Fluorine

[Satanic Mechanic]

Wow!  Look at the flow rate and thrust of methane vs. ammonia vs. water.  A flow rate lower than ammonia and a thrust a bit lower than water but still high.  That is impressive.
Also, methane can be created from CO₂ and water.  You could make the fuel on Mars for refueling.

SM

[Bob B.]

With liquid methane there are still some storability and density issues, but it is way better than hydrogen.  And its superior performance over ammonia seems to make it a very good compromise propellant.  And as you say, that fact it can be manufactured from in situ resources is a big bonus.

With a nuclear thermal rocket, performance will largely be linked to the molecular weight of the exhaust products.  One of the things that makes methane so much better than ammonia is that the carbon will exist in the exhaust in its elemental form, while nitrogen wants to combine into its molecular form.  Here are the propellants we've looked at so far (dissociation not included):

Fuel --> Product(s)      Avg. Molecular Wt.
   
  H2 --> H2                   2/1 = 2
  He --> He                   4/1 = 4
 CH4 --> C + 2 H2            16/3 = 5.33
B5H9 --> 5 B + 4.5 H2      63/9.5 = 6.63
 NH3 --> 0.5 N2 + 1.5 H2     17/2 = 8.5
N2H4 --> N2 + 2 H2           32/3 = 10.67
 H2O --> H2O                 18/1 = 18 

The exception to the "low molecular weight equals best performance" rule is helium, which is penalized by its very high specific heat ratio.  Note that I added hydrazine (N₂H₄) to the list even though I've done no calculations for it.  From this quick check, however, we can see that it will perform worse than most other candidates.  Are there any other compounds you can think of that might be worth a look?

I'm still mad at myself for messing up the methane thing.  The problem was that I was using the wrong thermodynamic data for carbon (gaseous instead of solid), so the methane wasn't dissociating like it should have when I ran the calculations.  As soon as I put in the correct data, the methane dissociated at a much lower temperature, drastically changing the results.  When using methane in a NTR, the carbon will exhaust as a solid ash.

[Satanic Mechanic]

 Are there any other compounds you can think of that might be worth a look?

I read the NTR white paper again and noticed that he wrote octane under propellants.  I wonder why he mentioned that over the other alkanes (butane, propane, etc...). 
I think methane is a good choice with the numbers you have shown.  It can be produced in areas with carbon dioxide and water and storage is not a hassle.

SM

[Satanic Mechanic]

How about other gases?  I know nitrogen is mentioned, I first thought of argon because I think it was used as a moderator for an experimental reactor.  While trying to find information on it, I found Argon fluorohydride but it decomposes quick above -256C.

I know, we are starting to get into the noble gases and it leads out of the solid core NTR and getting into the gaseous core and NEP designs.

SM

[Bob B.]

I read the NTR white paper again and noticed that he wrote octane under propellants.  I wonder why he mentioned that over the other alkanes (butane, propane, etc...).

Hum... octane kind of surprises me.  The better the hydrogen/carbon ratio the better the propellant.  This ratio is best with methane and gets worse as we move up to the heavier hydrocarbons.  Like you, I'd consider propane or butane first.  In fact, I was going to suggest that propane might be worth a look because of its better density and boiling point.  I'll have to go through a complete set of calculations to be sure, but I think propane will likely produce a specific impulse near 600 seconds.

Something else that looks a little intriguing is ethanol, not because its molecular weight is all that great but because the carbon will partially oxidize to generate some heat.  It might produce a 400-second ISP with a much smaller reactor.

C₂H₅OH --> CO + C + 3 H₂

I'll have to investigate some of these other options when I get home.

By the way, I got your PM but I haven't had a chance to read any of the papers yet.

[Bob B.]

I've decided to take a little different approach to compare these propellants.  My previous calculations assumed a single engine design through which different propellants were run.  That method didn't always give us the best comparison.  Here I've designed an engine for each propellant based on a specified performance.  The standard criterion used for all engines include (1) 500 kN thrust, (2) 600 s burn time, (3) 2,500 K chamber temperature, (4) 50 atm chamber pressure, and (5) 0.05 atm nozzle exit pressure.  Everything else was calculated from these standard specs.  Below I've tabulated the propellants side-by-side for easy comparison (listed in order of ISP).

Propellant                            H2      CH4     C3H8     B5H9    C2H8N2      He    C2H5OH      NH3      H2O    LOX-H2
------------------------------------------------------------------------------------------------------------------------- 
Propellant liquid range, K         14-20   91-112   86-231  226-333  215-337      1-4  159-351  195-240  273-373   varies
Propellant density, kg/l           0.071    0.423    0.582    0.618    0.791    0.145    0.789    0.682    1.000    0.338
Chamber temperature, K             2,500    2,500    2,500    2,500    2,500    2,500    2,500    2,500    2,500    3,388
Chamber pressure, atm                 50       50       50       50       50       50       50       50       50       50
Nozzle exit pressure, atm           0.05     0.05     0.05     0.05     0.05     0.05     0.05     0.05     0.05     0.05
Nozzle throat diameter, cm          26.4     24.2     23.9     24.8     24.4     28.1     24.9     26.4     25.4     25.6
Nozzle exit diameter, cm           188.8    252.7    262.0    237.5    247.2    129.2    235.3    189.5    220.7    214.9
Expansion ratio                     51.1    108.7    120.2     91.9    102.4     21.2     89.6     51.6     75.8     70.7
Exhaust molecular weight, g/mol    2.012    5.389    6.341    6.653    8.629    4.003    9.203    8.504   17.869   12.610
Specific heat ratio                 1.30     1.09     1.07     1.13     1.11     1.67     1.14     1.30     1.18     1.20
Thrust (vacuum), kN                  500      500      500      500      500      500      500      500      500      500
Specific impulse (vacuum), s         885      684      654      581      529      510      490      431      333      452
Reactor power, MW                  2,279    1,293      965    1,135      648    1,295    1,138    1,496    1,251      N/A
Engine flow rate, kg/s              57.6     74.6     78.0     87.8     96.4    100.0    104.1    118.2    153.1    112.9
Turbine flow rate, kg/s             1.58     1.84     1.89     1.63     1.67     3.08     1.59     1.36     2.42     1.25
Effective specific impulse, s        861      667      638      570      520      494      482      426      328      447
Pump power (75 atm, 75%), kW       8,451    1,830    1,391    1,466    1,256    7,205    1,357    1,777    1,576    3,423
Turbine Power (67%), kW           12,676    2,745    2,086    2,199    1,884   10,808    2,036    2,665    2,364    5,134
Burn time, s                         600      600      600      600      600      600      600      600      600      600
Propellant mass, kg               35,529   45,838   47,935   53,651   58,815   61,865   63,414   71,745   93,327   68,502
Propellant volume, l             500,406  108,364   82,363   86,814   74,356  426,656   80,372  105,197   93,327  202,688

H₂ - hydrogen, CH₄ - methane, C₃H₈ - propane, B₅H₉ - pentaborane, C₂H₈N₂ - UDMH, He - helium, C₂H₅OH - ethanol, NH₃ - ammonia, H₂O - water

(Edit #1)  Added "pump power" to the list - 75 atm pressure rise and 75% efficiency is assumed.

(Edit #2)  Added "turbine power" and "turbine flow rate".  Turbine efficiency assumed 66.67% and turbine mass flow rate calculated assuming 1,000 K inlet temperature, 50 atm inlet pressure, and 3 atm exit pressure (16.67 pressure ratio).

(Edit #3)  Added "effective specific impulse" - calculated on total propellant flow to both engine chamber and turbine.

(Edit #4)  Added LOX-LH₂ chemical engine for comparison - Mixture ratios:  5.5 combustion chamber, 1.0089 gas generator, 5.345 total.

[ijuin]

Given that methane also has a density of 0.42 kg/l compared to H2's 0.07 kg/ml, it may be the preferred propellant for situations in which the low temperature or high bulk of H2 is problematic (e.g. if you need to store it for months of cruise time until you reach the destination planet). I note from your table that in your equal-impulse design, methane has the second-lowest propellant mass requirement as well as being comparable in bulk to water or heavy-hydrocarbon (kerosene, etc) propellant. Also, H2 apparently requires about 80% more reactor energy, so by using methane you can get by with a reactor half as powerful.

[Bob B.]

Given that methane also has a density of 0.42 kg/l compared to H2's 0.07 kg/ml, it may be the preferred propellant for situations in which the low temperature or high bulk of H2 is problematic (e.g. if you need to store it for months of cruise time until you reach the destination planet). I note from your table that in your equal-impulse design, methane has the second-lowest propellant mass requirement as well as being comparable in bulk to water or heavy-hydrocarbon (kerosene, etc) propellant. Also, H2 apparently requires about 80% more reactor energy, so by using methane you can get by with a reactor half as powerful.

I also think propane looks pretty good.  It has a better density and boiling point than methane and requires even less reactor power, though it does have a lower specific impulse.

I think the best propellant will depend on the specific application.  In the example presented in my last post, I think the heavy inert mass of the hydrogen system (big tanks and reactor) may make the methane or propane system the better alternative.  However, if we had a system that wasn't going to burn for 600 seconds, but rather for 3,000 seconds, the situation may change.  In this case we might be able to afford the heavier inert mass of the hydrogen system because the savings in propellant mass would more than make up for it.  We really need to consider the delta-v to see which system is best for a given application.

Maybe my next step is to estimate the mass of each system (tanks, reactor, pump machinery, etc.) and calculate the delta-v with some dummy payload.  I could vary the burn duration to see where the breaking points are between swinging the advantage from one propellant to another. 

For future analyses, I think I'll drop pentaborane, helium, and water from the list.  Pentaborane seems to have some pretty good properties and performance, but we have concerns about boron's neutron absorption property.  Helium just doesn't give the performance to make its frigid cryogenic temperature and low density worth coping with.  And water's performance is too poor to consider it a primary propellant, though its use as a secondary propellant, as previously discussed, may still be viable.

[Bob B.]

Does anybody know what method was used to drive the turbopumps on the NERVA-style engines?  I can see three options:

(1)  Draw off some of the engine exhaust and duct it to the turbines – though the temperature (~2500 K) may be too high for the turbines to handle.

(2)  Use a separate heat exchanger that draws heat from either the engine exhaust or reactor to vaporize a small amount of propellant to drive the turbines.

(3)  Use a separate propellant, such as hydrogen peroxide, to generate gas to drive the turbines.  This is simple and effective but adds some weight.

I think option #2 would work very well for hydrogen or ammonia, but I’m concerned about using it with methane or propane.  Would the large amount of solid carbon in the exhaust foul the turbines?  I know that some kerosene burning engines have been able to overcome this issue, but they generally burn for only a few minutes (it’s my understanding the turbines get pretty fouled up in that short time).  I think a hydrocarbon NTR probably needs to use option #3.  In either case, we probably need to add a couple percent to the propellant mass to account for that needed to run the pumps.

[Satanic Mechanic]

Bob,
All the designs I have seen for a NTR used option #2 to run the turbopumps.  To answer your question about the turbines getting fouled using the hydrocarbons, I do not see that happening for methane or propane over that short amount of time and that propellant is a simple alkane.  I can see some problems if we were to go to a heavy alkane, like octane or higher, over the long term.
Something else to consider is the propellant will be be heated and will be broken apart cleanly.  I really want to use the analogy of the cleaness of the engine of propane-powered vehicles, but that is an internal combustion engine and not a NTR.  Off-topic.
Another thought that came to mind, but I will have to research this:  Does it take less energy to drive a turbopump sending hydrogen propellant than methane propellant?  I would think that hydrogen would be easier to pump but we would have to take a look at factors like flow-rate, temperature and few more that my M.E. friends (Fluids class!) would point out.

SM

[Bob B.]

Another thought that came to mind, but I will have to research this:  Does it take less energy to drive a turbopump sending hydrogen propellant than methane propellant?  I would think that hydrogen would be easier to pump but we would have to take a look at factors like flow-rate, temperature and few more that my M.E. friends (Fluids class!) would point out.

Pump power is the volumetric flow rate times the change in pressure divided by the pump efficiency.

Power = Q x delta-P / efficiency

where P is in watts, Q is in m³/s, and delta-P in is pascals.

Because the density of hydrogen is so low, we have to pump a very large volume of it.  This means it takes more pump power than the denser propellants. 

I think I'll add pump power to my previous table.

[Bob B.]

To answer your question about the turbines getting fouled using the hydrocarbons, I do not see that happening for methane or propane over that short amount of time and that propellant is a simple alkane.  I can see some problems if we were to go to a heavy alkane, like octane or higher, over the long term.

I would agree with you if we were burning the fuel because the carbon would oxidize into gaseous CO and CO₂, leaving only a trace of solid carbon.  But in a NTR we're simply decomposing the fuel, so virtually all the carbon will end up in the exhaust as solid carbon.  In the case of methane, 1/3 of the exhaust (by mole) is carbon.  I think this could be a problem in a turbine, and maybe even in the engine itself.  I just don't have any experience to tell me how much of a soot buildup there will be.

[Satanic Mechanic]

I put in the search engine for turbopumps and NERVA and came up with this website: http://klabs.org/history/ntrs_docs/other/nuclear/
Take a look at the article about Feed Systems and Rotational Machinery.

SM

[ijuin]

But in a NTR we're simply decomposing the fuel, so virtually all the carbon will end up in the exhaust as solid carbon.  In the case of methane, 1/3 of the exhaust (by mole) is carbon.  I think this could be a problem in a turbine, and maybe even in the engine itself.  I just don't have any experience to tell me how much of a soot buildup there will be.

Would flushing the reactor with a bit of oxygen at the end of a burn remove enough of this soot to be worth the logistics of adding the capacity to do so? If not, then how hot can we reasonably run the reactor in order to keep the buildup down?

[Satanic Mechanic]

Would flushing the reactor with a bit of oxygen at the end of a burn remove enough of this soot to be worth the logistics of adding the capacity to do so? If not, then how hot can we reasonably run the reactor in order to keep the buildup down?

Oxygen is corrosive to the fuel assembly.  The water used in reactors uses deionized water to reduce the corrosion.  My first thought was nitrogen but that is very inert.  Would it be enough to blast it with high pressure nitrogen to clean it or use another gas.
I think the carbon buildup would be small during the mission's duration.  Would there be a way to clean it after a mission for reuse?  I hope the NERVA would be re-used and not discarded after the mission like some of the scenarios I have read.

SM