Is ISP consistent between KSP and Real Life?

[G'th]

As part of my planned update to FASA, I want to include engine variants that I've had for a while as part of it. Things like the F-1A/B, J2/S/X etc.  However, I also want to make sure that the ISP is correct, because it now occurs to me that in all the tiem I spent on the variants getting them tweaked to perform just right at Kerbal scale, I never actually changed the ISP. 

Right now, I've gone ahead and matched the ISP's of the engines to what they should be in real life, but I'm concerned on whether or this would be accurate for simulating at the game's scale. So essentially my question is, does, for whatever reason, ISP work differently in the game than it does in real life?

[Frozen_Heart]

I think that ISP in KSP is generally a lot lower than in real life.

For example the Shuttle engine in real life has a sea level ISP of 366s and a vac of 452s.

Compare this to the Vector engine which is KSPs shuttle engine. It has a sea level ISP of 265s and a vac of 315s. Its vacuum efficiency is lower than the RL shuttles sea level efficiency.

[cantab]

KSP's engines all took an Isp nerf in 1.0 I think, when the aerodynamics got updated and drag took a dramatic drop.

But actually KSP's engines now aren't far out of line for real-world hypergolic fuelled engines.

[Streetwind]

3 hours ago, G'th said:

does, for whatever reason, ISP work differently in the game than it does in real life?

Nope. Specific impulse is a super simple definition that can't really be changed. It says: when putting out 1 N of thrust, and having 1 kg of fuel available (or: putting out 1 kN, and having 1 metric ton of fuel), the engine will run for this long. That's why the unit is in seconds.

What's different in KSP is that the celestial bodies, as well as the distances between them and their atmospheric ceilings, are much smaller. So you need a lot less dV. If you went and used real life stats for everything - hydrogen engines with 460s Isp, engine thrust-weight ratios of over 150, tank mass ratios of 20-30... well, let's just say, you wouldn't even need to stage your rocket to go to Duna anymore.

Thus the stats that the rocketry equipment has in KSP are mostly informed by gameplay considerations. Any higher than what they are now, and gameplay becomes trivially (and immersion-breaking-ly) easy.

Edited May 25, 2016 by Streetwind

[KerikBalm]

ksp engines mostly have Isps corresponding to Kerosene- liquod O2 rockets.

https://en.wikipedia.org/wiki/Rocketdyne_F-1

Vacuum Isp 304 : Sea level Isp: 263

https://en.wikipedia.org/wiki/RD-170

337/309

https://en.wikipedia.org/wiki/Merlin_%28rocket_engine_family%29#Merlin_1D

1C: 305/275    1D: 311/282     1D-vacuum: 348 (couldn't find SL stats)

https://en.wikipedia.org/wiki/AJ10

311  (Note, this is not a Kerlox fuel, but.. anyway... also couldn't find SL stats)

So vacuum Isps between 304 to 348 for Kerlox fuels in real life

In KSP, the inline LFO rocket engines all have Isps between 300 to 350... so that seems realistic. The radial ones go down as low as 290 (only for the small sizes)

Likewise, the sea level Isp goes as low as 263 (they don't bother listing SL stats for vacuum engines it seems). In KSP... we've got the LV-T45, and its 270 SL Isp... seems realistic... as does the KR-2L's 255 Isp (the KR-2L is sort of borderline as a vacuum engine).

So... it seems there is no Hydro-lOx fuel engines in KSP (which would give Isps well over 400)

The NERVA engine that they tests had an Isp of about 800... although I think they were making improvements and expecting perhaps 850... anyway, the KSP LV-N with 800 Isp is definitely realistic... except that as the above engines establish... liquid fuel seems to be a kerosene derivative, not liquid hydrogen. The LV-Ns 2:1 TWR is twice what the NERVA got... but more advanced designs could probably do better in terms of both TWR and Isp (1000 Isp, and perhaps up to 30:1 TWR)

https://en.wikipedia.org/wiki/Project_Timberwind

https://en.wikipedia.org/wiki/Space_Shuttle_Solid_Rocket_Booster

268/242

Best KSP SPR: 220/195 ... realistic range, a bit nerfed

So... yea, they are generally realistic Isps... the TWRs are generally too low, the tanks are generally to heavy... but the dV you need to get anywhere in the Kerbal's system is roughly 1/3 what it would take to get to an analagous location in our solar system(Mun=Moon, Kerbin=Earth, Duna= Mars, Jool = Jupiter, etc), due to the roughly 1/10th scale of the distances, but mostly maintaining similar surface gravities.

So in essence, playing stock KSP is like playing RSS and multiplying all Isps by a factor of 3, and then nerfing the TWRs and increasing the dry mass of many parts, while removing hydrolox as a fuel choice).

 

Jet engines on the otherhand, are roughly 2x as efficient as they should be, assuming liquid fuel is a Kerosene derivative.

The SR-71's engines would fit the description of turboramjets

https://en.wikipedia.org/wiki/Pratt_%26_Whitney_J58

https://en.wikipedia.org/wiki/Specific_impulse#Larger_engines

http://www.astronautix.com/stages/sr71m12.htm

Seems to be somewhere between 2084 and 1900, depending on flight regime I guess... The turboramjet in KSP (seems comparable... its peak thrust is mach 3 but can still push craft much faster, the SR-71 flew at over mach 3) has an Isp of 4000... roughly double.

Similar ratios are seen with the turbofans: Goliath: 12600 constant... relative to a general electric CF-6 at cruising speed of a 747... roughly 6,000 ... so again about a 2:1 ratio.

But the rapier... if we assume its acting like a sabre, using liquid hydrogen... well, then its Isp may be about right because one gets really really good Isps with airbreathing engines and liquid hydrogen... but then its vacuum engine is performing as if its just kerlox... but the rapier would be seriously overpowered if it was the only LFO engine to get stats as if hydrogen was the fuel.

[PB666]

2 hours ago, KerikBalm said:

The NERVA engine that they tests had an Isp of about 800... although I think they were making improvements and expecting perhaps 850... anyway, the KSP LV-N with 800 Isp is definitely realistic... except that as the above engines establish... liquid fuel seems to be a kerosene derivative, not liquid hydrogen. The LV-Ns 2:1 TWR is twice what the NERVA got... but more advanced designs could probably do better in terms of both TWR and Isp (1000 Isp, and perhaps up to 30:1 TWR)

Well that an area of non-discussion in the forums. There is the assumption that Lf in kerbin has properties of hydrogen or kerosine, but with 5 times the density. So in essense its an imaginary fuel.

The problem with KSP is that unless you mod your big-orange tank, there is no suitable storage of a LF only fuel tank, consequently that increases the mass ratio. And of course everytime there is a major update scale mods get ummphed. So its better to create a aerotank model and spec it to carry LF (or hydrogen if you add it to the game).

EDIT: I should point out that it is possible to get a high ISP with Kerosene in a NTR engine if the mass flow is decreased, the plasma is allowed to heat to a much higher temperature, meaning everything is made to be more stable. PV = nRT does not place a limit on how fast carbon can move when heated. Everything will be hotter when it exits. This is not theoretically a problem because the overwhelming energy in an NTR is wasted, because decay utilized is much much lower than potential decay.

Edited May 26, 2016 by PB666

[cantab]

The densities of liquid fuel and oxidizer, based on comparing the physical size of a fuel tank with the mass of fuel it contains, are both about equal to water. The in-game units are not litres, that is a misconception.

[kerbiloid]

6 minutes ago, cantab said:

The densities of liquid fuel and oxidizer, based on comparing the physical size of a fuel tank with the mass of fuel it contains, are both about equal to water.

One can measure only outer size of the tank, not the inner size.

P.S.
As most of parts "found near a road", these tanks and engines should be very rusty inside. So, probably 4/5 of their volume is full of rust.

[hypervelocity]

if you want concord between real life & the game's ISP you should play with the Realism Overhaul suite of mods

[KerikBalm]

Hyper- exactly... when the dV required to get anywhere is about 1/3rd of real values for the analogues, then one should basically take real Isp values and divide by 3... enjoy your LV-N getting ~280 Isp.

But also enjoy much higher TWR ratios and tank ratios.

 

So I'm looking at the space shuttle external tank... it seems 64% of the tankage volume was for the LH2, and the remaining 36% was for the O2 tank (this excludes the non-tankage Intertank part).

Dry mass for the O2 and H2 tanks was 26.5 tons

Hydrogen mass was 106.2 tons. So, using a STS external tank, just empty of O2, would get you a 5:1 wet:dry ratio.

If we reduce the dry mass to be 64% of the value because only 64% of the tank volume was for H2, then the estimated dry mass for a H2 only tank would be 17 tons   (106.2+17)/(17) = 7.24 .... in game... the space plane Mk2 and Mk3 LF only tanks have wet:dry ratios of 8:1 IIRC (while the 1.25m one is 9:1)... so the tank ratios aren't so far off from reality, if we assume they are all H2 tanks... but the ratio is a bit too high, but way too low for kerosene tanks.

For a hydrolox engine... well the space shuttle ET's overall wet:dry ratio was 760/26.5 = 28.7 

Much much much better than the LFO tanks we have at 9:1... and a Kerlox tank would have even better ratios

lets just take 30:1 vs 9:1, and assume 350 Isp (a realistic value and also an ingame value).

A single stage IRL with kerlox would get a maximum of 350*9.81* ln (30) = 11,678 m/s

In game, the best you can get: 350*9.81* ln (9)= 7,544 m/s

It gets a bit worse when you look at capsule mass (at least for the 2.5m crew capsule and lander cabin), and engine mass needed for X TWR... but doesn't come close to offsetting the 1/3 dV requirements

[magnemoe]

1 hour ago, PB666 said:

Well that an area of non-discussion in the forums. There is the assumption that Lf in kerbin has properties of hydrogen or kerosine, but with 5 times the density. So in essense its an imaginary fuel.

The problem with KSP is that unless you mod your big-orange tank, there is no suitable storage of a LF only fuel tank, consequently that increases the mass ratio. And of course everytime there is a major update scale mods get ummphed. So its better to create a aerotank model and spec it to carry LF (or hydrogen if you add it to the game).

EDIT: I should point out that it is possible to get a high ISP with Kerosene in a NTR engine if the mass flow is decreased, the plasma is allowed to heat to a much higher temperature, meaning everything is made to be more stable. PV = nRT does not place a limit on how fast carbon can move when heated. Everything will be hotter when it exits. This is not theoretically a problem because the overwhelming energy in an NTR is wasted, because decay utilized is much much lower than potential decay.

You can use the MK3 liquid fuel tanks with the LV-N so the orange tank dont need an liquid fuel option. 
On the other hand we have an lack of smaller 2.5 meter tanks for fuel only, now we have 1.25 meter and MK2 who is on the small side or jump to 3.75 who is too large for smaller ships. 

[KerikBalm]

A solid core NTR can't get hot enough to make the kerosene a plasma as far as I know.

A gas core one on the other hand... those things would be ridiculous....

Also, while the NTR is actually thrusting, the energy of fission should get captured with reasonable (~50%) efficiency.

A normal chemical rocket is about 60% efficient... I don't see why using heat from a nuclear reactor would be any different... however, the warm up and cool down periods would just be energy getting dumped as heat.

Edited May 26, 2016 by KerikBalm

[PB666]

31 minutes ago, KerikBalm said:

A solid core NTR can't get hot enough to make the kerosene a plasma as far as I know.

A gas core one on the other hand... those things would be ridiculous....

Also, while the NTR is actually thrusting, the energy of fission should get captured with reasonable (~50%) efficiency.

A normal chemical rocket is about 60% efficient... I don't see why using heat from a nuclear reactor would be any different... however, the warm up and cool down periods would just be energy getting dumped as heat.

The capture maybe efficient, but the fission lifetime of thebpellets is much longer than the burntime of gas in the rocket. 

For all intents and purposes above 2500'c most everything will be ionized in some form or the other the gas phase transition between diamond is 3600'c, so effectively carbon is atomized even from diamond at that temperature. However the ionization energies of the C-H bonds are lower and thermal denaturization begins around 800'C and includes cracking of the C12H24 sized carbon chain into succesively smaller bits.

thttps://en.m.wikipedia.org/wiki/Cracking_(chemistry)

This the produces free radicals of different types that includes elements of plasma. . 

The process can be favored by the radiation process, particukarly neutron flow from the core and heat ionized reactants result from fission. This of course would cause alot of the carbon free radicals to preciptitae at high temperatures and sort of gum up the reactor, but at least for a small period you could achieve reasonably high ISPs with kerosene. Again its not a recommendation for use, of all the evils you can through into an NTR hydrogen gas is the least. 

[KerikBalm]

"the fission lifetime of thebpellets is much longer than the burntime of gas in the rocket. "

You said "the overwhelming energy in an NTR is wasted,"

Its only wasted if you throw away the fuel pellets/rods/whatever. If its some sort of reusable nuclear tug, thats not the case. Such a tug and its payloads would be massively radioactive, and not suitable for human crews, of course.

 

"For all intents and purposes above 2500'c most everything will be ionized in some form or the other "

For sure it would start to decompose.. I guess since even a hot flame can be a plasma, these hot decomposed gasses would count as plasma... but solid core NTRs don't actually get so hot... the combustion chamber of the space shuttle main engines, for example, reached higher temperatures than what NERVA obtained... but anyway...

Referring to the rocket bible.... err... I mean the website atomic rockets

"The exhaust velocities are larger than what one would expect given the molecular weight of the propellants because in the intense heat they break down into their components. Ammonia is nice because it breaks down into gases (Hydrogen and Nitrogen). Methane is nasty because it breaks down into Hydrogen and Carbon, the latter tends to clog the reactor with soot deposits. Water is most unhelpful since it doesn't break down much at all."

If Methane would be nasty, Kerosene would be a nightmare, I don't think you'd be able to run the engine very long before it needed an overhaul.

Kerosene is even too dirty for normal rocket engines,

"RP-1 (alternately, Rocket Propellant-1 or Refined Petroleum-1) is a highly refined form of kerosene outwardly similar to jet fuel, used as rocket fuel.

...

Many early rockets had burned kerosene, but as burn times, combustion efficiencies, and combustion-chamber pressures grew, and as engine masses shrank, the engine temperatures became unmanageable. Raw kerosene used as coolant would dissociate and polymerize. Lightweight products in the form of gas bubbles, and heavy ones in the form of engine deposits, then blocked the narrow cooling passages. The coolant starvation raised temperatures further, accelerating breakdown. This cycle would escalate rapidly (i.e., thermal runaway would occur) until an engine wall ruptured.

This occurred even with the entire flow of kerosene used as coolant. Rocket designers turned to the fuel chemists to formulate a heat-resistant hydrocarbon. The specification was completed in the mid-1950s."

I don't think anyone would expose themselves to the trouble of trying to use a large hydrocarbon as a propellant for NTRs... unless they were stranded and desperate for propellant

[PB666]

51 minutes ago, KerikBalm said:

If Methane would be nasty, Kerosene would be a nightmare, I don't think you'd be able to run the engine very long before it needed an overhaul.

Kerosene is even too dirty for normal rocket engines,

"RP-1 (alternately, Rocket Propellant-1 or Refined Petroleum-1) is a highly refined form of kerosene outwardly similar to jet fuel, used as rocket fuel.

The figures they give for the NTR we for a 600 ISP engine temperature, the new engines are of course much hotter, You can run and ISP up to a 1000 in the NTR with a nozzle refit, you would just shorten its life.

The RP-1 would really clobber the NTR core with carbon, no doubt. Hydrogen is much much better because it undergoes plasma conversion at a much lower temperature, so you've got the sweetspot, and second and probably more important it only can bond with one other atom using its outer shell electron, so that the most it can do is coat the metal with a single atom thick layer of hydrogen. This is not as bad as it may look because in doing this it can inefficiently absorb neutrons and increase the surface heat relative to the core heat, carbon 12 on the other hand is not a big fan of neutrons.

I have never heard of anyone proposing an NTR as a recycling space tug, I suppose that is as good as use as any, but even at 600 ISP is has a rather short life.

EDIT: BTW I was using the average composition formula for RP-1 not kerosene, just using kerosene as a generic term to describe the fuel. RP1 is listed as a form of kerosene.

 

Edited May 26, 2016 by PB666

[KerikBalm]

"The figures they give for the NTR we for a 600 ISP engine temperature, the new engines are of course much hotter, You can run and ISP up to a 1000 in the NTR with a nozzle refit, you would just shorten its life."

Which figures are you talking about? Anyway, the ones relevant to my statement about temperatures: 3200K gets you over 800 Isp with H2... this is still cooler than the 3300K of the SSME combustion chamber

"The RP-1 would really clobber the NTR core with carbon, no doubt." Exactly... wouldn't be used

"Hydrogen is much much better because it undergoes plasma conversion at a much lower temperature, so you've got the sweetspot, and second and probably more important it only can bond with one other atom using its outer shell electron, so that the most it can do is coat the metal with a single atom thick layer of hydrogen. This is not as bad as it may look because in doing this it can inefficiently absorb neutrons and increase the surface heat relative to the core heat, carbon 12 on the other hand is not a big fan of neutrons."

Blah blah, yes, we know, just about everyone on this forum knows hydrogen is ideally what you run through a NTR.

"I have never heard of anyone proposing an NTR as a recycling space tug, I suppose that is as good as use as any"

I seem to recall it was part of some proposals for the STS... when the STS included reusable tugs (chemical and nuclear), a station, and a shuttle... of course they only went with the shuttle in the end... and that generally turned out terrible... although it was good as a jobs program I guess (ok ok... must not divert the topic to that again)

 

http://www.hq.nasa.gov/office/pao/History/taskgrp.html

"

The Nation now has the demonstrated capability to move on to new goals and new achievements in space in all of the areas pioneered during the decade of the sixties. In each area of space exploration what seemed impossible yesterday has become today's accomplishment. Our horizons and our competence have expanded to the point that we can consider unmanned missions to any region in our solar system; manned bases in earth orbit, lunar orbit or on the surface of the Moon; manned missions to Mars; space transportation systems that carry their payloads into orbit and then return and land as a conventional jet aircraft; reusable nuclear-powered rockets for space operations; remotely controlled roving science vehicles on the Moon or on Mars; and application of space capability to a variety of services of benefit to man here on earth.

Our opportunities are great and we have a broad spectrum of choices available to us. It remains only to chart the course and to set the pace of progress in this new dimension for man.

"

Edited May 26, 2016 by KerikBalm

[PB666]

4 minutes ago, KerikBalm said:

Blah blah, yes, we know, just about everyone on this forum knows hydrogen is ideally what you run through a NTR.

I was just correcting the facts about using other fuels, not promoting hydrogen. Personally I am not a fan of NTR, there is alot more fiction than fantasy concerning its potential use, and mostly as one limits it to liquid hydrogen or hydrogen gas, its not very useful compared to the alternatives, which do not carry political consequences.

 

[weissel]

On 25.5.2016 at 1:06 PM, Streetwind said:

Nope. Specific impulse is a super simple definition that can't really be changed. It says: when putting out 1 N of thrust, and having 1 kg of fuel available (or: putting out 1 kN, and having 1 metric ton of fuel), the engine will run for this long. That's why the unit is in seconds.

Force (N) == Force (kg * m/s²) = ISP (m/s) * mass flow (kg/s)

The proper, correct, right, sane, ... measurement for the specific impulse is the exit velocity.[1]  Meters per second.  What comes out.  Very intuitive, very simple. 

3000 m/s  /  9.81 m/s²  ~= 305.8s

If you use seconds, you are using propellant weight (i.e. mass / g0) as the basis, which is "funny", since the weight is zero in free fall and 4 times the mass at 4 gravities ... whereas the inert mass always stays the same.  Using g0 ~= 9.81 m/s² (the standard gravity on Earth at sea level), even when you are on the moon, on Eve's surface or wherever else.  Regularly confuses people terribly, just search the forums here.  "Wait, but the gravity (and gravity on surface) is completely different here, why is the gravity 9.81 m/s²??"

Of course that means we are stuck with seconds as the unit ...
 

[1] Effective exit velocity when you are in an atmosphere.
A high bypass turbojet engine will have a much much slower real exit velocity, as the engine will, instead of just pushing it's exhaust really fast, instead use much of the power to accelerate tons of cold air to a (in turn) much lower exit velocity (which is more effective for usual plane travel speeds).

[PB666]

 

5 hours ago, weissel said:

Force (N) == Force (kg * m/s²) = ISP (m/s) * mass flow (kg/s)

The proper, correct, right, sane, ... measurement for the specific impulse is the exit velocity.[1]  Meters per second.  What comes out.  Very intuitive, very simple. 

3000 m/s  /  9.81 m/s²  ~= 305.8s

If you use seconds, you are using propellant weight (i.e. mass / g0) as the basis, which is "funny", since the weight is zero in free fall and 4 times the mass at 4 gravities ... whereas the inert mass always stays the same.  Using g0 ~= 9.81 m/s² (the standard gravity on Earth at sea level), even when you are on the moon, on Eve's surface or wherever else.  Regularly confuses people terribly, just search the forums here.  "Wait, but the gravity (and gravity on surface) is completely different here, why is the gravity 9.81 m/s²??"

Of course that means we are stuck with seconds as the unit ...
 

[1] Effective exit velocity when you are in an atmosphere.
A high bypass turbojet engine will have a much much slower real exit velocity, as the engine will, instead of just pushing it's exhaust really fast, instead use much of the power to accelerate tons of cold air to a (in turn) much lower exit velocity (which is more effective for usual plane travel speeds).

Its probably a good idea not to use ISP once outside of LEO, unless you are going to express accelerations in g-forces. 

But then what do i know they expressed the growth of BEAM in inches. 

[SuperFastJellyfish]

1 hour ago, PB666 said:

But then what do i know they expressed the growth of BEAM in inches.

I'm from a suburb of Chicago, and that made me cringe.  Maybe they hoped that laypeople would logon to watch them inflate a useless module for 8 hours?  It's unbelievable, these days, to not use metric in scientific anything.

[Bill Phil]

23 minutes ago, SuperFastJellyfish said:

I'm from a suburb of Chicago, and that made me cringe.  Maybe they hoped that laypeople would logon to watch them inflate a useless module for 8 hours?  It's unbelievable, these days, to not use metric in scientific anything.

Tradition. Why do people anywhere still do what they did in the past? Tradition.

The UK has miles, albeit a different type. Germany has pounds, but a different type ( 1 german pound is half a kilogram). 

But here's the thing: the USA is a super power.

It was probably for the layman. But many scientists use imperial, and many use both. 

[PB666]

41 minutes ago, Bill Phil said:

Tradition. Why do people anywhere still do what they did in the past? Tradition.

The UK has miles, albeit a different type. Germany has pounds, but a different type ( 1 german pound is half a kilogram). 

But here's the thing: the USA is a super power.

It was probably for the layman. But many scientists use imperial, and many use both. 

Not in the lab, its stricky metric, never imperial, its just not done. 

[Bill Phil]

1 minute ago, PB666 said:

Not in the lab, its stricky metric, never imperial, its just not done. 

Labs only ever use measurements. I'm referring to calculations.

[PB666]

10 hours ago, Bill Phil said:

Labs only ever use measurements. I'm referring to calculations.

Measurements or calculations, doesn't matter. The biggest conversion question in the lab is cgs versus kms. There is the very ocaasional instance where I have to convert molarity to eqivilents at the atomic scale, very rarely. Ah yes and have to convert concentrations into LD50 values, ussually for other folks. The only time in recent memory i used an emperial measure is when we had to wax the floors in a heavily restricted area. 

[regex]

On 5/26/2016 at 5:55 AM, KerikBalm said:

ksp engines mostly have Isps corresponding to Kerosene- liquod O2 rockets.

KSP's engines are based on the theoretical maximums of Aerozine50/NTO engines.  I have this on good authority from the guy who did the recent rebalance.  Furthermore, the hsp of LF and O back that up in that they are the same as the hsp of Aerozine50 and NTO respectively.  KSP's fuels are also basically infinitely storable, like Aerozine50/NTO.  Because of KSP's nebulous "unit" volume you don't see those fuels matching their IRL counterparts, and I'm not entirely sure if that would have changed anyway due to it ruining existing craft in a major way.

Whether that matches Kerolox rockets in the main is purely coincidence.

While this may cause problems conceptually with air-breathing engines (not entirely sure how Aerozine50 would be handled in a turbine...) and NTRs (Aerozine50 is a fairly heavy fuel that would provide nowhere near 800s in an NTR AFAIK) it's an abstraction for gameplay.

E: To answer OP, yes, they are fairly realistic.  What isn't is the dry mass of parts but that's more because KSP's tiny, impossible solar system requires something we don't have IRL called "game balance".

Edited June 2, 2016 by regex