Larger Jet Engines

[Northstar1989]

A simple request- but the game could really use larger (1.875 meter or 2.5 meter) jet engines.

Even on relatively compact (can be launched from the Level 2 Runway) suborbital and air-launch spaceplanes I build, I end up having to spam 8-12 jet engines or more to get optimal capabilities/efficiency.

This isn't a problem relegated to mod playthroughs.  Even going all Stock, I can build some very large wings out of the structural wing parts (as I should be able to!  Although the part-count spam can get really annoying...) and end up needing to spam engines to go along with it.  If you think about it, this is very "Kerbal"- building it bigger instead of more efficient (the very limited selection of stock spaceplane parts, and oversimplified drag/aero models don't really allow for optimal efficiency).

It would be nice if the game catered better to this very common problem even in Stock KSP (and even more so with mods: which give us more interesting/useful heavy payloads to lift to orbit)- the need to spam far, far too many jet engines on our spaceplanes due to being limited only to 1.25 and 0.625 meter jets.  I would like to see some 1.875 and 2.5 meter Stock jet engine parts!

A spaceplane-focused DLC could even sell well to the community I bet.  The stock spaceplane parts selection is far too limited- hence why some of the most popular mods for KSP have long been spaceplane parts packs...  But console players can't share in the mod-fest, and many players prefer to play the game "out of the box" (Stock ONLY).  So it would really be nice to see more spaceplane parts, and particularly larger jet engines, as part of the base game...

[KerikBalm]

I would also like larger engines.

8-12 engines is nothing though... when I make large cargo SSTOs for 3x kerbin... I'm looking at like 40 engines.

A 2.5m rapier that replaces a quad cluster of 4 would massively help my part counts.

Testing my 3x SSTO in 1x... it becomes able to deliver >100 tons to jool intercept (but getting the return trajectory and aerobraking to land back at Kerbin is tricky)

Spoiler

I eventually changed to reusable 2 stage spaceplanes, which still have high engine counts, but do improve part count and performance a bit, but you can still see that I spam a lot of engines:

Spoiler

 

 

[Northstar1989]

57 minutes ago, KerikBalm said:

I would also like larger engines.

8-12 engines is nothing though... when I make large cargo SSTOs for 3x kerbin... I'm looking at like 40 engines.

A 2.5m rapier that replaces a quad cluster of 4 would massively help my part counts.

40 engines?

*Shudder*.  At that point I'm not sure which would commit self-immolation first: my graphics card or my CPU.

Larger engines are clearly desperately needed.

As for your spaceplane- I could probably give you some useful pointers.  Also some useful mods to try out.  I was the first (by at least a couple months) player I know of to ever build a 3.2x Kerbin spaceplane that worked in one stage (also could only haul about 1.2 tons of cargo, and 2 Kerbals, to orbit though...)  Was done with RealFuels and FAR, plus low-level KSP-Interstellar precoolers (which allow jets to operate efficiently at higher speed/altitude, like they would in real life- unlike Stock precoolers: which are nearly useless...)

 

Anyhow, enough about mods.  This thread was about needing new/larger STOCK jet engines so we DON'T have to rely on mods- or in this case, create our own- as I can't even find a mod with this functionality, other than OPT's 2.5 meter electrically-augmented Scramjets, or the 2m5 meter SABRE engines of the long-defunct original B7 Aerospace Parts...  2.5 meter jets, even in mods, seem to be incredibly rare...

I don't care if I have to buy a new DLC/expansion for it, this NEEDS to happen...

Edited May 13, 2020 by Northstar1989

[jimmymcgoochie]

There are some mods that add larger plane engines and other parts- Near Future Aeronautics has some good ones, Airplane Plus has more (I think) and OPT also has engines in it. NFAero is the one I’ve used most and it comes with a variety of 2.5m engines, from turbofans to turbo-ramjets and various air breathing rockets with sword-related names, and other parts to go with them like air intakes/nacelles/pre-coolers. There are also two nuclear air-fuelled engines that run on just atmosphere and work anywhere with a thick enough atmosphere, even Eve.

[Northstar1989]

7 minutes ago, jimmymcgoochie said:

There are some mods that add larger plane engines and other parts- Near Future Aeronautics has some good ones, Airplane Plus has more (I think) and OPT also has engines in it. NFAero is the one I’ve used most and it comes with a variety of 2.5m engines, from turbofans to turbo-ramjets and various air breathing rockets with sword-related names, and other parts to go with them like air intakes/nacelles/pre-coolers. There are also two nuclear air-fuelled engines that run on just atmosphere and work anywhere with a thick enough atmosphere, even Eve.

NFAero is not updated for 1.9.x and won't be for a lonnnggggg time.  OPT only has a single 2.5 meter, electrically-augmented (borderline cheaty) Scramjet- a fair number of fuselage parts, but zero upsized turbofans/ramjets.  I'll check Airplane Plus, since you mention it.

The bottom line here is that players SHOULDN'T have to trawl an endless and confusing sea of mods hoping to find such a basic functionality, though.  Some simple, upsized jet engines are easily in order for the base game: and quite a few players would probably pay for a spaceplane-themed expansion featuring it...

[KerikBalm]

1 hour ago, Northstar1989 said:

40 engines?

*Shudder*.  At that point I'm not sure which would commit self-immolation first: my graphics card or my CPU.

Hey, I was able to get part count below 200, 180-something if I recall. That leaves about 100 parts for the payload (300 parts is about the limit my computer can handle). Time did move awefully slow. Getting to orbit was like a 30 minut endeavour.

Quote

As for your spaceplane- I could probably give you some useful pointers.  Also some useful mods to try out.  I was the first (by at least a couple months) player I know of to ever build a 3.2x Kerbin spaceplane that worked in one stage (also could only haul about 1.2 tons of cargo, and 2 Kerbals, to orbit though...)  Was done with RealFuels and FAR, plus low-level KSP-Interstellar precoolers (which allow jets to operate efficiently at higher speed/altitude, like they would in real life- unlike Stock precoolers: which are nearly useless...)

Well, I wasn't using a "standard" 3.2 config, but my own 3x config (I can do the math for rotation period changes), with a 1.25x atmosphere rescale (the stock atmosphere is proportionately way too thick).

I'm not interested in part mods, I wanted the part mods to be limited to just Stock+DLC (even if the solar system was modded). I don't play with FAR, because I think if I did far, then I'd want to ditch the lego piece wings and go with some sort of procedural or mod wing set, when I want to keep my craft stock+DLC.

The precoolers that you mentioned for instance... are probably OP'd. IRL Sabre engines would operate to mach 5.something. In KSP, rapiers do that. IRL, 1,600 m/s surface/1800 orbital  is a small fraction of orbital velocity (nearly 8,000 m/s), in stock its almost all the way there, in 3x (or 3.2x), it gets you roughly half way to orbit... IRL to get half way there, you'd need to go at least 3,500 m/s in the atmosphere... so I don't think better jet performance is realistic for a 3.2 or 3x scale.

I also don't want to part clip or exploit, and I want to be able to carry large and oversize payloads to orbit... so a 1.2 ton to orbit in a just slightly larger system doesn't impress me much (IIRC, I was the first, to post on the forums anyway, to break 30% payload fraction when 1.0 dropped and they changed the aero, and I did it with a massive SSTO capable of carrying oversized payloads).

Still, with those contstraints in mind, I'm interested in hearing any improvements you could suggest to this sort of SSTO:

Spoiler

LV-N burn to orbit while approaching Ap, with an oversized payload:

Just above the "top" of the atmosphere, taking another spaceplane piggyback as the payload:

Rendevous with a large interplanetary ship being constructed in orbit, adding a rotating habitat module (iirc, that was the 4th trip)

I think this was delivering the first module:

and landing back on kerbin:

an earlier model, again doing a piggy back lift for a smaller spaceplane intended for duna:

Showing the redocking procedure after deploying a payload:

Recovery on runway:

An earlier LOL shot, when I was willing to use part mods for surface bases:

But it was all delivered in one package:

lulz:

I managed to get 17% payload fraction in 3x (1.25x atmosphere), with increasing use of nukes increasing my fraction... but making it take fooorrrrrreeeeevvvvveeeerrrr to get to orbit.

That's why I switched to re-usable 2 stage designs... much faster time to orbit, and less time spent playing with a craft in the higher part counts.

[Northstar1989]

2 hours ago, KerikBalm said:

Well, I wasn't using a "standard" 3.2 config, but my own 3x config (I can do the math for rotation period changes), with a 1.25x atmosphere rescale (the stock atmosphere is proportionately way too thick).

I'm not interested in part mods, I wanted the part mods to be limited to just Stock+DLC (even if the solar system was modded). I don't play with FAR, because I think if I did far, then I'd want to ditch the lego piece wings and go with some sort of procedural or mod wing set, when I want to keep my craft stock+DLC.

The precoolers that you mentioned for instance... are probably OP'd. IRL Sabre engines would operate to mach 5.something. In KSP, rapiers do that. IRL, 1,600 m/s surface/1800 orbital  is a small fraction of orbital velocity (nearly 8,000 m/s), in stock its almost all the way there, in 3x (or 3.2x), it gets you roughly half way to orbit... IRL to get half way there, you'd need to go at least 3,500 m/s in the atmosphere... so I don't think better jet performance is realistic for a 3.2 or 3x scale.

I said 3.2x didn't I?

My bad.  I don't think I've even ever played at that scale- despite its recent popularity.  Memory fails me at times about gaming.

What I meant to say was RSS 64k: a 64% rescale of Earth.  At that point it's 6.5 or 7 km/s to orbit, and just as hard as RSS at 100% scale with the full-sized (upsized from Kerbal defaults: as they tend to be about 50-64% the size of their real-life equivalents.  The Realism Overhaul series of mods actually up-sizes engines to the sizes of the engines they are supposed to represent, if I'm not mistaken).

As you can see, getting to orbit in RSS 64k is quite a bit more challenging of an endeavor (roughly 2x the Delta-V requirements of 3.2x, in fact).

I don't exploit, or abuse part-clipping (although given the way REAL parts are built into each other, a little light-clipping is OK, if you ask me).

Pre-coolers might be OP for stock-alike systems, but they are necessary and *realistic* for larger RSS sizes.  The fact is, pre-coolers are a real thing (SABRE pre-coolers have actually been built and demonstrated on the ground, for instance) that just hasn't seen any use- but CAN be built.

Just because it's from KSP-Interstellar doesn't mean it's automatically overpowered: and most of the early to mid-game parts for the mod are 100% accurate to the performance we expect in real life (I know because I advised FreeThinker on the balance, and dug up published engineering studies to justify nerfing or buffing certain parts/fuels.  Most of the time we buffed, because the balance in the original mod was based on stock Kerbin, not balanced against something like RealFuels and RSS- and real life rockets are usually very OP'd compared to Stock KSP parts in terms of TWR and ISP...)

The nuclear thermal turbojets were a bit awesome, I'll admit.  But since I limited myself to rather low-tech reactors that are on the drawing boards today, in our real Earth, it was nothing that won't be feasible someday.  Remember, a true Horizontal Takeoff, Horizontal Landing Orbital Spaceplane has NEVER been built and flown on Earth...

It's nearly impossible in RSS 64K without FAR, a degree in aerospace engineering, and tons of mods to at least get the shape right, if you're limited to only today's engines (no RAPIER, no RAMJETS- since they've never been used on such large scale at such high speeds, at least definitely no Scramjets, amd a lot of other limits most KSP players would gag at...)  Since I'm only one man, and my background is in Biology, I think using nuclear thermal rockets that could be built in the next 60 years is probably excusable for RSS 64k...

I said nothing of my payload fraction: but I know it was in excess of 3%.  Which is all real life rockets achieve- and spaceplane fractions would likely be even lower...

Anyways, enough about what I accomplished: it was enough to have a few insights.  However, if you refuse to use FAR, what I can tell you is very limited and not guaranteed to work: the stock aero model is so bad...

What I can tell you, First of all, is to use Structural wing parts- not Delta Wings.  They're larger (have more per-part Lift) if I recall: though it's been a long time since I used anything but Procedural Wings (my CPU thanks me for the reduced part-counts, and FAR loves the better shapes.  Why don't you love them toooooo???) so I could be wrong- in which case, ignore the rest of this paragraph.  The stock aero model may or may not penalize you for this, as it doesn't realize you're building them all into a single swept mega-wing the way FAR does, but it will still give you more total lift for your part-count at least.  This will raise your Lift/Drag ratio for your spaceplane as a whole, even if it lowers it for the wings (because the wings will be larger relative to the fuselage).  That will help you get larger payloads to orbit: though you will have less Delta-V if you insist on hauling these huge wings all the way to Jool for some reason, and probably won't make it without refueling... (if you really must do that, and refuse to build specialized craft for atmospheric vs. space operations, or have the payload fly on its own transfer-stage, then I can't help you: you're crippling yourself in too many ways for no reason...)

Second, your spaceplane needs Canards.  Big, all-moving canards, ideally (the kind you can't build in Stock!)  But if you insist on staying stock, then build a long/thin wing put f5om the nose of your craft, and attach control surfaces to BOTH the leading and trailing edges- the largest surfaces you have access to.  It's not a large canard, but it's the best you can do in Stock...  No guarantee this will work in Stock aero- but canards are VITAL for spaceplanes in FAR...

Alternatively, instead of one large main wing, make a forward-swept front wing AND a back-swept rear wing.  Sweep both as strongly as possible: the objective is to create lifting surfaces far from the Center of Mass to get more Torque for a given force (long lever-arm).  Then attach control surfaces to the front of the forward wing (note you may have to invert the Pitch/Roll control authority with Stock tweakables.  Yaw should be disabled for any horizontal control surface anyways...) and rear of the rear wing, near the wing tips.  Set the control surfaces closest to the tips (should have many in a row) to Roll control ONLY, while those further in should be limited just to Pitch.

Raised Tailplanes.  Make yours longer- they can be made even further from Center of Mass than wingtips of ypur main wings if long enough, and also provide Keel Effect stability (having CoM BELOW Center of Lift in the vertical axis provides passive stability against Roll- reducing need to correct with control authority/thrust-vectoring or the game's unbelievably, unrealistically OP'd reaction wheels).  I'm pretty sure Stock aero doesn't really care much for biplane/ wing interference effects (see point below), so 2 wings in parallel make 2x as much Lift as 1.  If you're gonna use Stock aero you might as well work WITH it, not against it.

Consider using a supersonic bi/tri-plane design.  This doesn't work in FAR, because it doesn't model the effect, but in real life, setting 2-3 wings the correct distance apart can cause the shockwaves at the leading edge to destructively interfere with each other, REDUCING Drag in much the same way a bulbous bow below the waterline can actually reduce Drag on a ship (though, like a bulbous bow, it might only work for a narrow envelope of speeds and altitudes- requiring very careful piloting to take advantage of in real life...)  In Stock KSP it's just a bit more structurally stable- as you can strut the wings to each other near the ends to reduce wobble.  Should be combined with reduced wing-chord so that your overall wingspan and part-count remain the same.  NEVER do this in FAR though...

Observe your Trim during high your ascent yo orbit.  Are you usually fighting to keep the nose up, or pull it down?  Depending on which:

- If you're trimming to pull the nose up most of the flight, consider moving your Center of Mass back, or Center of Lift forward a bit and adding Canards- which generate POSITIVE Lift to pull the nose up.  Control Surfaces BEHIND the Center of Mass, on the other hand, generate NEGATIVE Lift: and this actively work against your ascent.

- If you're trimming the nose down, just the opposite: move CoM forward/ CoL back and favor rear control-surfaces over Canards

As nose-draggers are more stable, most players build over-dampened darts.  But, as proportional Canards for large planes are impossible to build in Stock, most players end up relying on Negative Lift to control pitch most of the way up: which is counterproductive and harmful in spaceplanes.  If you won't build the "Canards" I described before (ling/thin nose-mounted wings whose sole purpose is to attach as many control surfaces as possible), you should make your plane less stable and more tail-heavy to reduce Drag from trim, at least.

Use fewer air intakes.  Looking at your design, it has too many for the number of engines: generating excess Drag.  Consider swapping some of your shock intakes for Tail Extenders facing forward (the Tail Connector A has the lowest Drag of any 1.25 meter nosecone in the game when facing directly into the airstream) or at least some type of long/thin nosecone.  The remaining intakes should provide enough air for your RAPIER's up until the point you should be switching to rocket-mode.  Consider using reduced thrust on takeoff if you get flameouts on the runway (and gradually upping your throttle as you build speed: you shouldn't need max throttle to get off the runway if you have as much max Thrust and wing area as you ought to...)

Consider swapping to a lighter cockpit. Those Mk3 cockpits are a sin against efficiency if you're not shuttling Kerbals to orbit for some reason (or removing them from orbit in rescue contracts).  Use a Mk1 cockpit, THEN transition to a Mk2 Fuselage (and a Mk3 as necessary if you need the fuel capacity and don't want to spam a longer line of Mk2 parts, pushing up part-count) to save weight.

 

Consider some of the advice here.  Not all of it is well-explained, some is incomplete thoughts, but there's a good reason for all of it...

[KerikBalm]

3 hours ago, Northstar1989 said:

I said 3.2x didn't I?

My bad.  I don't think I've even ever played at that scale- despite its recent popularity.  Memory fails me at times about gaming.

What I meant to say was RSS 64k: 

...

As you can see, getting to orbit in RSS 64k is quite a bit more challenging of an endeavor (roughly 2x the Delta-V requirements of 3.2x, in fact).

No, not in fact. Keeping surface gravity the same but increasing diameter increases orbital velocity (and thus dV, ignoring air resistance) by a factor of the square root of the rescale.

Scaling kerbin up 10x results in √ 10 times the orbital velocity... 3.16x as much. Scaling it up 6.4 x increases it √ 6.4, which is 2.5x the orbital velocity of stock. 6.4x is 2.13x the size of 3x, and 2x the size of 3.2x so, √ 2. = 1.41x the orbital velocity. The total dV to orbit changes depending on a number of factors, air resistance is a big factor that increases it, because you lose a lot more to drag when your at an equivalent pressure to 30km altitude in stock, and going at 4km/sec instead of 2 km/sec.... for example.

A rocket doing a gravity turn is less affected by this than a spaceplane that will be accelerating more deeper in the atmosphere.

 

Quote

I don't exploit, or abuse part-clipping (although given the way REAL parts are built into each other, a little light-clipping is OK, if you ask me).

Yes, light clipping is fine with me... light being the key word

Quote

Pre-coolers might be OP for stock-alike systems, but they are necessary and *realistic* for larger RSS sizes.  The fact is, pre-coolers are a real thing (SABRE pre-coolers have actually been built and demonstrated on the ground, for instance) that just hasn't seen any use- but CAN be built.

But rapiers perform as if they already have precoolers, there's no other way they'd work to mach 5+

Quote

What I can tell you, First of all, is to use Structural wing parts- not Delta Wings.  They're larger (have more per-part Lift) if I recall: though it's been a long time since I used anything but Procedural Wings (my CPU thanks me for the reduced part-counts, and FAR loves the better shapes.  Why don't you love them toooooo???) so I could be wrong- in which case, ignore the rest of this paragraph. 

You are wrong. Those delta wings are the Big-S delta wings. There is only one stock wing that has more lift per part, and that is the FAT-455 wing. The FAT-455 wing has a much lower heat tolerance, and is not suitable for orbital use.

Thats why those SSTOs I showed used the large delta wings, but the 2 stage recoverable designs uses mainly the FAT-455 wings.

I would like procedural wings to reduce part counts, but I also like to see what others build, and compare what I'm able to do to others, so I use stock parts as a reference.

An all stock design is sharable and comparable with no mods.

So in addition to bigger engines, I'd also like bigger wings.

[Pds314]

Yeah.... 8-12? more like 50... this design for a space bus is from awhile ago but it's very representative of the issue. I believe it was made in oldFAR, but that's actually a lot more similar to current stock than it is to oldstock.

Edited May 14, 2020 by Pds314

[Northstar1989]

6 hours ago, KerikBalm said:

No, not in fact. Keeping surface gravity the same but increasing diameter increases orbital velocity (and thus dV, ignoring air resistance) by a factor of the square root of the rescale.

Scaling kerbin up 10x results in √ 10 times the orbital velocity... 3.16x as much. Scaling it up 6.4 x increases it √ 6.4, which is 2.5x the orbital velocity of stock. 6.4x is 2.13x the size of 3x, and 2x the size of 3.2x so, √ 2. = 1.41x the orbital velocity. The total dV to orbit changes depending on a number of factors, air resistance is a big factor that increases it, because you lose a lot more to drag when your at an equivalent pressure to 30km altitude in stock, and going at 4km/sec instead of 2 km/sec.... for example.

The numbers I quoted are for rocket ascents.

It takes 4 - 3.5 km/s vacuum Delta-V to reach orbit in stock-sized Kerbin.  Most players underestimate this because they forget most engines get less Thrust/ISP in atmosphere.  I also use SpaceX-style launch stage recovery (including in RSS 64k), which is more comparable to spaceplane ascents in terms of cost, but requires more Delra-V...

It definitely takes 6.5 - 7 km/s in RSS 64k: because aero-drag doesn't decrease (in fact increases, like you said), but goes from maybe 1-2 km/s of the ascent costs (with 2 - 2.5 km/s for grwvity-losses) to 1.5 - 2.5 km/s with gravity-losses being 2.5x as much (sqrt 6.4 is 2.5) around 5 - 6 km/s.  The numbers I gave before add up...

 

[KerikBalm]

I can get to orbit in stock with 3.2 km/s... 3.5 to 4 .... Especially 4, is a large overestimate

Also, you said: " with gravity-losses being 2.5x as much (sqrt 6.4 is 2.5)"... but that's 2.5 as much as 1x, you had previously said: "roughly 2x the Delta-V requirements of 3.2x"

Now, aero losses will go up more because of the V^2 relationship of drag. take a 4x system, your orbital velocity increases by 2x. If your rocket is designed to have the same drag at the same speed for a given air density (and Isp and TWR increase proportionately), then the force of drag goes up by 2^2.

So my intution is that 3x has roughly 3x the aero losses, 6.4x has roughly 6.4x the aero losses (but going to FAR will change that relative to stock). Orbital velocity increases by root(3) or root(6.4)... thus the total change in velocity needed (which is dV, after all) goes up by that much, but you have more losses.

On top of that, if you aren't playing with some mods that increase part performance (most stock KSP parts are underpowered), you will probably end up packing on more fuel, and having a lower TWR (more gravity losses)... and if you try to get around that, due to limited thrust to cross section ratios (although vectors are by far the best), your cross section and thus drag goes up.

Anyway, its complicated, and it depends a lot on your flight profile.

Its even worse for spaceplanes when considering the *closed cycle* dV needed. In 1x, suppose you get to 1,800 m/s orbital on airbreathers... you're now about 550 m/s away from orbital.

In 3x, you'll have a bit more ground speed, (but also a heavier plane that won't get as high), lets be generous and say that you reach 2,000 m/s orbital... thats still 2100 m/s shy of orbit... a roughly 4x increase in closed-cycle dV. And that dV is a lot harder to come by at 305-340 Isp, than from airbreathers at 3200-4000 Isp.

Being so much farther below orbital velocity, you now also need much more lift to maintain altitude, and much more lift for all that extra closed cycle fuel you need to carry, this means a lot more drag at the same speed, and then you'll need to accelerate faster to get up higher.

That's why I switched to high TWR airbreathers that do a 30 degree or so zoom-climb/parabolic trajectory after hitting their power band... climb up and out of the soup, go closed cycle and keep pushing, get the Ap up, and then release the 2nd stage. 2nd stage gets into a stable orbit, and one can switch back to the 1st stage before it falls back into the atmosphere.

Fully reusable with a better payload fraction than an SSTO, much faster time to orbit, and much higher TWR - no LV-Ns.

Of course, the disadvantage is that you need to recover not one, but 2 spaceplanes... and the 1st stage spaceplane needs to do a 180 back in the atmosphere, and fly back to base (with its much lower wing loading at this point, and its already high TWR, it cruises back at over 1700 m/s surface

 

 

Edited May 14, 2020 by KerikBalm

[Northstar1989]

16 hours ago, KerikBalm said:

That's why I switched to high TWR airbreathers that do a 30 degree or so zoom-climb/parabolic trajectory after hitting their power band... climb up and out of the soup, go closed cycle and keep pushing, get the Ap up, and then release the 2nd stage. 2nd stage gets into a stable orbit, and one can switch back to the 1st stage before it falls back into the atmosphere.

Fully reusable with a better payload fraction than an SSTO, much faster time to orbit, and much higher TWR - no LV-Ns.

Of course, the disadvantage is that you need to recover not one, but 2 spaceplanes... and the 1st stage spaceplane needs to do a 180 back in the atmosphere, and fly back to base (with its much lower wing loading at this point, and its already high TWR, it cruises back at over 1700 m/s surface

Drag goes up less than you think, because in an optimal ascent you spend less time in the lower atmosphere- even in a rocket (of course it's better to build up a lot of vertical velocity early, and then thrust horizontally later, than it is to hold a rocket above the atmosphere with Thrust...)  So for 64k, drag-losses go up a lot less than 6.4x (maybe 3x), and gravity-losses at least 3.5x (you trade lower aero drag for higher gravity-losses: but the trade is unequal)

The bottom line, though, is that 64k spaceplanes are a LOT harder than 3.2x ones: and that's all I'll say further on that...

As for spaceplane ascent-profiles, there's a BIG gap between "in the soup", and altitudes where airbreathers no longer work.  An optimal ascent will climb into the upper atmosphere, mostly level off for speed (and a bit mire altitude), and only THEN go closed-cycle.

And contrary to your assertions, realistic SABRE's with precoolers are predicted to continue working until about Mach 6 (Mach 5 is only where it becomes no longer useful to remain open-cycle with the chosen ascent profile for SKYLON- you don't want to push the open-cycle right until their limits, Drag and shock-heating becomes prohibitive before then: past a certain point it's better to just use rockets and climb higher...)

Similarly, my spaceplane engines in 64k could operate to about Mach 6.4 with precoolers: but went closed-cycle before then.  And it was a nuclear thermal hybrid (dual-mode, could switch to rocket mode like a SABRE) turbojet.  A fundamentally different propulsion process than SABRE- one where no actual combustion occurs, and all heat cones from the reactor (and thus just slightly more capable than a SABRE- though it only barely surpassed one when you accounted for the huge reactor weight, and enormous fuel tanks needed for Liquid Hydrogen- no LOX tanks like with SABRE.  Its thrust was actually quite a bit lower than a SABRE's in closed-cycle, even with an initial "kicker" of Liquid Nitrogen used early in the closed-cycle part of the ascent.  But ISP was higher: more similar to a stock NERVA...)

2-stage spaceplanes have always been an optimization of the basic spaceplane ascent profile: indeed without knowing it, your piggyback design probably re-created certain historical designs for 2-stage piggyback spaceplanes on Earth (prior to SABRE, it's the only way we could have built a useful spaceplane AT ALL, if desired)  But within that design, there are still a number of optimizations I'd like to point you towards...

[Incarnation of Chaos]

This thread went completely to crap rather quickly without my involvement; UNACCEPTABLE!

Jokes aside; i'd like bigger jets and bigger intakes especially. And some tweaks to the amount of inlet air they produce, as it becomes the largest issue with larger designs tbh. And i would like to see them as a update, not DLC.

But stock 1.875, 2.5 and even 3.5m jets of all varieties need to be represented in KSP, and we should have pure ramjets, scramjets and the like also. But that's just because i enjoy building planes occasionally, Two-Stage reusable rockets are my workhorses whenever i actually want to get a mission done.

[KerikBalm]

Well, yea, I think the discussion with @Northstar1989 is rather unproductive, we play at different scales, with different part mods, in different atmosphere models.

We agree that larger jet engines can reduce part count. Larger or procedural wings would also decrease part count, and I would love them as well.

[Incarnation of Chaos]

2 hours ago, KerikBalm said:

Well, yea, I think the discussion with @Northstar1989 is rather unproductive, we play at different scales, with different part mods, in different atmosphere models.

We agree that larger jet engines can reduce part count. Larger or procedural wings would also decrease part count, and I would love them as well.

Stock procedural wings would be lovely; especially with all the issues iv'e had getting them semi-working in 1.8.1.  "Semi-Working" because for some reason FAR isn't patching them correctly now, and i'm not getting anywhere close to the lift i should (Stock wings are being successfully patched though, so it's likely a issue with my install and not FAR itself.....). Which if we had stock procedural wings wouldn't even be an issue.

[Northstar1989]

On 5/15/2020 at 3:35 AM, KerikBalm said:

Well, yea, I think the discussion with @Northstar1989 is rather unproductive, we play at different scales, with different part mods, in different atmosphere models.

We agree that larger jet engines can reduce part count. Larger or procedural wings would also decrease part count, and I would love them as well.

Was just trying to give you some pointers for your spaceplane bro.

Then I got sidetracked.

[Northstar1989]

Anyhow...

Larger jet engines.  Stock.  Are there maybe some modders who could draw some up, and then offer to work with SQUAD to make them Stock? (After this whole ESA update is done)

[KerikBalm]

On 5/19/2020 at 1:30 PM, Northstar1989 said:

Was just trying to give you some pointers for your spaceplane bro.

Then I got sidetracked.

Yea... But its pretty much useless advice.. 

So... A whole bunch of stuff about arranging surfaces for control and trim, when my designs are balanced just fine (really, what is with the tip about disabling yaw on horizontal surfaces, I already do that), something about biplanes that 1) doesn't work in stock, 2) doesn't work in FAR, 3) I don't think works in RL because no proposed designs use it (are you trying to refer to the design of the xb-70 wings)...

An insistence on using non stock canards when my sstos already use large canards, incorrect info about stock wings, etc.

And some advice which is fine and I was already aware of, but don't use for aesthetic reasons

Edited May 23, 2020 by KerikBalm

[Northstar1989]

11 hours ago, KerikBalm said:

A whole bunch of stuff about arranging surfaces for control and trim, when my designs are balanced just fine

It's not about stability.

Optimizing trim is about reducing Drag from control authority.

If you rely on rear-mounted control surfaces to push the nose up in an nosedragger, you are generating NEGATIVE Lift you have to cancel out with even more wing area.

Not only does this add to mass, it adds to Drag significantly.

A lot of my advice was useful for Stock KSP, and actually HARDER to implement in FAR.  Suoersonic Biplanes work in Stock KSP, for instance.  The Drag isn't reduced like in real-life, but the wings both generate much more Lift than they do in a real-life biplanes or FAR: so it is possible to use dual wings to improve the Lift/Drag ratio of a plane as a whole in Stock KSP.

Edited May 23, 2020 by Northstar1989

[KerikBalm]

You realize that all the wings and cntrol surfaces on those designs have incidence?

There's no negative lift from the rear surfaces until deflection angle exceeds the angle of incidence.

Even positive lift in the front can be bad if its achieved through a high AoA, which is very draggy

[Northstar1989]

On 5/24/2020 at 3:29 AM, KerikBalm said:

You realize that all the wings and cntrol surfaces on those designs have incidence?

There's no negative lift from the rear surfaces until deflection angle exceeds the angle of incidence.

Built-in Angle of Attack, or are you just talking incodence from nosing up during flight?

Negative lift is *very easily* generated at pretty modest levels of control input.

If your Angle of Attack is 5-7 degrees (about what it should be in FAR for a supersonic spaceplane, at least) and your pitch control input is 50% upwards, with default Control Authority (20 degrees in KSP!!), you'll get *negative* 3-5 degrees angle of attack for control surfaces mounted behind the Center of Mass at 100% Pitch Control Authority (you SHOULD be creating separate surfaces for Yaw, Pitch, and Roll- and disabling the other 2 axes for each...)

What's more, even if the rear control surfaces are at completely neutral (0 degree) incidence, such as 50% pitch-up input at 10 degree angle of attack, the control surface is still bringing down your Lift:Drag ratio- as it is still generating Drag for MUCH less than optimal (5-7 degrees FAR, more like 7-12 degrees Stock) Angle of Attack...

By contrast, if you have LARGE canards (I don't think your canards were what I'd consider "large"- also keep in mind that on the stock winglets only a SMALL PORTION of the lifting surface actually acts as a control surface and deflects- the rest just provides Lift like a normal wing with a much smaller control surface attached...) and are flying at 5 degrees AoA at 50% pitch-up, with control authority limited to just 10 degrees (what I would suggest- sometimes even less) then your canards are only at 10 degrees incidence- which has a better Lift/Drag ratio than the same control surface at 0 degree incidence.

What's more, on many designs with rear-mounted engines the CoM is well BEHIND the midpoint of the plane- which means nose mounted Canards have a much longer lever-arm with the CoM than do rear-mounted control surfaces.  This means less force is required to generate the same torque and turn the plane.  For instance, to maintain level flight at a certain altitude you might only need 6 square meters of frontal control surface deflected at 5 degrees from the rest of the wing (10 degree incidence), generating 2 kN of extra Drag, with Canards 20 meters from your CoM; whereas you might need 10 square meters of rear-mounted control surface, deflected at 6 degrees (negative 1 degree incidence) generating 2 kN of Drag directly and forcing the rest of the wing to generate an additional 2 kN of Drag (through either increased AoA, or larger wing-area of the design) to make up for the loss of Lift from the Elevators at this deflection (whereas the Canards generate MORE Lift when they deflect the nose up).

These differences are *particularly* pronounced in KSP with spaceplanes (vs. regular planes), because spaceplanes end up with higher wing-loading and greater demand for Lift due to all the rockets/fuel they need to get to orbit, their payload, the poor mass-ratios of fuel tanks in KSP, and the need for spaceplanes to cruise at higher altitudes and speeds (with jets optimized accordingly, and often producing reduced Thrust at high speed/altitude- necessitating use of more/larger jet engines...) than normal planes.

KSP jet engines also have all their mass in the nozzles- unlike REAL jets, which are embedded in the fuselage: meaning the Center of Mass tends to end up further back in KSP than in real life, and rear-mounted control surfaces become even LESS effective (and act more similarly to Spoilers and less like Elevators) as a result of their shorter lever-arms.

To get back to the *topic* of this thread, though: the higher wing-loading and reduced effectiveness of Elevators (due to CoM too far back, thanks to unrealistic engine Center of Mass- which SHOULD be offset forward of the nozzle) in KSP spaceplanes mean that you need to fight against even more Drag, and thus require larger jet engines than Stock KSP currently provides.  Hence one more reason why we need larger (1.875, 2.5, and maybe even 3.75 meter) jet engines in KSP...

 

P.S. Send me a PM with an image.  I would love to see a close-up of your Canards and their Actuation/Control settings (you need to enable Advanced Tweakables to see/change these in the editor.  I thin this is a Difficulty or Gameplay setting?  I always have them on, so I forget...)  Many players in KSP under-size their Canards, either because they don't understand the value in having larger Canards with reduced Control Authority (less chance of Stall, and less extra Drag for a given torque generated), or fail to realize that the majority of the area of the stock winglets is NOT actually control surface at all, but just normal wing-area (and thus, functionally, they have tiny Canards that have to deflect at very high incidence to provide the needed Torque...)

Edited May 26, 2020 by Northstar1989

[KerikBalm]

1 hour ago, Northstar1989 said:

Built-in Angle of Attack, or are you just talking incodence from nosing up during flight?

I said incidence, I meant incidence. Nosing up in flight doesn't give incidence.

Quote

By contrast, if you have LARGE canards (I don't think your canards were what I'd consider "large"- also keep in mind that on the stock winglets only a SMALL PORTION of the lifting surface actually acts as a control surface and deflects- the rest just provides Lift like a normal wing with a much smaller control surface attached...) and are flying at 5 degrees AoA at 50% pitch-up, with control authority limited to just 10 degrees (what I would suggest- sometimes even less) then your canards are only at 10 degrees incidence- which has a better Lift/Drag ratio than the same control surface at 0 degree incidence.

...

KSP jet engines also have all their mass in the nozzles- unlike REAL jets, 

...

due to CoM too far back, thanks to unrealistic engine Center of Mass- which SHOULD be offset forward of the nozzle)

...

fail to realize that the majority of the area of the stock winglets is NOT actually control surface at all, but just normal wing-area (and thus, functionally, they have tiny Canards that have to deflect at very high incidence to provide the needed Torque...)

#1) as I said multiple times, I am using stock parts, no larger canards for control are possible. In the ssto case, and the 2 stage case, the com is where the payload is, and hence leverage is about the same.

#2) only the small delta winglets, the shuttle style tail, and whatever that 1200 heat tolerance one is that goes with the fat-455 wings act like that. The fins with canards in the name are all moving planes

#3) you have to look at total l/d, not part l/d. A part with 0 AoA will have no lift, but still some drag. It won't affect craft L/D much in such a configuration. A canard at high AoA on the verge of stall will generate lift, but also a lot of drag. If the part l/d is worse than total l/d, it brings the total l/d down. In some cases, its better to gave a little drag with no luft than some lift with a lot of drag

#4) jets in ksp do have their com offset forward, the effect is less and less useful as planes get bigger.

You could always put them in a tube, and mount them in the middle (as in many RL designs), or like skylon and airliners, and mount them close to center.

I use fore and aft mounted engines to keep com from shifting too much as fuel burns off

 

[Northstar1989]

1 hour ago, KerikBalm said:

I said incidence, I meant incidence. Nosing up in flight doesn't give incidence.

Then you're referring to built-in Angle of attack (what "Angle of Incidence" means in aviation).

Has the same effects as flying at higher Angle of Attack, with regards to the discussion we were having.  The important differences have to do with wing position on the ground (landing), fuselage drag/lift in level flight (should NOT be done with Lifting Body fuselages, like the Mk2 fuselage parts), and the angle of the engines relative to the horizon (better to angle the engine nozzles themselves if you have a lifting body, better to angle the fuselage and engines together if not).

None of it changes the Trim Drag you experience from Elevators vs. Canards (less for Canards, both in-game and real life).

[Northstar1989]

15 hours ago, KerikBalm said:

#1) as I said multiple times, I am using stock parts, no larger canards for control are possible. In the ssto case, and the 2 stage case, the com is where the payload is, and hence leverage is about the same.

It doesn't matter where the payload is.  This is about the relative length of the lever-arms for the Canards vs. the Elevators.

If the lever-arms are about the same, the difference is smaller.  But Canards are still better, as they don't reduce the Lift of the plane as a whole or induce Negative Lift when they push the nose up.

(Note: in a TAIL-DRAGGER design, where you have to push the nose DOWN, Elevators are better and Canards are worse- but these designs are inherently unstable, and so rarely used...  With Autopilot Module Manager mod, which is a utility mod- no parts- that introduces basic Fly-By-Wire and stabilizing algorithms, however, it is possible to keep designs that are slightly unstable under control- and they tend to require less control input to maneuver, making them actually slightly better-performing.  Military fighter jets in real life are sometimes tail-draggers for this reason...)

15 hours ago, KerikBalm said:

2) only the small delta winglets, the shuttle style tail, and whatever that 1200 heat tolerance one is that goes with the fat-455 wings act like that. The fins with canards in the name are all moving planes

That they are- but I didn't know if maybe you were using one of the other parts.

You CAN create your own, much larger Canards by attaching the control surfaces (normally used as Elevators in Stock) to the front of Swept/Structural/Delta Wings though.  3 or 4 of these on the leading-edge of a frontal winglet of your own design will get you more control area than any of the Canard/fin parts, which are tiny (letting you reduce Control Authority, so each surface deflects less to give you needed control...)

15 hours ago, KerikBalm said:

#3) you have to look at total l/d, not part l/d. A part with 0 AoA will have no lift, but still some drag. It won't affect craft L/D much in such a configuration.

This is exactly what I just told you.  Re-read what I wrote.

Elevators, in neutral or very low deflection, generate little/no Lift but still generate Drag- lowering the Lift/Drag ratio ("l/d") of the vessel as a whole.  This is why Canards (which generate EXTRA lift to push the nose up) are better.

It WILL harm overall L/D a little- all of these differences are small, but add up (less efficient aerodynamics means you need more Thrust, more fuel, and then more wing area to support those- causing more Drag, adding mass, meaning more fuel still to reach orbit...)

15 hours ago, KerikBalm said:

A canard at high AoA on the verge of stall will generate lift, but also a lot of drag. If the part l/d is worse than total l/d, it brings the total l/d down. In some cases, its better to gave a little drag with no luft than some lift with a lot of drag

You have to remember that Elevators generating no Lift are basically deadweight you have to push to orbit.  You might as well have replaced some of them with wing area with negative incidence, that is at neutral AoA in level flight, or just moved the Center of Lift forward, or Center of Mass back...

The whole "edge of stall" case is *precisely* why you include LARGE amounts of Canard surface area, so each of them is nowhere near stall (only at a few degrees higher AoA than the rest of the wing).

The Lift/Drag curve looks like an upside-down "U"- and your spaceplane should at slightly *below* the peak (having excess wing area like this increases Drag, but has benefits like reducing takeoff/landing velocity and improving cross-range by increasing the cruise altitude and V*L/D, even as L/D falls).  So, having Canards deflect just a few degrees will bring them to peak L/D, and having them deflect a little more won't cause L/D to fall by much (while significantly increasing Lift).  The key is simply not too place too much burden on the Canards relative to their area- they should be able to maintain control with maybe 6 or 7 degrees MAX deflection (so most of the time they only deflect 2-3 degrees).

To achieve this in Stock, you need to build large "custom Canards" using the method I outlined.  Even then, Canards should SUPPLEMENT control authority from your tailplane- allowing you to have less Control Authority there as well.

If you use default deflection, your Elevators *ARE* generating Negative Lift, and this IS harming your performance.  Canards help avoid this- by letting you turn down max Control Authority on your tailplane.

15 hours ago, KerikBalm said:

#4) jets in ksp do have their com offset forward, the effect is less and less useful as planes get bigger.

You could always put them in a tube, and mount them in the middle (as in many RL designs), or like skylon and airliners, and mount them close to center.

I use fore and aft mounted engines to keep com from shifting too much as fuel burns off

The default CoM offset is very slight- the CoM still lies within the nozzle.

In a REAL jet, the nozzle is very light.  The weight is found in the Compressors, Combustion Chamber, and Turbomachinery- which are located much further forward in the aircraft.

For KSP to be realistic, you would have to increase CoM offset to the point that jet engine CoM laid several meters in front of the nozzle part you place, at least.

Tubes add drag.  Fore-mounted engines require mountings- and also add Drag.

Anyways, we're off-topic.  Once again, PM me about your plane, [snip]

Edited May 27, 2020 by Gargamel
Portions Redacted by Moderator

[KerikBalm]

15 hours ago, Northstar1989 said:

Then you're referring to built-in Angle of attack (what "Angle of Incidence" means in aviation).

Yes, I was referring to incidence, I don't know why you want to call it built in angle of attack, when you can have positive incidence and negative angle of attack, and your statement about getting incidence by pitching up is just rubbish.

Quote

None of it changes the Trim Drag you experience from Elevators vs. Canards (less for Canards, both in-game and real life).

Dude, there is no inherent difference in trim drag from elevators vs canards. if you trim an elevator down, it makes more lift behind the CoM, if you trim a canard up, it makes more lift in front of the CoM. If a plane is balanced so that you need to pitch down as much as you need to pitch up, then it doesn't matter.

14 hours ago, Northstar1989 said:

It doesn't matter where the payload is.  This is about the relative length of the lever-arms for the Canards vs. the Elevators.

And the payload in those designs is at the CoM, and thus if the CoM is at the approximate center, then the lever arms are approximately equal.

Quote

If the lever-arms are about the same, the difference is smaller.  But Canards are still better, as they don't reduce the Lift of the plane as a whole or induce Negative Lift when they push the nose up.

They reduce lift when you need to pitch down.

If an elevator is constantly making lift because of incidence then it doesn't reduce lift of the whole plane for small deflections.

Quote

This is exactly what I just told you.  Re-read what I wrote.

Elevators, in neutral or very low deflection, generate little/no Lift but still generate Drag- lowering the Lift/Drag ratio ("l/d") of the vessel as a whole.  This is why Canards (which generate EXTRA lift to push the nose up) are better.

And canards in neutral or very low defelction generate little/no lift but still generate drag. That canard can generate extra lift to push the nose up is exactly equal to an elevator that is generating extra lift to lift the tail up... 

This is not complicated.

Quote

You have to remember that Elevators generating no Lift are basically deadweight you have to push to orbit. 

You have to remember that Canards generating no lift are basically deadweight you have to push to orbit.

Quote

The whole "edge of stall" case is *precisely* why you include LARGE amounts of Canard surface area, so each of them is nowhere near stall (only at a few degrees higher AoA than the rest of the wing).

And the same holds true for elevators with incidence, so the amount of deflection isn't that great (since to avoid stall, you also need deflection that isn't that great.

Quote

The key is simply not too place too much burden on the Canards relative to their area- they should be able to maintain control with maybe 6 or 7 degrees MAX deflection (so most of the time they only deflect 2-3 degrees).

...

If you use default deflection, your Elevators *ARE* generating Negative Lift, and this IS harming your performance.  Canards help avoid this- by letting you turn down max Control Authority on your tailplane.

If you use incidence, your elevators are generating positive lift. The key is simply not to place too much burden on the elevators relative to their area . they should be able to maintain control with maybe 6 or degrees MAX deflection (so most of the time they only deflect 2-3 degrees)

Seriously, your arguments are so easy to defeat by simple word substitution... [snip]

Quote

The default CoM offset is very slight- the CoM still lies within the nozzle.

It most definitely lies outside the nozzle

[snip]

 

Edited May 27, 2020 by Gargamel
Portions Redacted by Moderator