RAPIERs don't suck!: A complete performance evaluation

[Aragosnat]

To go off on that point about how RAPIERs can hold onto their intake air longer, I'd like to also point out that since RAPIERs are both hybrid engines, and they don't have any delay between modes, there is pretty much zero chance of your craft going into a flatspin due to the unpredictable nature of engine flameouts.

Depends on the craft as well as the persons pc. As I have gotten flatspins just using two raipers consitantly too. But, if I close the intakes while both are still on jet mode. Then they both switch over with no flatspins.

[numerobis]

The rapier can get into a flat spin just fine -- on a 2-engine plane, from when you have full air to 2/3rds air (which is a climb of about 2km), both engines operate but one gets decreasingly much thrust, so you yaw; let's say it's the right engine that loses thrust, so you yaw right. From 2/3rds air to 1/6 air (a climb of 7km) you have a rocket engine on one side and a jet on the other; the rocket gets fixed thrust, whereas the jet thrust is decreasing both from your speeding up and from the loss of air, so you now yaw left.

So you have to fight a yaw moment for 9km of climb before both engines switch to rockets, unless you switch over manually.

When the rapier switches, you're at 1/3rd thrust at, say, 1500 m/s. That means your rapier engine is developing all of 50 kN in air breathing mode, and you're still climbing. Then you switch to a whopping 175 kN of rocket thrust! If you instead switched to a turbojet and a pair of 48-7S your total mass would be 1.4t of engines, saving 350 kg, and your thrust would be steady. You'd burn a bit of that in extra fuel from the 10s difference in Isp, but not much. So, stats-wise, compared to the turbojet + other rockets, the rapier is not great. Compared to the payload fraction you can get with pure rockets, of course, it utterly kicks ass (but not as much as the turbojet).

I think both jets are drastically OP: either can get you to orbital velocity or very close to it. By comparison, the SABRE air-breathing stage gets up to 2km/s or so, leaving the rocket to develop another 5km/s or so to reach LEO. I think a good setting for the KSP turbojet and rapier would be a top speed around 1km/s or 1.5km/s, so you get half your speed off jets and half off rockets.

[Captain Sierra]

-snip-

Yeah, I did not bother to take into account orbital maneuvering. For my SSTOs which require that, I usually strap on a pair of more efficient radial engines (24-77s work good) strictly for orbital maneuvering. The higher TWR of the RAPIERs is reserved primarily for the ascent.

In other news, I've accomplished some of the testing I wanted to do but have more flights to run before I have more results to add.

[Claw]

Also, if people only ever used the absolute best thing every time, then everyone's designs would look the same. Whether you call it the LV-N, Aerospike, whatever, space planes and rockets would get a little boring if there was only ever one obvious choice.

If the RAPIER was a TurboJet combined with a aerospike (or insert your favorite engine) and weighed less than both, then we would probably see complaints that they were overpowered. If it weighed the same, then it would be pointless.

I've seen them flame out before TurboJets, and I've seen them flame out after. I've used them in combination with TurboJets as a TWR aid during climb, but shut them down to divert air to the TurboJets, and used the RAPIERS for rockets when I couldn't support the TurboJets anymore. I think there are uses for it beyond being the ultimate power in the galaxy.

Although I agree that it probably also needs some adjustment. I think it's unfortunate that you take an air breather thrust hit and a rocket ISP hit and it doesn't seem to have a corresponding reduction in weight. To me, that's what makes the problem because in both cases, you're pushing more weight with less benefit. That's also a hard thing to balance without making it overpowered.

[regex]

I'm pretty new to spaceplanes, mainly because I don't see a real use for them beyond being a "toy". When the RAPIER engine came out I thought I'd give them a try and eventually settled on a fairly useful little crew shuttle that I could use to bring Kerbals up or down from LKO, usually from single-man ships that I wanted to save weight on (otherwise I use a proper capsule to bring a bunch down). It reliably gets to orbit using a single RAPIER, can land pretty much anywhere, and has enough delta-V to rendezvous with most anything in LKO that isn't in a terrible orbit.

Well, after reading this thread, I decided to give the other engine combinations a try and see what was up. I used the exact same plane sans engine setup. I'm not a dirty stinking cheater so I don't alt-f12 part-clip, which also means my plane could probably be optimized even further and potentially look even cooler. Aside from the Engineer unit, the plane is bone-stock.

My process for all planes:

65 degree climb to 12km, screenshot at lowest climb speed, level off at 13km, screenshot at 14km. Maintain consistent 15 degree AOA until 1km/s, screenshot, then switch to a 10 degree AOA while building speed. Screenshot at 1.5km/s. Screenshot at flameout, switch over to closed cycle mode/rocket engines, pitch to 45 degrees, push apoapsis to 100km and coast. Once out of atmosphere, push apoapsis to 100km again, coast, and circularize. Screenshot at 100km x 100km orbit.

This may not be the best ascent profile but it worked pretty darn well for me in all cases. Here is the gallery, in order of planes and flights:

Javascript is disabled. View full album

My conclusion is that the RAPIER is an excellent engine to introduce new players to the world of spaceplanes (as has been said) but is otherwise fairly sub-standard for these small planes. OTOH, it does the job quite well and reduces part count (not like that's a big consideration here; all versions of the Night Heron were under 40 parts). I think good compromise would be to increase its space ISP while lowering the top speed for the turbofan by a bit (maybe 100m/s or something). At flame-out both of my turbofan planes were well over 2.3km/s while the RAPIER barely managed 2km/s.

After running this test, I'd have to say I favor the 909 setup. The TWR for both the turbofan and rockets is great, and the plane looks far better than the 48-7s version with its engines stuck on adapters. Granted, the 48-7s version is a bit better on fuel and makes for a more precise rendezvous, but I don't favor those engines because I think they need a serious nerf.

Edited February 1, 2014 by regex

[Mobjack]

You guys are doing this backwards. You shouldn't build your ships around your engine choices, but build your engines around your ship design.

The main advanage with a RAPIER is that you can take a Kerbal into space using one engine. An Aerospike Turbojet combo requires a 3 engine setup which will end up using more fuel. Probably the best way to do this is with a Turbojet and two 48-7Ss which are both overpowered in my opinion.

For bigger ships you can get creative too. Instead of having 2 turbojets and a Aerospike, try 2 Turbojets and a RAPIER. Or two RAPIERS and an Turbojet.

[Rhomphaia]

Yeah, I did not bother to take into account orbital maneuvering. For my SSTOs which require that, I usually strap on a pair of more efficient radial engines (24-77s work good) strictly for orbital maneuvering. The higher TWR of the RAPIERs is reserved primarily for the ascent.

In other news, I've accomplished some of the testing I wanted to do but have more flights to run before I have more results to add.

I'm not getting this.

24-77s (and in-fact all radial engines) are less efficient in a vacuum than the RAPIER. The advantage they offer in conjunction with turbojets comes from the lower mass, they don't offer anything in conjuction with a RAPIER . If high TWR is an issue for fine maneuvers you would be better off setting the thrust limit on the RAPIER once in orbit.

[Dispatcher]

You guys are doing this backwards. You shouldn't build your ships around your engine choices, but build your engines around your ship design. ...

I think this is one reason why the assembly buildings require you to start with certain parts (functionally, this means control pods; and yes, the structure tree has its hierarchy requirements). Its all about the payload, and where you want to take it. And for that matter, form factor is another (I like to build landers which are wider than they are tall, because its easier for me to land the things and they tend to tip over less, even on steep inclines). This does dictate the engines used to some degree. And ultimately it boils down to dV (especially when a target lacks an atmosphere suitable for lifting surfaces, oxidizer [or drag for aerobraking]). That said, people come up with a variety of ways to achieve their goals, and some of those methods can be unexpected. Besides, a lot of the Challenges pretty much force contestants to choose engine over payload and design. Now if only the aerospike and the jets had attachment nodes on the bottom...

[Mobjack]

This is what RAPIERs are good for:

Try building something comparable with an Aerospike/Turbojet combo. Just testing them out with an MK1 cockpit won't give you the whole picture.

One thing I noticed is while Aerospikes have better ISP than a RAPIER engine, once you add the weight of a Turbojet, it is actually produces less dV in a closed cycle.

From what I see, the RAPIER engine is nicely balanced.

Edited February 2, 2014 by Mobjack

[Nebur Wolf]

I was thinking on running some tests with rapier engines, but my work is cut down due to this thread.

[Jouni]

My conclusion is that the RAPIER is an excellent engine to introduce new players to the world of spaceplanes (as has been said) but is otherwise fairly sub-standard for these small planes. OTOH, it does the job quite well and reduces part count (not like that's a big consideration here; all versions of the Night Heron were under 40 parts). I think good compromise would be to increase its space ISP while lowering the top speed for the turbofan by a bit (maybe 100m/s or something). At flame-out both of my turbofan planes were well over 2.3km/s while the RAPIER barely managed 2km/s.

The real problem with Rapiers is that turbojets are even more ridiculously overpowered than the 48-7S engines.

Real-world turbojets have not been used succesfully past mid-supersonic speeds (around 800 m/s), and even turbojet-ramjet hybrids (like in SR-71) have peaked at around 900-1000 m/s. Pure ramjets are inefficient until you get close to Mach 1, but after that, they can be used efficiently until low hypersonic speeds (maybe up to 1800 m/s). Scramjets become useful at the speeds where ramjets peak off, so you would really need four different types of engines, if you wanted to switch from jets to rockets at close to (Kerbin) orbital speeds.

The promise of Sabre engines and other similar hybrid designs is that the same engine can be used in air-breathing mode from 0 to 1600 m/s, after which it switches to using stored oxidizer. When a single engine replaces three different ones, it can make real spaceplanes so much lighter and simpler, that they might actually work some day.

[Levelord]

The real problem with Rapiers is that turbojets are even more ridiculously overpowered than the 48-7S engines.

Real-world turbojets have not been used succesfully past mid-supersonic speeds (around 800 m/s), and even turbojet-ramjet hybrids (like in SR-71) have peaked at around 900-1000 m/s. Pure ramjets are inefficient until you get close to Mach 1, but after that, they can be used efficiently until low hypersonic speeds (maybe up to 1800 m/s). Scramjets become useful at the speeds where ramjets peak off, so you would really need four different types of engines, if you wanted to switch from jets to rockets at close to (Kerbin) orbital speeds.

The promise of Sabre engines and other similar hybrid designs is that the same engine can be used in air-breathing mode from 0 to 1600 m/s, after which it switches to using stored oxidizer. When a single engine replaces three different ones, it can make real spaceplanes so much lighter and simpler, that they might actually work some day.

You're arguing that the RamJet is overpowered because it doesn't conform to it's real world counterpart, while implying the RAPIER is balanced because it conforms to a hypothetical and unproven real world counterpart. It is flat out an unfair comparison.

[inigma]

Captain Sierra, please feel free to post on your OP the following engine design pioneered by Shuttlecraft Systems Inc. just a few days after the RAPIER became available, and is in use in our highly successfull Proton III Type-1A Shuttlecraft, a demonstration of which is seen in the Proton III video (see in my sig for the thread link):

Javascript is disabled. View full album

A Turbojet clipped into a RAPIER, stacked with a Nuke. This engine achieves the best of all worlds. A little heavy, but the the pros outweigh the cons. Test it out sometime.

[Jouni]

You're arguing that the RamJet is overpowered because it doesn't conform to it's real world counterpart, while implying the RAPIER is balanced because it conforms to a hypothetical and unproven real world counterpart. It is flat out an unfair comparison.

No. What I am arguing is that the performance of real-world jet engines is pretty close to the basic jet engine in Kerbal, while the turbojet engine in Kerbal more like a hybrid of a turbojet/turbofan engine, a ramjet engine, and a scramjet engine. When you already have access to that kind of superengines, a hybrid that combines the performance of a turbojet/turbofan, a ramjet, and a rocket engine is no longer such a promising concept.

[regex]

No. What I am arguing is that the performance of real-world jet engines is pretty close to the basic jet engine in Kerbal, while the turbojet engine in Kerbal more like a hybrid of a turbojet/turbofan engine, a ramjet engine, and a scramjet engine. When you already have access to that kind of superengines, a hybrid that combines the performance of a turbojet/turbofan, a ramjet, and a rocket engine is no longer such a promising concept.

So basically what you're saying is that the RAPIER should top out at 1.6km/s in air-breathing mode while the turbojet, which I'll assume is a hybrid SR-71 type engine, should top out at 1km/s. Realistically you'd also have to cut that speed to accommodate the pea soup, maybe about in half since Kerbin takes a bit more than half the delta-V to orbit as Earth. Let's not get into part mass, though, because KSP's payload fraction is pretty ridiculous.

[Jouni]

So basically what you're saying is that the RAPIER should top out at 1.6km/s in air-breathing mode while the turbojet, which I'll assume is a hybrid SR-71 type engine, should top out at 1km/s. Realistically you'd also have to cut that speed to accommodate the pea soup, maybe about in half since Kerbin takes a bit more than half the delta-V to orbit as Earth. Let's not get into part mass, though, because KSP's payload fraction is pretty ridiculous.

I'm not sure if there is a good way to fix the situation. One part of the problem is that you can get so close to orbital speeds with air-breathing engines that the jet stage is the only thing that really matters. If you capped all air-breathing engines to around 1500 m/s, the rocket stage would also need to do a significant amount of work to reach orbit, so different trade-offs would become possible. But even then, it would be very hard to balance the engines so that neither Rapier nor a jet/rocket combo would be superior under most circumstances.

If Rapier replaced three engine types instead of two, like Sabre is supposed to do in the real world, things could be different. Maybe you would then use Rapiers in lighter designs, while heavier spaceplanes could be made more efficient by using normal jets from 0 to 500 m/s, switching to high-speed air-breathing engines (that would not work at lower speeds) for 500 to 1500 m/s, and finally firing the rockets to reach the orbit. Or maybe you would end up with two-stage designs, where a normal jet carries the spaceplane to a high altitude, where it can start using its high-speed air-breathing engines.

[TimePeriod]

TL : DR

RAPIER engines, yes/no?

[Captain Sierra]

TL:DR

RAPIER engines, yes/no?

Long and short of it is they open up a lot of new doors for SSTO design while not really invalidating most designs (except possibly a few jet + rocket combos depending on the rocket used).

[Levelord]

TL : DR

RAPIER engines, yes/no?

(1.) New to SSTO design and need an introductory engine to teach you flight ascent paths for SSTOs: RAPIER

(2.) SSTO veterans that want to maximize efficiency and payload fraction: Turbojet/lv-909/48-7s etc etc combo

I honestly don't know why people deliberately want to shoehorn themselves into choosing one engine over another. The thread starts off with people swearing by the RAPIER as a more powerful/efficient engine, but when confronted with the raw data and efficiency of the TurboJet, they jump to the argument that the Turbojet or the 48-7s (or likely even the ant/ion engine for that matter) etc etc... are overpowered. As if every other engine in the game is responsible for the RAPIER's problems.

They are missing the point that the RAPIER is there to fill a niche, to bridge the learning curve for SSTO design. Then they feel threatened that if the other engines do better than the RAPIER, they must either all be overpowered or invalidate the RAPIER's reason for being in the game at all. It's not the point at all, and I'll repeat again that it's purely there to bridge the learning curve gap for its ease of use, while having the drawback of being less efficient overall compared to other methods. If the RAPIERs were ever slated o be removed from the game I would in fact argue to keep it in the game because it is a very good tool to help teach newer players on how to design SSTOs.

Edited February 2, 2014 by Levelord
spelling

[BubbaWilkins]

(1.) New to SSTO design and need an introductory engine to teach you flight ascent paths for SSTOs: RAPIER

(2.) SSTO veterans that want to maximize efficiency and payload fraction: Turbojet/lv-909/48-7s etc etc combo

I honestly don't know why people deliberately want to shoehorn themselves into choosing one engine over another. The thread starts off with people swearing by the RAPIER as a more powerful/efficient engine, but when confronted with the raw data and efficiency of the TurboJet, they jump to the argument that the Turbojet or the 48-7s (or likely even the ant/ion engine for that matter) etc etc... are overpowered. As if every other engine in the game is responsible for the RAPIER's problems.

They are missing the point that the RAPIER is there to fill a niche, to bridge the learning curve for SSTO design. Then they feel threatened that if the other engines do better than the RAPIER, they must either all be overpowered or invalidate the RAPIER's reason for being in the game at all. It's not the point at all, and I'll repeat again that it's purely there to bridge the learning curve gap for its ease of use, while having the drawback of being less efficient overall compared to other methods. If the RAPIERs were ever slated o be removed from the game I would in fact argue to keep it in the game because it is a very good tool to help teach newer players on how to design SSTOs.

you and others also keep bringing up this same argument, but the math doesn't support it either.

(1) Rapier:

190 Thrust in Air

175 Thrust in Vac

1.75 Mass

(1) Turbo Jet and (6) 48.7s

200 Thrust in Air

180 Thrust in Vac

1.8 Mass

I(1) Turbo Jet and (4) LV-909

200 Thrust in Air

200 Thrust in Vac

3.2 Mass

(1) Turbo Jet and (3) LV-N 's

200 Thrust in Air

180 Thrust in Vac

7.95 Mass

In each of those combinations, the non rapier combo weighs more than the rapier for near comparable results. You would have to compare fuel consumption rates, ISP curves, etc to really get into the nuts and bolts and declare a true winner. The upside is that smaller craft like most of those posted in this thread don't need to match the thrust in Vac of the Rapier, so it makes more sense to just use a few smaller engines. But, on those larger craft where you need the higher thrust of the Rapier, those smaller engine combo's with (2) 48-7S's and such just don't have enough punch to move the craft. This is where the Rapier shines.

Now what would be interesting is to see if you throttled a Rapier to match the thrust of the smaller engine combo's in VAC if it actually out performs them as well. Perhaps I'll give this a try.

[Levelord]

you and others also keep bringing up this same argument, but the math doesn't support it either.

(1) Rapier:

190 Thrust in Air

175 Thrust in Vac

1.75 Mass

(1) Turbo Jet and (6) 48.7s

200 Thrust in Air

180 Thrust in Vac

1.8 Mass

I(1) Turbo Jet and (4) LV-909

200 Thrust in Air

200 Thrust in Vac

3.2 Mass

(1) Turbo Jet and (3) LV-N 's

200 Thrust in Air

180 Thrust in Vac

7.95 Mass

In each of those combinations, the non rapier combo weighs more than the rapier for near comparable results. You would have to compare fuel consumption rates, ISP curves, etc to really get into the nuts and bolts and declare a true winner. The upside is that smaller craft like most of those posted in this thread don't need to match the thrust in Vac of the Rapier, so it makes more sense to just use a few smaller engines. But, on those larger craft where you need the higher thrust of the Rapier, those smaller engine combo's with (2) 48-7S's and such just don't have enough punch to move the craft. This is where the Rapier shines.

Now what would be interesting is to see if you throttled a Rapier to match the thrust of the smaller engine combo's in VAC if it actually out performs them as well. Perhaps I'll give this a try.

Your math is over-simplified and it leaves out the ISP, TWR and final horizontal velocity, fuel payload fraction and ascent profile.

The fuel consumption rates, ISP, intake ratio have already been discussed in detail in the previous pages of the thread by me and a few others (Yes we did the homework; on page 2 and page 6). We also came up with the conclusion that the traditional Jet/rocket combo is better for smaller craft while the RAPIER is suited for larger craft. Which when you think about it, newer players to SSTO designs tend to build fairly large aircraft to carry a lot more fuel with them while the older players tend to be more minimalistic, so the engines fit their roles. There are slightly heavier craft I had seen in the past in 0.18 that uses ion engines to breach the atmosphere from Jet powered flight. I've addressed before that it's not the TWR that gets you into space, but the final horizontal velocity you achieve (roughly 2250m/s) which is near orbital velocity where you only need a very small amount of thrust to raise the apoapsis into space.

Edited February 2, 2014 by Levelord

[Jouni]

They are missing the point that the RAPIER is there to fill a niche, to bridge the learning curve for SSTO design. Then they feel threatened that if the other engines do better than the RAPIER, they must either all be overpowered or invalidate the RAPIER's reason for being in the game at all.

I believe that insulting other users is forbidden by forum rules.

Let's get to the point. We can view KSP as a casual game, where performance statistics can be pretty much arbitrary, if it makes the game more fun. Or we may view the game as a simulation, where different engines should occupy pretty much the same roles as their real-world inspirations. Depending on which view you choose, you get pretty different conclusions on what is right and wrong with different engines.

[Comrade Jenkens]

That fact that there is so much arguing about which is better proves that its balanced pretty well!

[Levelord]

I believe that insulting other users is forbidden by forum rules.

Let's get to the point. We can view KSP as a casual game, where performance statistics can be pretty much arbitrary, if it makes the game more fun. Or we may view the game as a simulation, where different engines should occupy pretty much the same roles as their real-world inspirations. Depending on which view you choose, you get pretty different conclusions on what is right and wrong with different engines.

It's not insulting as it is what really happened in the thread. For example, the Turbojet engine was stated to be underpowered compared to the RAPIER because proponents couldn't get Turbojet powered spaceplanes into orbit. Then, when confronted with raw data and a successful test flight, they changed their argument that the Turbojet engine is overpowered. When the Turbojet was presented in either light it was touted as underpowered or overpowered despite the data which means that they are simply finding excuses to find flaws with the turbojet engine whether it does well or poorly. It's like saying a student failed an exam because he's stupid and lazy, but when confronted with an 'A' in another test result, you conclude he cheated. It's not a fair conclusion and it puts the student in a position where no matter what he does, good or bad, that he's just not satisfactory to the examiner. That's the same thing with the TurboJet engine. In many ways I should be the one that feels insulted that people think that I would simply ignore the objective data and throw my lot in with one engine over the other.

I've reconciled my preference for both engines in differing situations and I find both useful in different situations, but there are some people in the thread who thinks that one method is clearly useless while the other is the pinnacle of SSTO design. What I brought to light is that the differences in ascent profile and raw stats determine the performance and choice of engine.

We were all newbies at some point when we bought the game, so it's also not insulting to suggest that the RAPIER engine is for newbies. It took me several weeks to figure out how to get to orbit myself. I find it unreasonable to expect people who begin playing this game to immediately be experts in the game's ins and outs and how KSP's physics function in it's own unique way. Even seasoned astronauts have problems playing KSP when they were first introduced to it (as shown in a Scott Manley video). So I feel that the RAPIER as an introductory engine has a very reasonable place in KSP and I strongly encourage first time SSTO builders to give it a try.

I just don't like how the RAPIER is hailed as a one size fits all engine when the data is not presented in a fair and objective manner. So I provide counter-points to show how both are valid in different situations.

edit: I also disagree about the statistics in KSP being meaningless and arbitrary because that would mean that by extension, mods that rely on those numbers like Kerbal engineer and protractor are themselves rendered meaningless. Truth of the matter is, you can build any craft big enough with enough engines, and you can ignore what the numbers are and still get into orbit. However for the purposes of this thread where the primary topic is about efficiency, minimalist craft rely on those numbers because the margin for error is significantly smaller.

Edited February 2, 2014 by Levelord

[BubbaWilkins]

We were all newbies at some point when we bought the game, so it's also not insulting to suggest that the RAPIER engine is for newbies. It took me several weeks to figure out how to get to orbit myself. I find it unreasonable to expect people who begin playing this game to immediately be experts in the game's ins and outs and how KSP's physics function in it's own unique way. Even seasoned astronauts have problems playing KSP when they were first introduced to it (as shown in a Scott Manley video). So I feel that the RAPIER as an introductory engine has a very reasonable place in KSP and I strongly encourage first time SSTO builders to give it a try.

With 1000's of hours of gamplay under my belt, I find that I prefer the Rapiers in most of my craft. If that makes me a newb, then so be it. Many also refer to grenade launchers in 1st person shooters as newb tubes. 1 on 1 sure, but if you manage to wipe out 2-3 in one hit, that is the proper and efficient application of ordinance. The Rapier is easy to use, but I think it's niche extend well beyond that point. Agreed, it's not the best choice in all instances, but it's also not the worst.