Favourite spaceplane engine layout

[AeroGav]

What's the most common one you use on your designs / what'd you recommend for a NEWB?

 

•  All RAPIER

Pros - fully reusable, fuel always drains evenly, good (best?) payload fraction to LKO, high thrust weight ratio

Cons - low ISP in space , not optimal for missions beyond low orbit

• 2 RAPIER 2 NERV 

Pros - fully reusable, can fly to space without oxidizer. Symmetrical layout. Decent delta V.  Once fuel type for air breathing and rocket engines.

Cons - one nerv can only lift 15 tons whereas a single rapier can do twice that, at a pinch, so you are carrying one more rapier than you really need - and those suckers are heavy.  2 tons each.  Also the NERVs drain fuel front to back of the stack like a rocket , so won't take from wing tanks and try their hardest to mess up CG

• 1 RAPIER 2 NERV

https://kerbalx.com/AeroGav/Astrojet-Citation

 

Pros -  fully reusable, can fly to space without oxidizer. Symmetrical layout.  Good Delta V 

Cons - poor performance when subsonic, hard to bust through the sound barrier.  Swapping RAPIER for Whiplash improves this at the expense of a little delta V though lower airbreathing max speed.

• 1 RAPIER 2 NERV  2 disposable panthers (coupled to rear of NERVs)

Pros - solves the poor subsonic performance of the above layout, nice controllable climb not as banzai as a pure RAPIER/Whiplash design at high speed.

Cons - not fully reusable, though Panthers are only a third of the price of a RAPIER and enable to sort of missions you'd need a pretty big rocket for.

https://kerbalx.com/AeroGav/K211-Tundra-Goose

• NERV  + disposable whiplash on decoupler behind NERV

Pros - very high delta V, can go to Duna and back, since you're not dragging the  dead weight of air breathing engines once they outlive usefulness.

Cons - not fully re-suable, though still only 2000 kredits a pop

https://kerbalx.com/AeroGav/whippynerv

 

 

• NERV  + disposable RAPIER

https://kerbalx.com/AeroGav/Astrojet-Citation-X

 

The last word in performance.   Re-using engines is for peasants.

Edited June 8, 2016 by AeroGav

[teag2]

My first SSTO used 2 Rapiers and 1 NERV. I didn't bring enough oxidizer and fell back down to 55km before achieving orbit.

I might try a two RAPIER two NERV sometime. It shouldn't be too hard to achieve orbit (and beyond).

[Brainlord Mesomorph]

Actually I have decent results w/  2 turbojets and 2 thuds.

[mk1980]

plain Rapiers are probably the easiest for a new player.

something like the simple "Dove" stock design works nicely - you don't have to be a genius to get that thing to orbit and it's a fairly simple design.

or a somewhat bigger Mk2 plane with multiple rapiers. ignore efficiency at first. it's much easier to get an overengineered plane to orbit (with more thrust than it really needs).

though the whiplash & aerospike combination is also quite nice to fly (there's also a stock plane with that engine config).

i wouldn't recommend Nervs for a new player. long range SSTO's are more of a  "niche" thing (or an "advanced" use - depending on what you actually want to do with those SSTO planes once they are in orbit)

 

[AeroGav]

57 minutes ago, mk1980 said:

. ignore efficiency at first. it's much easier to get an overengineered plane to orbit (with more thrust than it really needs).

though the whiplash & aerospike combination is also quite nice to fly (there's also a stock plane with that engine config).

i wouldn't recommend Nervs for a new player. long range SSTO's are more of a  "niche" thing (or an "advanced" use - depending on what you actually want to do with those SSTO planes once they are in orbit)

 

Its a tradeoff between not having enough thrust and not having enough fuel.    My first designs were rapier only or 2 rapier 1 nuke,  and it would zoom off like a rocket (!) as soon as i went closed cycle, but usually run out of fuel before making stable orbit.

My 2 NERV / 1 Whiplash design has about 30kn drag and 120kn thrust in nuke mode, though the drag gets lower as we get higher up.   That's holding a nice constant 5 degree angle of attack from mach 5 onward with Mechjeb  -  if you pull the nose too far up to 15 degrees you'll get 50% more drag, and if you really mishandle things it might stop accelerating altogether.    But it generally makes orbit with 30% remaining fuel.   With 3 rapiers instead we'd have 4 or 5 times as much thrust, so you can brute force your way up, but then ISP is 3x worse so you'll run out of fuel if you don't fly the profile exactly right.

The only hard thing about a Rapier design is fuel feed.

 

Edited June 8, 2016 by AeroGav

[mk1980]

True, but with  enough thrust you can just carry a lot more fuel without any issues.  Getting a rapier ssto to orbit is really simple. It only gets a bit harder when you also want to carry a significant payload,  but thats more of an advanced lesson. 

[cantab]

Well my experience is in FAR, stock can be significantly different. But so far my preference has been RAPIERs as primary propulsion, and Panthers as VTOL lift engines. The former offer simplicity and high jet speeds, while the latter have the highest static TWR of any jet in wet mode. In a design where the Panthers can thrust forwards, such as a "tailsitter" or a tilt engine/wing design, then they can do double duty - extra power to blast through the early ascent, and range extenders for approach and landing.

Not enough TWR is a problem during the push to hypersonic jet speeds and the switch to rockets. One iteration of my 40 ton SSTO Gallifrey used four Panthers and one RAPIER, and I found the RAPIER just squandered its thrust fighting drag and it failed to make orbit. Switching to two RAPIERS and three Panthers, with everything else the same, made the difference - even though the airframe was a little heavier, it wasted less thrust as a percentage and thus made LKO.

Ascent profile matters too. I admit I still know very little, but one thing I've found is it seems good to have some vertical speed before making the switch to rockets, that way you aren't pulling such high draggy AoA trying to pitch up to climb out of the atmosphere. A corollary is it's not so good to do the jet acceleration as high as possible where the plane can barely fly level, being a bit lower seems like it can be helpful.

I like the various non-SSTO concepts suggested. Letting a chunk of mass go is not to be sneezed at. It might be fun to take an established SSTO spaceplane, and see how much more payload you can get out of it using drop engines or drop tanks.

[AbacusWizard]

I generally use a single Aerospike rocket engine on the central fuselage flanked by a pair of Whiplash jet engines attached to the wings. All it can really do is get to low orbit and back, but that's all it needs to do--I use it to shuttle crew (and science) between the ground and Starbase Alpha, and then use re-usable high-efficiency orbital crew transport ships to move them from Starbase Alpha to... well, anywhere, really.

[suicidejunkie]

My sized SSTMinmus has 2x Whiplash + 2x Nukes. Mk2 size with Fat-455 wings, 6 crew and ~2 tons of cargo.

 

AP of 72km opposite KSC after ~12 minutes on the MET, with 1500+ m/s in the tank for landing on Minmus next to the fuel refinery.

Profile is a rapid climb to 15km, level off and accelerate to 1500m/s during a slow climb up to ~22km, finally nukes on and maintain ~5m/s climb and ~4deg AoA while accelerating to 2km/s.  Faster climb and less AoA as the speed approaches orbital velocity.

[Rune]

If what you are after is dV in orbit (as it seems you are). Minimizing engine mass for a given takeoff weight is key, to maximize mass ratio once in orbit. So it is rather obvious that you should use only the minimum amount of RAPIERs required to get out of the atmosphere, and then the absolute minimum number of Nervs, I.E. one.

Yup, better to load some oxidizer for the jump to almost-orbital velocity, since it is only a bit of extra tank weight on orbit and you already need the RAPIERs to climb through air. Case in point, giving something like 3.5km/s on LKO with a respectable payload and two RAPIERs plus a single Nerv:

 

Rune. For everything else, plain RAPIERs will do.

[mk1980]

in general i agree with Rune , but there is also a point where minimizing engine count seems to be counter productive. it's probably more of an edge case and not very relevant for long range SSTOs, but if you don't pack enough punch to get out of the atmosphere reasonabyl fast, you may end up losing more deltaV in the ascent than you would gain from having a few tons less (unused) engine mass in space.

well, at least that was my experience when experimenting with minimum engine count planes. there's currently a challenge in the subforum about getting a 72 ton payload (2 orange tanks) to a 300km orbit and my design used only 8 rapiers. in that case, i think some additional engines would actually have improved the performance.

though in the long run, the reduced engine mass will probably trump the additional fuel you burned in the ascent - empty tanks are less dead weight than unused jet engines.

[stenole]

A combination of whiplashes and rapiers. Enough rapiers to get sufficient TWR in closed cycle mode for the last part of the ascent. Whiplashes are there to supplement the rapiers to blast through the sound barrier and gain altitude relatively fast.

[Red Iron Crown]

All Rapiers here outside of stunts. Spaceplanes for me are about ferrying to LKO and back down, so they don't need more orbital dV than a pittance for rendezvous and deorbiting. Specialized high efficency vacuum engines are a waste of mass in this scenario as they rarely save enough fuel to offset their additional mass.

When I did Laythe-and-back in an SSTO (definitely in the "stunt" category), I went with 2xRapier 1xLV-N.

[Foxster]

My basic Mun-return SSTO...

[RocketBlam]

For early spaceplanes, its possible to use the Panthers on afterburner to get up to 1000 m/s or so, and then switch over to a regular rocket like an LTV-30 or maybe two Terriers. After I get Rapiers, I find that using regular jet engines just isn't efficient. 

[Van Disaster]

Let's talk generics a little, first.

* What's the spaceplane for? I generally build spaceplanes because I like recovering as much as I can direct to the runway ( and it makes re-entry with a payload rather easier ). This means unless it's a LKO shuttle I'm effectively building a winged rocket, and what's most important is how it works in orbit as long as it can fly pretty well when it's mostly empty. Ergo, brute-forcing it off the runway & up to altitude is fine ( even if that means strapping it to a rocket & launching straight up ).
* Spaceplane doesn't equate to SSTO, which follows on from above.

I use FAR which rather punishes you for strange shapes at high speeds, and I use part mods. Having got that out of the way, my favourite shuttle shape looks like this:

Tanks are balanced round the cargo bay, the dead mass of the engines is on top of CoM ( the engines are stock RAPIER equivalent, but I'd do the same with the bigger B9 engines also ). Give me dual-mode engines and a reasonably sized craft and I'll try and duplicate that shape with whatever.

This career run I've used a lot of spaceplanes though. Very early days with Juno & LV-T needed vertical launch - initially with SRBs to make sure it worked, and then some more recoverable boosters:

You *can* SSTO with Junos - there's a thread about it in Craft Exchange - you just can't do anything particularily big. Useful for putting up fully recoverable sats for contracts.

Swapping the LTv for a Terrier helps the early plane a little but doesn't change anything about launch. The next major step for me was the Panther ( and Thud ), because the packaging & performance balance made SSTO doable. I could probably have built something a little more efficient old-style out of a pair of Panthers and a LV-T but it'd have been pretty large ( and draggy ), and this worked ok. The engine is the B9 equivalent of the Panther - slightly more powerful & heavier, so not really overall that different.

I tried a Mk2 version but it didn't really work out until I unlocked LV-Ns - and at that point I gave up on SSTO again for larger craft because the terrible LV-N TWR. So... going back to brute-forcing it to orbit, SRBs on the wingtips to give it a kick up work just fine - they cost a little jet fuel to haul up to transition altitude, but who cares when it's just jets burning?

Unfortunately wingtip SRBs somewhat preclude engine pods out there, but that craft is balanced full or empty anyway. I should note that that *isn't* just a LKO shuttle.

And that's pretty much where I am in this career game, no funds to upgrade the R&D building higher yet. I've built tons of spaceplanes over the years - certainly not going to dig up pics of all of them - but the arrangement of the first one in the post is usually the most trouble free.

Edited June 13, 2016 by Van Disaster

[Adelaar]

Nice and sleek SSTO designs Van Disaster! 

My own SSTO's are of the brute force kind, and are almost always RAPIERS only, except a few early career SSTO testbeds using ramjets and aerospikes. 

When constructing SSTO's I kind of like to have one RAPIER for every 10-13 tons. So for a 50000 kg craft I use 4 engines. This has pros and cons of course. The pro is a smooth ascentprofile and controlability. The cons an inability to properly do interplanetary trips. 

[SpaceplaneAddict]

  1 Rapier, plus 1 Nuke clipped inside

[KerikBalm]

Pure Rapier, or 2 rapier, 1 nerv (or even higher rapier:nerv ratio)... my spaceplanes are for going to/from orbit...the vacuum Isp of the rapier isn't a problem for me.

Orbit can be of Kerbin, Laythe, or Duna.

Laythe its pretty much the same... rapiers get you very very close to orbit before going to closed cycle. I do pure rapiers or a rapier- turbo combo (turbos for better atmo Isp when cruising around looking for a landing site, plus better TWR on takeoff which is important for rough terrain)

Duna... LV-Ns... and some chemical rocket for an initial boost... skippers, poodles, aerospikes...