Author Topic: RP-1, methane, impulse density Q&A  (Read 95121 times)

Offline Proponent

  • Senior Member
  • *****
  • Posts: 7282
  • Liked: 2784
  • Likes Given: 1462
Re: RP-1, methane, impulse density
« Reply #40 on: 10/15/2014 10:25 am »
I'm still wondering why methane seems to be the clear favorite, when it's so much less dense than other hydrocarbons.  Even if ISRU methane is used as a fuel on Mars someday, that's some distance into the future and in the meantime an awful lot more stuff is going to be and will continue to be launched from Earth than from Mars.  Justifying methane over the others on this basis seems to be a case of the tail wagging the dog.

Methane, with its simple C-H bonds, probably is less subject to coking, but is coking really a significant problem with the others?  Surely there must be some information out there about coking, like reaction coefficients for polymerization as a function of temperature for the various fuels.  This would reduce the arm-waviness of the discussion.

It's not just the coking, while that is nice. Methane rich gas has a very high specific heat. This means that for a fixed turbine inlet temperature, you can get ungodly amounts of power out of it.

This is why methane, like hydrogen, optimizes at a fuel-rich preburner for staged combustion. The lack of coking just helps close the case.

Allows you either to have a very low turbine temp and get a moderate chamber pressure, which is good for reusability, or a very high chamber pressure for a typical turbine inlet temp (900-1200K), which is good for performance.

This has seemed to me to be the strongest argument for methane.  But I've been thinking about it a little more.

My earlier post containing heat capacities of light hydrocarbons shows that methane's is a bit higher than those of other light hydrocarbons.  Thus, at a given temperature, methane packs somewhat more thermal energy for running a turbopump.  That's obviously good.

But... that energy is used to pump propellants, and the power required by a pump depends on the volume rate that's pumped, not on the mass rate.  So, let's compute the heat capacity per unit volume of propellant (see the third attachment for the calculations).  The results are plotted below, with underlying data from the NIST Chemistry WebBook.  The first plot shows heat capacity of the fuel per unit volume of propellant for hydrogen at O/F=5.5, methane at 3.5, ethane at 3.2, ethylene (ethene) at 2.6, propane at 3.9, and propylene (propene) at 2.7.  This figure is meant to represent fuel-rich staged combustion.  The second plot is the same except that the heat capacity of oxygen is added in, corresponding to full-flow staged combustion, where the temperatures at the inlets of the two turbines are the same.

To make visual sense of the plots, note that deeply-cryogenic hydrogen is plotted in the coldest color, blue.  The colors for the hydrocarbons will make sense if you know the resistor color code; brown = 1 (carbon atom), red = 2, orange = 3.

In FRSC at 700 K, the hydrocarbon to beat is propane, with a heat capacity per unit volume of propellant of 740 kJ K-1 m-3.  Methane comes in about 10% lower at 670 kJ K-1 m-3.

Propane also comes out tops In FFSC at 700 K, with a heat capacity per unit volume of propellant of 1450 kJ K-1 m-3.  Methane at 1340 kJ K-1 m-3 is several percent lower and is the worst of hydrocarbons considered here.

Fold in methane's disadvantage in bulk density (830 kg/m3 vs. 920 kg/m3 for propane), and its few seconds' worth of Isp advantage over propane (and disadvantage in comparison to propylene) doesn't seem worth it, especially for a booster stage.

Since the dudes at SpaceX (FFSC) and Blue Origin (FRSC) are smart and know a lot more about rocket engines than I do, I'm sure there are good reasons for preferring methane over other light hydrocarbons, but it doesn't look to me like heat capacity is one of them.

EDIT:  Added missing 'r' in "NIST Chemistry WebBook"
« Last Edit: 01/05/2015 11:31 am by Proponent »

Offline Steven Pietrobon

  • Member
  • Senior Member
  • *****
  • Posts: 39304
  • Adelaide, Australia
    • Steven Pietrobon's Space Archive
  • Liked: 32874
  • Likes Given: 8357
Re: RP-1, methane, impulse density
« Reply #41 on: 10/24/2014 07:37 am »
Impulse density is a subject I've been interested in for quite a while. Here's a table showing the impulse density of various propellant combinations. The best is F2/NH3! The worst is O2/H2, with O2/CH4 at the low end compared to many other propellants.

MR = Oxidiser to fuel mixture ratio
dp = Propellant density (kilograms per litre)
ve = Effective exhaust speed (divide by g = 9.80665 m/s^2 to get Isp in seconds)
Id = Impulse density (Newton seconds per litre)

Propellants  MR   dp (kg/L)  ve (m/s) Id (Ns/L)
O2/H2        5.0   0.3251     4455     1448
O2/H2        6.0   0.3622     4444     1610
O2/H2        7.5   0.4120     4365     1798
O2/B2H6      2.0   0.7447     4056     3021
O2/NH3       1.4   0.8896     3399     3024
O2/CH4       3.6   0.8376     3656     3062
O2/CH4O      1.4   0.9640     3238     3121
O2/Atsetam   1.8   0.9041     3622     3275
O2/C2H6O     1.9   0.9928     3307     3283   
O2/C2H6      3.2   0.9252     3634     3362
O2/C3H8      3.1   0.9304     3613     3362
O2/C3H4      2.4   0.9666     3696     3573
O2/C3H6      2.7   0.9782     3681     3601
O2/RP–1      2.8   1.0307     3554     3663
O2/C3H4O     1.5   1.0572     3572     3776
O2/C10H16    2.6   1.0471     3608     3778
O2/C7H8      2.4   1.0954     3628     3974

N2O4/B2H6    1.3   0.7196     3859     2777
N2O4/NH3     2.0   1.0428     3097     3230
N2O4/UDMH    2.9   1.1823     3350     3961
N2O4/AZ50    2.2   1.1979     3366     4032
N2O4/MMH     2.4   1.2051     3366     4056
N2O4/N2H4    1.4   1.2156     3371     4097

N2O/C2H6O    5.7   1.1301     3042     3438
N2O/C2H6    10.1   1.1123     3117     3467
N2O/RP-1     9.2   1.1626     3099     3603

HNO3/RP-1    5.2   1.3162     3085     4060

HTP/H2      17.0   0.6925     3592     2487
HTP/B2H6     1.84  0.7947     3970     3156
HTP/NH3      3.1   1.1216     3068     3441
HTP/CH4      8.5   1.1440     3245     3712
HTP/C2H6     8.0   1.2245     3248     3978
HTP/C3H8     7.8   1.2284     3242     3982
HTP/C4H6     6.9   1.2274     3274     4019
HTP/C3H4     6.6   1.2573     3319     4173
HTP/C3H6     7.3   1.2694     3296     4184

HTP/CH4O     3.2   1.1968     3102     3712
HTP/C2H6O    4.5   1.2458     3151     3926
HTP/C3H8O    5.2   1.2625     3167     3998
HTP/N2H4     2.2   1.2608     3283     4139
HTP/RP–1     7.3   1.3059     3223     4209
HTP/C6H6     6.6   1.3201     3210     4238
HTP/C3H4O    4.2   1.3010     3283     4271
HTP/C10H16   7.1   1.3190     3264     4305
HTP/C7H8     6.6   1.3496     3287     4436

F2/H2       14.6   0.6553     4704     3083
F2/H2O       2.1   1.2942     2876     3722
F2/HTP       0.88  1.4689     2966     4357
F2/NH3       3.4   1.1770     4115     4843
F2/B2H6      6.4   1.1314     4416     4996
F2/N2H4      2.3   1.3073     4212     5506

AZ50 (50% UDMH and 50% N2H4 by mass)
HTP (98% H2O2 and 2% H2O by mass)
Atsetam (32% C2H2 and 68% NH3 by mass)

        MP     BP
H2    -259.1 -252.9 Hydrogen
O2    -218.3 -182.9 Oxygen
CH4   -182.5 -161.5 Methane
B2H6  -164.9  -92.5 Diborane
C2H6  -182.8  -88.6 Ethane
C2H2   -82.2  -75   Acetelyne
C3H8  -187.6  -42.1 Propane
NH3    -77.3  -33.3 Ammonia
C3H6  -129    -33   Cyclopropane
C3H4  -102.7  -23.2 Methylacetylene
C4H6  -109     -4.5 1,3-Butadiene
C4H6  -119      2   Cyclobutene  HF?
N2O4   -15     21.2 Nitrogen Tetroxide
CH4O   -98     64.7 Methanol
C2H8N2 -57     54   Unsymmetrical Dimethylhydrazine (UDMH)
C2H6O -114     78   Ethanol
C6H6     5.5   80   Benzene
C3H8O  -89.5   82   Isopropanol
C7H8   -27     88   Quadricyclane
CH6N2  -52     91   Monomethylhydrazine (MMH)
H2O      0.0  100.0 Water
N2H4     1    113.5 Hydrazine
C3H4O  -53    114.5 Propargyl Alcohol
CH1.95 -62    147   RP-1
C10H16        158   Syntin


Aerozine50 density = 0.8818 kg/L

Efficiency = 97.4%
Chamber Pressure = 20.7 MPa
Expansion Ratio = 77.5
« Last Edit: 10/24/2014 08:20 am by Steven Pietrobon »
Akin's Laws of Spacecraft Design #1:  Engineering is done with numbers.  Analysis without numbers is only an opinion.

Offline Steven Pietrobon

  • Member
  • Senior Member
  • *****
  • Posts: 39304
  • Adelaide, Australia
    • Steven Pietrobon's Space Archive
  • Liked: 32874
  • Likes Given: 8357
Re: RP-1, methane, impulse density
« Reply #42 on: 10/24/2014 08:20 am »
Here's a graph showing how poorly methalox performs. We plot delta-V versus the ratio of propellant volume to final mass. Two lowest curves are hydrolox with a nominal mixture ratio (MR) of 6 and an impractical one of 7.5 (8 is stoichiometric). The next worst is methalox. All the other combinations perform better.

Say for example you want your first stage to have a 4 km/s delta-V, about what you need to get to LEO for the first stage of a two vehicle with the same propellants. Hydrolox requires 4 litres of propellant for every kg of your total burnout mass (which includes the first stage dry mass, second stage and payload). Methalox requires 2.35 L/kg. Kerolox requires 2.0 L/kg. That is, your first stage needs 18% more propellant volume which corresponds to about 18% more propellant tank mass.
Akin's Laws of Spacecraft Design #1:  Engineering is done with numbers.  Analysis without numbers is only an opinion.

Offline Proponent

  • Senior Member
  • *****
  • Posts: 7282
  • Liked: 2784
  • Likes Given: 1462
Re: RP-1, methane, impulse density
« Reply #43 on: 10/24/2014 09:01 am »
I'm not sure I understand the rationale for using the ratio of propellant volume to burn-out mass as a metric.   If I'm using a bulky propellant combination like lox-hydrogen, I'm going to tend to have large, heavy tanks, which is bad.  But since the tank mass appears in the denominator, in some sense the combination is rewarded for being bulky.

EDIT: "I''m" -> "in" in final sentence.
« Last Edit: 10/25/2014 08:26 am by Proponent »

Offline R7

  • Propulsophile
  • Senior Member
  • *****
  • Posts: 2725
    • Don't worry.. we can still be fans of OSC and SNC
  • Liked: 992
  • Likes Given: 668
Re: RP-1, methane, impulse density
« Reply #44 on: 10/24/2014 10:51 am »
I'm not sure I understand the rationale for using the ratio of propellant volume to burn-out mass as a metric.   If I'm using a bulky propellant combination like lox-hydrogen, I'm going to tend to have large, heavy tanks, which is bad.  But since the tank mass appears in the denominator, I'm some sense the combination is rewarded for being bulky.

Smaller number -> smaller tanks -> smaller portion of the final mass is tank mass -> the better.
AD·ASTRA·ASTRORVM·GRATIA

Offline MP99

Re: RP-1, methane, impulse density
« Reply #45 on: 10/27/2014 08:18 am »


Here's a graph showing how poorly methalox performs. We plot delta-V versus the ratio of propellant volume to final mass. Two lowest curves are hydrolox with a nominal mixture ratio (MR) of 6 and an impractical one of 7.5 (8 is stoichiometric). The next worst is methalox. All the other combinations perform better.

Say for example you want your first stage to have a 4 km/s delta-V, about what you need to get to LEO for the first stage of a two vehicle with the same propellants. Hydrolox requires 4 litres of propellant for every kg of your total burnout mass (which includes the first stage dry mass, second stage and payload). Methalox requires 2.35 L/kg. Kerolox requires 2.0 L/kg. That is, your first stage needs 18% more propellant volume which corresponds to about 18% more propellant tank mass.

Musk has confirmed that his methalox will be sub-cooled close to freezing temps.

How does that affect the density and other properties (EG having to add more heat to reach the same combustion temps [impact to Isp?], reduced energy to pump a smaller volume, viscosity effects, extra energy required to autogenously pressurise)?

Cheers, Martin

Offline simonbp

  • Science Guy
  • Senior Member
  • *****
  • Posts: 7138
  • Liked: 314
  • Likes Given: 183
Re: RP-1, methane, impulse density
« Reply #46 on: 10/27/2014 12:13 pm »
In Steven's formulation, the best propellants that have ever flown extensively are hypergols...

However, I'm not sure that volume is the correct normalization here, as tank mass much more closely scales to surface area, which scales as volume^(2/3). If you apply that correction, the more exotic dense fuels will appear not much better than methane.

Offline Proponent

  • Senior Member
  • *****
  • Posts: 7282
  • Liked: 2784
  • Likes Given: 1462
Re: RP-1, methane, impulse density
« Reply #47 on: 10/27/2014 12:58 pm »
That's assuming constant wall thickness, but the mass of a pressure vessel scales as the product of pressure by volume for constant tensile strength.  This tends to suggest that volume normalised in some way is a valid metric.

As Whitehead points out (see page 8 of the attached paper), however, the mass of the tanks is by no means the whole story -- supporting structures can be quite significant too.

EDIT:  Added missing 'a' near end of second sentence.
« Last Edit: 10/28/2014 08:49 am by Proponent »

Offline baldusi

  • Senior Member
  • *****
  • Posts: 8356
  • Buenos Aires, Argentina
  • Liked: 2539
  • Likes Given: 8287
Re: RP-1, methane, impulse density
« Reply #48 on: 10/27/2014 03:49 pm »


Here's a graph showing how poorly methalox performs. We plot delta-V versus the ratio of propellant volume to final mass. Two lowest curves are hydrolox with a nominal mixture ratio (MR) of 6 and an impractical one of 7.5 (8 is stoichiometric). The next worst is methalox. All the other combinations perform better.

Say for example you want your first stage to have a 4 km/s delta-V, about what you need to get to LEO for the first stage of a two vehicle with the same propellants. Hydrolox requires 4 litres of propellant for every kg of your total burnout mass (which includes the first stage dry mass, second stage and payload). Methalox requires 2.35 L/kg. Kerolox requires 2.0 L/kg. That is, your first stage needs 18% more propellant volume which corresponds to about 18% more propellant tank mass.

Musk has confirmed that his methalox will be sub-cooled close to freezing temps.

How does that affect the density and other properties (EG having to add more heat to reach the same combustion temps [impact to Isp?], reduced energy to pump a smaller volume, viscosity effects, extra energy required to autogenously pressurise)?

Cheers, Martin

Just subcooling the CH4 (which you get "for free" with a small common bulkhead), gives less than 3% improvement in propellant mass for same volume. Doing full CH4@93K and LOX@68K is a little better 8.5%. It might not seem that much, but this is the rough performance improvement expected:

Densification\OrbitLEOGTO
LOX+CH4+15.00%+23.00%
CH4 Only+6.50%+10.50%

Which is quite interesting if you ask me. It is roughly like adding two solids to an EELV, for example. And almost like the RS-68 to RS-68A improvement on the Delta IV Heavy. Say that your rocket does 5.3 tonnes to GTO, full densification would bring it to 6.5 tonnes. And CH4 would allow 5.85 tonnes. So, for cases where you are a bit below your target performance, you could apply this and get an extra decade out of your design. Or save this as an option in design and have some margin for any other performance shortcoming that you might have.

Offline Hyperion5

  • Full Member
  • ****
  • Posts: 1681
  • Liked: 1373
  • Likes Given: 302
Re: RP-1, methane, impulse density
« Reply #49 on: 10/27/2014 04:22 pm »
In Steven's formulation, the best propellants that have ever flown extensively are hypergols...

However, I'm not sure that volume is the correct normalization here, as tank mass much more closely scales to surface area, which scales as volume^(2/3). If you apply that correction, the more exotic dense fuels will appear not much better than methane.

Hypergolics are only the best if you don't factor in their terrible toxicity.  If you're looking for a non-toxic, high impulse density mix, it's hard to beat RP-1/H2O2 (95%+ Hydrogen Peroxide).  That mix is denser than almost any hypergolic mix, though its not-so-great Isp prevents it from winning out on impulse density.  On both cost of propellants and of complications, "keroxide", as Lobo has dubbed it, should win out.  Only problem with the mix for Spacex is that it freezes at water-like temperatures and you can't produce RP-1 on Mars (at least, not for a long while).  Which brings them right back to CH4. 



Here's a graph showing how poorly methalox performs. We plot delta-V versus the ratio of propellant volume to final mass. Two lowest curves are hydrolox with a nominal mixture ratio (MR) of 6 and an impractical one of 7.5 (8 is stoichiometric). The next worst is methalox. All the other combinations perform better.

Say for example you want your first stage to have a 4 km/s delta-V, about what you need to get to LEO for the first stage of a two vehicle with the same propellants. Hydrolox requires 4 litres of propellant for every kg of your total burnout mass (which includes the first stage dry mass, second stage and payload). Methalox requires 2.35 L/kg. Kerolox requires 2.0 L/kg. That is, your first stage needs 18% more propellant volume which corresponds to about 18% more propellant tank mass.

Musk has confirmed that his methalox will be sub-cooled close to freezing temps.

How does that affect the density and other properties (EG having to add more heat to reach the same combustion temps [impact to Isp?], reduced energy to pump a smaller volume, viscosity effects, extra energy required to autogenously pressurise)?

Cheers, Martin

Just subcooling the CH4 (which you get "for free" with a small common bulkhead), gives less than 3% improvement in propellant mass for same volume. Doing full CH4@93K and LOX@68K is a little better 8.5%. It might not seem that much, but this is the rough performance improvement expected:

Densification\OrbitLEOGTO
LOX+CH4+15.00%+23.00%
CH4 Only+6.50%+10.50%

Which is quite interesting if you ask me. It is roughly like adding two solids to an EELV, for example. And almost like the RS-68 to RS-68A improvement on the Delta IV Heavy. Say that your rocket does 5.3 tonnes to GTO, full densification would bring it to 6.5 tonnes. And CH4 would allow 5.85 tonnes. So, for cases where you are a bit below your target performance, you could apply this and get an extra decade out of your design. Or save this as an option in design and have some margin for any other performance shortcoming that you might have.
 

I just want to be clear on this.  Are you shrinking the propellant tanks or using the extra propellant within the same volume tanks to enable more powerful engines to fling the rockets skyward? 


Offline strangequark

  • Full Member
  • ****
  • Posts: 1072
  • Co-Founder, Tesseract Space
  • San Francisco, CA
  • Liked: 226
  • Likes Given: 12
Re: RP-1, methane, impulse density
« Reply #50 on: 10/27/2014 08:39 pm »
I'm still wondering why methane seems to be the clear favorite, when it's so much less dense than other hydrocarbons.  Even if ISRU methane is used as a fuel on Mars someday, that's some distance into the future and in the meantime an awful lot more stuff is going to be and will continue to be launched from Earth than from Mars.  Justifying methane over the others on this basis seems to be a case of the tail wagging the dog.

Methane, with its simple C-H bonds, probably is less subject to coking, but is coking really a significant problem with the others?  Surely there must be some information out there about coking, like reaction coefficients for polymerization as a function of temperature for the various fuels.  This would reduce the arm-waviness of the discussion.

It's not just the coking, while that is nice. Methane rich gas has a very high specific heat. This means that for a fixed turbine inlet temperature, you can get ungodly amounts of power out of it.

This is why methane, like hydrogen, optimizes at a fuel-rich preburner for staged combustion. The lack of coking just helps close the case.

Allows you either to have a very low turbine temp and get a moderate chamber pressure, which is good for reusability, or a very high chamber pressure for a typical turbine inlet temp (900-1200K), which is good for performance.

This has seemed to me to be the strongest argument for methane.  But I've been thinking about it a little more.

My earlier post containing heat capacities of light hydrocarbons shows that methane's is a bit higher than those of other light hydrocarbons.  Thus, at a given temperature, methane packs somewhat more thermal energy for running a turbopump.  That's obviously good.

But... that energy is used to pump propellants, and the power required by a pump depends on the volume rate that's pumped, not on the mass rate.  So, let's compute the heat capacity per unit volume of propellant (see the third attachment for the calculations).  The results are plotted below, with underlying data from the NIST Chemisty WebBook.  The first plot shows heat capacity of the fuel per unit volume of propellant for hydrogen at O/F=5.5, methane at 3.5, ethane at 3.2, ethylene (ethene) at 2.6, propane at 3.9, and propylene (propene) at 2.7.  This figure is meant to represent fuel-rich staged combustion.  The second plot is the same except that the heat capacity of oxygen is added in, corresponding to full-flow staged combustion, where the temperatures at the inlets of the two turbines are the same.

To make visual sense of the plots, note that deeply-cryogenic hydrogen is plotted in the coldest color, blue.  The colors for the hydrocarbons will make sense if you know the resistor color code; brown = 1 (carbon atom), red = 2, orange = 3.

In FRSC at 700 K, the hydrocarbon to beat is propane, with a heat capacity per unit volume of propellant of 740 kJ K-1 m-3.  Methane comes in about 10% lower at 670 kJ K-1 m-3.

Propane also comes out tops In FFSC at 700 K, with a heat capacity per unit volume of propellant of 1450 kJ K-1 m-3.  Methane at 1340 kJ K-1 m-3 is several percent lower and is the worst of hydrocarbons considered here.

Fold in methane's disadvantage in bulk density (830 kg/m3 vs. 920 kg/m3 for propane), and its few seconds' worth of Isp advantage over propane (and disadvantage in comparison to propylene) doesn't seem worth it, especially for a booster stage.

Since the dudes at SpaceX (FFSC) and Blue Origin (FRSC) are smart and know a lot more about rocket engines than I do, I'm sure there are good reasons for preferring methane over other light hydrocarbons, but it doesn't look to me like heat capacity is one of them.

Ah, so what you forgot is that the specific heat for fuel rich combustion products using methane is what matters, rather than methane itself. You set a combustion temperature, based on what your turbine can handle, then determine the appropriate O/F ratio to reach that temperature. The specific heat advantage is much more pronounced for the combustion products, both due to the T vs. O/F curve for methane, and the uniquely high hydrogen content. As a result, the trade between methane vs other LHCs is stronger for staged combustion cycles than something like expander.

Offline Proponent

  • Senior Member
  • *****
  • Posts: 7282
  • Liked: 2784
  • Likes Given: 1462
Re: RP-1, methane, impulse density
« Reply #51 on: 10/27/2014 10:30 pm »
I assumed the output of the fuel-rich preburner would be a lot like the fuel, since only a small amount of the fuel is burned in the preburner (that has to be the case, otherwise staged combustion would not be efficient).

Offline R7

  • Propulsophile
  • Senior Member
  • *****
  • Posts: 2725
    • Don't worry.. we can still be fans of OSC and SNC
  • Liked: 992
  • Likes Given: 668
Re: RP-1, methane, impulse density
« Reply #52 on: 10/30/2014 12:54 pm »
I assumed the output of the fuel-rich preburner would be a lot like the fuel, since only a small amount of the fuel is burned in the preburner (that has to be the case, otherwise staged combustion would not be efficient).

It is but incomplete combustion means the rest is quite messy soup. Combustion of methane happens in numerous steps and it is a lottery how far each methane molecule gets in the intermediate steps before there's no more oxygen to drive the reaction to final products, water and CO2.

Attaching a capture what species RPA thinks a notional methane fuel-rich preburner contains (800K at injector, 100bar pressure, small 1.5 contraction, miniscule 1.0001 expansion because AIUI traditional turbomachinery does not operate with clearly supersonic flow). Substantial amount of free carbon and hydrogen.

Compare to similar but oxidizer rich case, very clean.
AD·ASTRA·ASTRORVM·GRATIA

Offline Proponent

  • Senior Member
  • *****
  • Posts: 7282
  • Liked: 2784
  • Likes Given: 1462
Re: RP-1, methane, impulse density
« Reply #53 on: 01/05/2015 12:04 pm »
I assumed the output of the fuel-rich preburner would be a lot like the fuel, since only a small amount of the fuel is burned in the preburner (that has to be the case, otherwise staged combustion would not be efficient).

It is but incomplete combustion means the rest is quite messy soup. Combustion of methane happens in numerous steps and it is a lottery how far each methane molecule gets in the intermediate steps before there's no more oxygen to drive the reaction to final products, water and CO2.

Indeed, the RPA results you provide demonstrate that I was naive in assuming that the exhaust of a fuel-rich preburner would be essentially the fuel.  In my naive analysis, I pointed out that while methane's heat capacity of  3.99 kJ K-1 kg-1 at, e.g., 800 K is 13% higher than propane's (3.51 kJ K-1 kg-1), the greater volume of lox-methane propellant to be pumped results in a net advantage for propane.

So let's compare specific heats of fuel-rich preburner exhausts for lox-methane and lox-propane.  Repeating your calculation for lox-methane at 10 MPa and 800 K (attached), I too get a specific heat of 4.90 kJ K-1 kg-1.  A similar calculation (also attached) for lox-propane (O/F = 0.085) gives a specific heat of 4.40 kJ K-1 kg-1.  Now methane's advantage in specific heat is just 11% -- less than in the naive case.  In other words, considering more realistic preburner chemistry appears only to increase propane's advantage over methane from a thermodynamic viewpoint.

On the other hand, the output of the lox-propane preburner appears to be 30% solid carbon by mass.  Does anybody really want to run a turbine with that much solid carbon, which would then be fed into the main combustion chamber?  Sounds like a recipe for disaster to me.

So, that might seem to be a good reason to prefer methane to propane for staged combustion.  But ... have a look at the methane exhaust products: 8% solid carbon.  That's better than 30%, but it  still sounds bad.

The bottom line is that while I'm sure SpaceX and BO have good reasons for choosing methane for their staged-combustion engines, it's still not obvious to me that heat capacity is what makes methane superior to other light hydrocarbons.

EDIT: "I" -> "it" in penultimate sentence.
« Last Edit: 08/24/2015 06:01 am by Proponent »

Offline Steven Pietrobon

  • Member
  • Senior Member
  • *****
  • Posts: 39304
  • Adelaide, Australia
    • Steven Pietrobon's Space Archive
  • Liked: 32874
  • Likes Given: 8357
Re: RP-1, methane, impulse density
« Reply #54 on: 06/08/2015 08:57 am »
I found an interesting company called Synfuels International that is able to make Ethylene (C2H4), gasoline and jet fuel from alcohol. When I ran ethylene with the USAF Isp code, I got the following results.

Propellants  MR   dp (kg/L)  ve (m/s) Id (Ns/L)
O2/CH4       3.6   0.8376     3656     3062
O2/C2H4      2.7   0.9007     3678     3313
Efficiency = 97.4%
Chamber Pressure = 20.7 MPa
Expansion Ratio = 77.5

        MP     BP
O2    -218.3 -182.9 Oxygen
CH4   -182.5 -161.5 Methane
C2H4  -169.2 -103.7 Ethylene


Compare to methalox, ethylox (liquid ethylene and liquid oxygen) has a density that is 7.5% greater and an Isp that is 0.6% greater, giving an impulse density that is 8.2% greater!

On the way to make their jet fuel they make a butene (C4H8)/hexene (C6/H12) mixture. Unfortunately, I don't know what the density and heat of formation of this combination is, so I can't determine its performance.

Their jet fuel has a density of 0.803 kg/L. The paper below examined p-cymene (C10H14) and pinane (C10H18) which both have a density of 0.86 kg/L! That could be a great rocket fuel if it gives a good Isp. The freezing point of both chemicals is less than -70 C.

If anyone can help with heat of formations, carbon to hydrogen ratios and densities of these chemicals or mixtures, that would be much appreciated. Thanks.

Attached are some slides and a paper on their interesting brews.
« Last Edit: 06/08/2015 10:05 am by Steven Pietrobon »
Akin's Laws of Spacecraft Design #1:  Engineering is done with numbers.  Analysis without numbers is only an opinion.

Offline Steven Pietrobon

  • Member
  • Senior Member
  • *****
  • Posts: 39304
  • Adelaide, Australia
    • Steven Pietrobon's Space Archive
  • Liked: 32874
  • Likes Given: 8357
Re: RP-1, methane, impulse density
« Reply #55 on: 06/08/2015 10:32 am »
I found a heat of formation -174.7 kJ/mol for Pinane (C10H18) on the NIST website. I get the following results

Propellants  MR   dp (kg/L)  ve (m/s) Id (Ns/L)
O2/CH4       3.6   0.8376     3656     3062
O2/C2H4      2.7   0.9007     3678     3313
O2/RP–1      2.8   1.0307     3554     3663
O2/C10H16    2.6   1.0471     3608     3778
O2/C10H18    2.7   1.0531     3599     3790


C10H16 is Syntin, formerly used by Russia in their Soyuz vehicles. Compared to kerolox, Pinane/LOX has a 1.2% better Isp and 2.2% better density, giving a 3.5% better impulse density.
Akin's Laws of Spacecraft Design #1:  Engineering is done with numbers.  Analysis without numbers is only an opinion.

Offline philw1776

  • Full Member
  • ****
  • Posts: 1836
  • Seacoast NH
  • Liked: 1842
  • Likes Given: 984
Re: RP-1, methane, impulse density
« Reply #56 on: 06/08/2015 01:22 pm »
A question follows.  Could a vehicle with these higher impulse fuels be launched from Earth benefiting from the higher impulse efficiency and be in situ refueled on Mars using old fashioned methane? The same engine with two different fuels.

Obviously I'm neither a chemist or a rocket engineer.
FULL SEND!!!!

Offline Impaler

  • Full Member
  • ****
  • Posts: 1283
  • South Hill, Virgina
  • Liked: 372
  • Likes Given: 0
Re: RP-1, methane, impulse density
« Reply #57 on: 06/09/2015 03:22 am »
Steven Pietrobon:  I think Zubrin himself has fully acknowledged that Ethylene is completely superior to Methane and if he had the whole thing to do over again he would have pushed that instead as it's synthesis is almost as easy as methane, higher hydrocarbons not so much.

Lower hydrogen needs for Ethylene and easier refrigeration (practically none on Mars) are considered even more important then the density and impulse values.  The only reason to go for Methane now is that fact that everyone is developing LNG based engines for launch vehicles now and you could reuse thouse engines on Mars, but even then I suspect a dual fuel engine would be possible and advantageous.

Offline Steven Pietrobon

  • Member
  • Senior Member
  • *****
  • Posts: 39304
  • Adelaide, Australia
    • Steven Pietrobon's Space Archive
  • Liked: 32874
  • Likes Given: 8357
Re: RP-1, methane, impulse density
« Reply #58 on: 06/09/2015 05:56 am »
A question follows.  Could a vehicle with these higher impulse fuels be launched from Earth benefiting from the higher impulse efficiency and be in situ refueled on Mars using old fashioned methane? The same engine with two different fuels.

The differences in mass and volume flow rates would be such that is probably not practical. The engine has to be designed for the particular fuel used.
Akin's Laws of Spacecraft Design #1:  Engineering is done with numbers.  Analysis without numbers is only an opinion.

Offline Burninate

  • Full Member
  • ****
  • Posts: 1145
  • Liked: 360
  • Likes Given: 74
Re: RP-1, methane, impulse density
« Reply #59 on: 06/09/2015 09:14 am »
Steven Pietrobon:  I think Zubrin himself has fully acknowledged that Ethylene is completely superior to Methane and if he had the whole thing to do over again he would have pushed that instead as it's synthesis is almost as easy as methane, higher hydrocarbons not so much.

Lower hydrogen needs for Ethylene and easier refrigeration (practically none on Mars) are considered even more important then the density and impulse values.  The only reason to go for Methane now is that fact that everyone is developing LNG based engines for launch vehicles now and you could reuse thouse engines on Mars, but even then I suspect a dual fuel engine would be possible and advantageous.

Refrigeration advantages are moot when sharing a thermal environment with LOx.  Ethylene is a moderate problem there, because to maintain it in liquid phase at the same temperature you would need to raise LOx tank pressure to 5+ atmospheres (ethylene freezes at the boiling point of LOx at about 3.5atm). That's manageable, but adds weight.  Zubrin  is currently working on ethylene-N2O green hypergolics - http://www.parabolicarc.com/2015/05/15/pioneer-astronautics/
« Last Edit: 06/09/2015 09:23 am by Burninate »

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
1