Buff Rapier vacuum Isp

[KerikBalm]

Since MH came out, with the skiff and very OPd wolfhound in particular, I've been taking a look at part balance, and the real world part analogues.

It occurs to me that the rapier closed cycle is of very little use in stock - the margins between a LF only design with nukes, a Rapier only design, and rapier+aerospike, 909, and now a wolfhound, are pretty small.

 

Rapiers are used for their superior airbreathing qualities, and many designs don't even make use of their close cycle function, or only make brief use of it.

Rapier closed cycle Isp should not be comparable to 1st stage lifter engines meant to operate at 1 atm. They should be optimized for the air from 20km and up (so vacuum and near vacuum). Likewise the "real-ish" SABRE would be very competetive in a vacuum, and it would make sense - since it will be using turbines and atmospheric air lower in the atmosphere, and it  would actually deliver most of the dV needed to get to orbit in closed cycle mode.

I think the rapier would be better situated to fill its gameplay purpose (being a dual mode engine, not just a super-turbo-ramjet, and being an analogue of a real engine concept) if it was given the Isp curve of the Rhino, or perhaps the Skiff. It would thus have relatively good vacuum Isp, but still less than dedicated vacuum engines (terrier, cheetah, poodle, and... ughh.... the OPD wolfhound)

And it would still be a terrible option to use as just a rocket engine, because an aerospike would generate the same thrust at the same or better Isp (or just outright better if the Skiff Isp curve is used), for half the mass and it would include an alternator. The new cheetah engine can also be used as a 1.25m engine, giving better vac Isp and TWR while keeping vectoring AND adding an alternator.

Thoughts? 

[GoSlash27]

I like the RAPIER the way it is. It's poor stats reflect the reality that combining a jet turbine and a rocket into a single engine yields a result that isn't very good at either job.

Despite it's shortcomings, it's still the best engine out there for it's intended purpose; SSTO space planes.

JMHO,
-Slashy

Edited April 10, 2018 by GoSlash27

[Ace in Space]

It's the spork of engines - its strength is it does two jobs at once, but at the cost of not being as good at either as a dedicated engine. Jack of all trades, master of none.

[KerikBalm]

6 hours ago, GoSlash27 said:

I like the RAPIER the way it is. It's poor stats reflect the reality that combining a jet turbine and a rocket into a single engine yields a result that isn't very good at either job.

Despite it's shortcomings, it's still the best engine out there for it's intended purpose; SSTO space planes.

JMHO,
-Slashy

"reflects the reality". The reality is that the SABRe its modled after would have an Isp comparable to the SSMEs.

Also, its not true that its not very good at either job. Its excellent at one of its jobs, its the best airbreathing engine there is for high speed and high altitude flight. You could remove the closed cycle mode completely, and it would still be used for its high speed and altitude airbreathing performance.

1 hour ago, Ace in Space said:

It's the spork of engines - its strength is it does two jobs at once, but at the cost of not being as good at either as a dedicated engine. Jack of all trades, master of none.

As above, its the best air-breathing engine for high speed and high altitude flight. The closed cycle mode is the only area where it really sucks as an engine.

Basically the only reason to use close cycle mode is to avoid the drag penalty of adding another stack to use another engine type (this can be avoided with some use of the offset tool/part clipping). In stock this doesn't matter so much, and now that there's the Wolfhound, I'm looking at the rapier closed cycle mode as basically completely obsolete in both stock and a 3x resized system.

Edited April 10, 2018 by KerikBalm

[AeroGav]

2 hours ago, KerikBalm said:

As above, its the best air-breathing engine for high speed and high altitude flight. The closed cycle mode is the only area where it really sucks as an engine.

Basically the only reason to use close cycle mode is to avoid the drag penalty of adding another stack to use another engine type (this can be avoided with some use of the offset tool/part clipping). In stock this doesn't matter so much, and now that there's the Wolfhound, I'm looking at the rapier closed cycle mode as basically completely obsolete in both stock and a 3x resized system.

You are correct in that the SABRE, unlike the RAPIER, has a plug type nozzle that gives decent performance across a wide range of air pressures,  purportedly not far below the state of the art in terms of vacuum isp.      The problem is you are going to hard time attracting support from the community because it is already seen as one of the most OP engines in the game.  You'd really have to package this up with a set of nerfs to address the following :

• The Rapier should require liquid hydrogen.  LH2 has more heat absorption per KG than even water, no other fuel comes close in this regard.     This means to have a RAPIER you have to deal with boil off and more importantly,  very bulky (draggy) fuel tanks.   Without LH2,  the use of fuel to pre-cool the intake air - the thing that differentiates it from a Whiplash - is not viable.
• The Rapier gets this crazy ramjet boost to max thrust at high speed, peaking at over 8 times the static thrust rating at mach 3.7.   This is not a feature of the real SABRE and seem to ignore flow limits of the intake/precooler/compressor systems, rpm limits of the turbines and max chamber pressures.    This banzai surge from mach 1.5 - 4 makes possible the blatantly unrealistic flight profile of flying level after takeoff and accelerating to mach 4/5 at sea level, before pitching up and holding a constant pitch  angle till clear of the atmosphere.     The fact that thrust goes to near infinite levels, allows the massive drag to be overcome.    In reality no aircraft has gone more than a hair over mach 1 at sea level.    Any airframe that generates enough lift to get airborne before the tyres explode will be too draggy to push high speeds at sea level.    At altitude,  it's a different story.     Same with the SABRE engine -  the ram air effect of increasing speed compensates for the thinning air at altitude, but will not allow it to greatly exceed its sea level thrust due to choke limits.  The pre-cooling deals with adiabatic heating, up to a point.

 

Doing both of those would require major changes,  but they should perhaps give the RAPIER a much flatter thrust curve in its config files, to at least acknowledge the above, if they're going to buff the vacuum ISP like you suggest.    That will still break a load of legacy craft files, but then that's something with come to expect with version updates !

 

 

13 hours ago, KerikBalm said:

t occurs to me that the rapier closed cycle is of very little use in stock - the margins between a LF only design with nukes, a Rapier only design, and rapier+aerospike, 909, and now a wolfhound, are pretty small.

 

My preferred propulsion is now one Rapier ,  one Panther and two NERVs per 40 tons of craft weight.    I don't use the close cycle mode at all.    Can only see it making sense in edge cases like what Matt Lowne does with RAPIER SSTO lifting an ION system into orbit.     I can get better payload fraction going liquid fuel only (mass of the nervs more than offset by  absence of massive LFO tanks)  but RAPIER only gives lowest dry mass, which enables the Xenon drive payload to get more delta V.

 

[klgraham1013]

If someone smarter than me came up with a good balance change, I'd love to see it.  Even just a nerf to the air breathing side in exchange for a buff to the closed cycle side.  I'll be keeping an eye on this thread, and what the community is thinking.

[KerikBalm]

1 hour ago, AeroGav said:

The problem is you are going to hard time attracting support from the community because it is already seen as one of the most OP engines in the game.  You'd really have to package this up with a set of nerfs to address the following :

• The Rapier should require liquid hydrogen.  LH2 has more heat absorption per KG than even water, no other fuel comes close in this regard.     This means to have a RAPIER you have to deal with boil off and more importantly,  very bulky (draggy) fuel tanks.   Without LH2,  the use of fuel to pre-cool the intake air - the thing that differentiates it from a Whiplash - is not viable.
• The Rapier gets this crazy ramjet boost to max thrust at high speed, peaking at over 8 times the static thrust rating at mach 3.7.   This is not a feature of the real SABRE and seem to ignore flow limits of the intake/precooler/compressor systems, rpm limits of the turbines and max chamber pressures.    This banzai surge from mach 1.5 - 4 makes possible the blatantly unrealistic flight profile of flying level after takeoff and accelerating to mach 4/5 at sea level, before pitching up and holding a constant pitch  angle till clear of the atmosphere.     The fact that thrust goes to near infinite levels, allows the massive drag to be overcome.    In reality no aircraft has gone more than a hair over mach 1 at sea level.    Any airframe that generates enough lift to get airborne before the tyres explode will be too draggy to push high speeds at sea level.    At altitude,  it's a different story.     Same with the SABRE engine -  the ram air effect of increasing speed compensates for the thinning air at altitude, but will not allow it to greatly exceed its sea level thrust due to choke limits.  The pre-cooling deals with adiabatic heating, up to a point.

First point:

Well, I'm sure they won't implement cryogenic fuels and boiloff and LH2 tanks. I would be ok if they did though. It would hurt the LV-N more than it would hurt a craft that just needs to get to orbit, dump off a payload, and retroburn. Boil-off for a surface to orbit cargoshuttle won't be an issue. It would be for an interplanetary LV-N tug. Anyway, we also don't have many good LF only tank options (the mk2 and mk3 tanks are very draggy, compared to cylindrical tanks, particularly when flying with some AoA), so I often end up using jumo 64 tanks with some oxidizer emptied.That tends up fitting the bulky fuel tank requirement.

As for the whiplash, the whiplash still gets better static thrust/TWR, has an alternator, and consumes 20% less fuel for the same amount of thrust, so... if you don't want to go to space, but you do want high speed atmospheric travel, it does seem pretty competitive. If there is no room for the Whiplash, then other engines would also have major problems (panther, the turbofan).

Second point:

"peaking at over 8 times the static thrust rating" No, not really/in practice. It has a peak thrust multiplier of 8 at mach 3.7, but it also has a flow multiplier cap of 3x. It cannot produce more than 3x the thrust it produces at mach 0 at 1 atm.

When it gets in thinner air, and fuel flow and thrust drop in the thinner atmosphere, then it can have 8x the thrust it would have if it were stationary at that altitude (but why would it ever be?)

At high speed and altitude, the velocity curve may give an 8x multiplier, while the atmcurve gives a 0.1x multiplier, and the result is its not getting some crazy TWR.

Also, accelerating that fast at low altitude is not the flight profile I use for it, particularly in 3x gameplay where every bit of fuel counts.

Due to the atmCurves of the whiplast and rapier, the "thrust to drag" ratio gets better at higher altitudes, and I find that its best to go into a steady subsonic climb, and then start the "speed dash" at some altitude after 5km (I normally aim to hit the optimum part of the "power curve" around 10km, so I can pretty much max out the speed as I'm passing 20-25km and thrust starts dropping fast).

 

My main point is that its an excellent jet engine (indeed the advance of the SABRE engine is in jet tech, not rocket tech), but the closed cycle mode is so bad that its really kind of useless, people will just use other LFO engines. The rapier closed cycle isn't really serving much purpose with such bad stats.bumping it from 305 to 320/330/340 along with the other LFO vacuum engines (or even 315, to be like the vector/mammoth) would go a long way towards making closed cycle more than a novelty.

It would still be bad as an LFO engine by itself, due to its TWR (again, the aerospike is half the mass for the same thrust, and currently has better Isp). The only LFO engines with worse vacuum Isp than the rapier are the small radial engines, and the specialty twin boar booster. A sabre analogue should do well in a vacuum, and even for gameplay, the only time it makes sense to use closed cycle mod is when you're nearly in a vacuum. So why does it have pretty good atmospheric Isp (it beats the poodle, terrier, spark, rhino, reliant, swivel, etc) and terrible vacuum Isp?

Edited April 10, 2018 by KerikBalm

[AeroGav]

5 minutes ago, KerikBalm said:

 

As for the whiplash, the whiplash still gets better static thrust/TWR, has an alternator, and consumes 20% less fuel for the same amount of thrust, so... if you don't want to go to space, but you do want high speed atmospheric travel, it does seem pretty competitive. If there is no room for the Whiplash, then other engines would also have major problems (panther, the turbofan).

You can see why the RAPIER is such a good jet engine even if you're not using close cycle mode.   The thrust multiplier remains above 3 till mach 5.5,  given that orbital velocity (at stock scale) is about mach 7,  you only need 1.5 mach's worth of rocket delta V.         The Whiplash may have 20% better fuel consumption in the air breathing phase, but it drops to a multiplier of 3 by mach 4.5, which means you need 2.5 machs worth of rocket delta V.       Even if you are using LV-Ns,  rocket mode fuel consumption is 4x that in air breathing,  so you can see why the Rapier's higher top speed always wins.

Quote

 

First point:

Well, I'm sure they won't implement cryogenic fuels and boiloff and LH2 tanks. I would be ok if they did though. It would hurt the LV-N more than it would hurt a craft that just needs to get to orbit, dump off a payload, and retroburn. Boil-off for a surface to orbit cargoshuttle won't be an issue. It would be for an interplanetary LV-N tug.

 

Well, there are other fuels that an NTR can run off than LH2.         For example,  Ammonia (which can also be used in jet engines).  Liquid ammonia is a storable propellant.   It won't give you anything like the ISP of NTR / LH2  (though it does partially dissociate to hydrogen in the heat of the reactor) , but it's still heaps better than storable chemical propellants.      The main thing is that large LH2 tanks mean drag, which is worse on a space plane than a rocket.

BTW, in this game rocket engines correctly attached  give very little drag, even if you don't clip nose cones into their rear attach nodes.   The vast majority of the drag is coming from the fuselage, ie.  the LFO tanks of your thirsty closed cycle RAPIERS.

 

19 minutes ago, KerikBalm said:

"peaking at over 8 times the static thrust rating" No, not really/in practice. It has a peak thrust multiplier of 8 at mach 3.7, but it also has a flow multiplier cap of 3x. It cannot produce more than 3x the thrust it produces at mach 0 at 1 atm.

When it gets in thinner air, and fuel flow and thrust drop in the thinner atmosphere, then it can have 8x the thrust it would have if it were stationary at that altitude (but why would it ever be?)

At high speed and altitude, the velocity curve may give an 8x multiplier, while the atmcurve gives a 0.1x multiplier, and the result is its not getting some crazy TWR.

Sounds like the best compromise would be to give it the Vacuum / Atm  closed cycle ISP values of the Dart aerospike,    and to reduce this flow multiplier to 1.5.

People will have to use a proper flight profile.    You can still hit the same airbreathing top speed, but you'll need to do that in level flight at high altitude, not while passing through 8km  in a steep climb.      Flights to orbit will take a minute or so longer, but if you don't like flying, why are you building space planes.

I'll have to redesign my craft to incorporate more RAPIERs,  but will probably start using Oxidizer again and fit fewer NERVs/Panthers.

[KerikBalm]

1 hour ago, AeroGav said:

You can see why the RAPIER is such a good jet engine even if you're not using close cycle mode. ....  so you can see why the Rapier's higher top speed always wins.

I never said that it didn't. I said pretty much the opposite in response to it being compared to a "spork" that can perform 2 functions, but excells at neither. Its clearly not a spork, its airbreathing mode is great if you intend to reach orbit. That doesn't mean that the whiplash is useless, just as the whiplash doesn't make the panther or wheesley useless. The whiplash still has advantages over the rapier (that aren't so relevant when you want to get a craft/payload to orbit).

 

Quote

Well, there are other fuels that an NTR can run off than LH2.         For example,  Ammonia (which can also be used in jet engines).  Liquid ammonia is a storable propellant.   It won't give you anything like the ISP of NTR / LH2  (though it does partially dissociate to hydrogen in the heat of the reactor) , but it's still heaps better than storable chemical propellants.

For a solid core NTR (what the LV-N seems to be), Ammonia would only give an Isp of about 510s when lH2 is giving about 800s. The 800s value of the LV-N clearly means its running on lH2.

Quote

Sounds like the best compromise would be to give it the Vacuum / Atm  closed cycle ISP values of the Dart aerospike,    and to reduce this flow multiplier to 1.5.

People will have to use a proper flight profile.    You can still hit the same airbreathing top speed, but you'll need to do that in level flight at high altitude, not while passing through 8km  in a steep climb.      Flights to orbit will take a minute or so longer, but if you don't like flying, why are you building space planes.

I'll have to redesign my craft to incorporate more RAPIERs,  but will probably start using Oxidizer again and fit fewer NERVs/Panthers.

I don't know about the aerospike profile... but I'm not sure its atmospheric Isp value really matters at all... There are only 4 bodies that one can land on with an atmosphere. 2 of them (kerbin & laythe) have O2, and thus you'd use open cycle and not closed cycle mode. Duna's atmosphere is so thin that you basically get the Vacuum value anyway.

That just leaves Eve... and it makes no sense to send a spaceplane to eve with rapier engines on it. The rapier engine mass is still much higher.... so if it doesn't have O2 in the atmosphere, why would you bring the rapiers, and if it does, why would you use closed cycle in the atmosphere? Sure there are single stage designs that can go to minmus, duna, etc, and back (as opposed to just dropping payloads off in orbit like I mostly do), but none of those are going to eve, so the atmospheric Isp again doesn't really matter.

Still, I'd give it the Rhino's Isp curve (205-340) instead of the Dart's Isp curve (290-340)

 

I think the flow cap (and in a way, TWR) being halved to 1.5 instead of 3 might be a bit harsh for an 11.5% Isp increase... perhaps change the flow cap to 2? I just checked the Whiplash file, and its flow cap is 2 (although note that the whiplash has a higher starting thrust).

*edit*

Quote

The problem is you are going to hard time attracting support from the community because it is already seen as one of the most OP engines in the game.

I don't think this is the case. In fact, I've posted in many threads in defense of the Rapier's utility. Its often accused of being not any good.

Spaceplanes in general are often accused of being OP, and that really has little to do with the engines, but its more to do with the size of Kerbin (try Sigma Dimensions and scaling kerbin up to 3, or 6x size). The engines do have an Isp that is too good for a kerosene burning jet, but would be just fine for a lH2 burning jet.... now the RSS folks might like to say that KSP rocket fuel most closely resembles aerozene and N2O4... but clearly that isn't what the LV-N is using. Squad seems to intentionally want "liquid fuel" to be generic "whatever you want it to be/whatever the engine needs it to be to make sense".

Of course... saying that the engines aren't OP'd, kerbin is just too small would also be an argument against the wolfhound being OP'd, and that engine is definitely OP'd (although as an SPS analogue, its Isp is higher than the real thing, but it is inline with... say a J-2, leading to some to suspect the stats are switched.).

At least the rapier Isp stats are less than the SABRE's predicted stats (as opposed to the wolfhound having a full 100s better than its real analogue... 412 vs 312)

https://en.wikipedia.org/wiki/SABRE_(rocket_engine)

SABRE/Rapier

Vacuum Isp: 460s /305s

Airbreathing Isp: 3600s / 3200s

I'm not suggesting that we take 32/36 and multiply that by 460s to give it a vacuum Isp of 408s... just that it should have a top tier vacuum Isp. So in KSP at the moment (OP wolfhound excluded), that would be 340-350. I'd even take 330 (of which the Skiff is the only engine with an Isp in that range)

Edited April 10, 2018 by KerikBalm

[TheFlyingKerman]

The problem now is bottom tier engines like the Swivel or Terrier are totally useless...

[KerikBalm]

Thats a problem whether or not the rapier closed cycle gets an Isp buff.

Also... the terrier is not useless - its fairly in line with the cheetah and poodle. Its the swivel and reliant that are quite bad compared to their larger diameter counterparts, but that's a topic for another thread. (IMO, the bobcat also needs a buff, as does the kodiak)