit to achieve orbital velocity. OTOH, when you exceed the mass that your engines can handle and have to add another engine, *not*increasing the intakes proportionally along with it would surely lead to flameout problems.

This is the reason why intakes/mass is the relevant metric:

One turbojet, one intake, and a 171x36 km orbit. Something you could reasonably expect with 5 intakes/10 tonnes, but not with 1 intake/turbojet.

Of course, if you have a rule saying that there should be 1 turbojet per 12-15 tonnes, and another rule saying that you should have 3 intakes per turbojet, you also have a rule for 2-2.5 intakes/10 tonnes. On the other hand, if you build different designs, such as jet-boosted rockets or low-TWR planes, you get very different results with 3 intakes/turbojet.

[panzer1b]

Cheerfully granted - I am being subjectively lenient and avoided using the word "Must". Which is just as well because ...

Drives a truck through the hole in either statement (you can tell I'm not desperately concerned about the idea). I'd definitely have to grant 'limited orbital adjustments' and landing so maybe it should read "A SSTO, by definition, must only be required to perform ONE significant manoeuvre while in orbit - a de-orbit burn."

Specifically, however, I am arguing against:

Since the TO stands for To Orbit. If something's designed to go to Mun or somewhere in a single stage then all well and good but it is only briefly and incidentally a SSTO; a significant and necessary part of the design and construction being specifically NOT To or even In Orbit.

Elsewhere I have argued for the generic term "Reusable launch-vehicle" if Single Stage To Orbit And Back is meant, but I realise the futility of even getting the message To People, let alone getting them to appreciate it. (Most people will continue to just use SSTO to mean 'spaceplane', overlooking everything else a spaceplane might do and everything else that can be a SSTO).

Yeah, well my SSTOs all are duna/return capable, with teh ability to land on duna and get back in orbit (provided i dont need to get back to kerbin).

And i guess i shouldnt technically call it a SSTO, but a purely universal do whatever i want with craft. Heck, my most recent flier is SSTO, LATHE CAPABLE (forget the return trip though, got to lathe and landed, thats it). I guess its a SSTOAB (single stage to orbit and beyond). Also able to drag 2 missiles to jool/lathe, so i guess its a little overkill (i usually dump the missiles and hardware for em when im really long distance flying).

Btw, specs for the SSTO are:

16 ram intakes, 1 shock intake (ram doesnt look right in its place)

3 rapiers, 1 nuke

20tons fully loaded

This can make it to a 50km AP on rapiers, given the proper ascent profile (its extremely finicky). Need to get to 20km asap (gun it vertical), level out at 30km, gun it until you are max velocity, then increase your altitude to 50km and gun nuke there, climbing at around 100m/s (a little more initially, a little less later).

gets me into a 89.9KM-AP 88.9KM-PE LKO with 693 LF and 844 OX left in the tanks. Started out with 900 LF, and 880 OX. While far from perfect, i think for a completely "aesthetic" design that i insisted had to have 3 rapiers in a triangular pattern as its main power plant, came out rather overkill in terms of performance.

Here is a pic of it in orbit (100% stock no mods at all)

This tank is all that was drained (and still has some left after getting to orbit) All other tanks are 100% full.

Are 16 rams and 1 cone enough? cant seem to get out of atmo with any less, and adding more seems to eat fuel economy in atmo and especially later once im out of it.

Edited December 6, 2014 by panzer1b

[GoSlash27]

haha yeah 16 rams and 1 shock cone are enough for the job. Honestly, 3 cones or 3 rams would suffice if you fly a proper profile.

(edit) ^ This should read 6 cones or 6 rams. Forgot to carry the 1...

Best,

-Slashy

Edited December 7, 2014 by GoSlash27

[numerobis]

One thing I often forget about: the shock cone is by far the best in terms of intake area per part. The radial and the ram are similar in intake area per part (a ram often requires two parts).

So if CPU time is what limits you, shock cones are awesome.

[panzer1b]

haha yeah 16 rams and 1 shock cone are enough for the job. Honestly, 3 cones or 3 rams would suffice if you fly a proper profile.

Best,

-Slashy

ok, if you know what ur doing, heres a challenge: Take my base craft, and get it INTO ORBIT USING less then 10 intakes (NO WAY YOU CAN DO IT WITH 3)!

http://www./download/rjuv46naa8ma8ps/HK-303+%27Tri-Fighter+V%27.craft#39;Tri-Fighter_V'.craft

conditions:

appearance may not change (muck around with internals as much as you want, but you may not change the generic external appearance outside of minor things like tiny section of wings visible ect). this also means that the 3 rapiers/nuke must stay as is!

total part count must not go up, but you CAN remove the weapons mountings (they add some extra drag/mass that is useless outside of shooting people in MP.

total mass of craft must not go up

you cannot add any fuel tanks (tweak the amount of fuel in each as you please though)

you must get it into a 80km AP/PE OR HIGHER orbit with at the bare minimum 844 oxidizer and 690 fuel left (i made almost 90km with this much oxy and a little extra liquid fuel).

if you can do this, then you are the best SSTO designer ive ever seen!

sorry i dont have any prize to give you if you succeed, but rest assured i will be very impressed, and id love some specific help with an actual design, instead of theory.

this challenge is open to all, anyone who can do this is just pure win!

Edited December 7, 2014 by panzer1b