A few days ago I came up with an idea for a hopefully fairly simple upgrade to a Raptor engine to run on a mix of methane and kerosene. It keeps the methane turbopump running on a full-flow cycle, though with a lower flow rate, and adds a shaft extension to run a separate kerosene pump. By powering the fuel pump(s) with methane, fuel-rich preburner combustion doesn’t cause problems like it would with kerosene. Taken to the extreme, including changing the engine coolant from methane to kerosene, as on the Merlin, it would end up being a high-efficiency staged combustion engine with a methane powered kerosene pump.
This would allow an increase in the first stage propellant mass by 15 to 20%, with an almost corresponding increase in payload to the same staging velocity, without increasing the external size of the stage. The advantage gets back to why SpaceX decided on RP-1 for the Falcon 9 first stage. For a given chamber pressure and engine cycle, it offers more capability in the same size vehicle. Since Super Heavy is Earth based and won’t be launched from Mars, it doesn’t really need to rely on just methane.
It’s should be a fairly cheap and simple modification which doesn’t touch the LOX pump and adds the kerosene pump to an extension of the methane pump’s upper shaft, with the methane infeed line teeing in between the two pumps.
Sounds like added complexity needing another stage of control of combustion. Single fuel with single oxidizer usually gives linear thrust variation with added flow. Going to a second-order equation of thrust and flow gets to a point of complexity not worth the effort. Your example gives a secondary fuel supply directly proportional to the primary flow. Both fuels only give best burn at one ratio. Your two-stage pump has no variation available to compensate for the varying combustion conditions. A series-fed two-turbine, two-compressor pump might be needed to make it work. There is currently a lot more known and tested on single-fuel single-oxidizer turbo pump rocket engines. Dual-fuel engines have many more unstable possibilites than the single-fuel models.
Are there any models available for this combination of fuels you propose?
The Russians have suggested using dual-fuel engines before, but I don’t know if they’ve ever built a prototype. I think they were looking at RP-1/LH2, and the idea was that instead of using two stages, such as the common RP-1 first stage and an LH2 second stage, you just shift fuels, saving the weight of an engine and an interstage. But that’s a much bigger jump than I’m suggesting.
The Raptor is already a dual pump, dual turbine engine. The LOX pump and turbine is a full flow design, with all the LOX flowing through the pump, preburner, and turbine. But unlike the NK-33, RD-170, and RD-180 engines, the Raptor doesn’t use its oxygen rich preburner and turbine to drive the fuel pump. By only driving the LOX pump, the load on the Raptor’s LOX turbine is only about 55% of what it would take to drive both pumps, which should result in a less fuel consumed in the pre-burner for the large but fixed amount of oxygen, and thus lower turbine inlet temperatures.
The separate methane turbopump on the Raptor is also full flow, so all the methane flows through pump, preburner, and turbine, mirroring the oxidizer pump, with a small amount of oxygen burned in its preburner. Nobody tries using a kerosene rich pre-burner for staged combustion because the exhaust is horribly sooty, so US gas-generator engines like the Merlin just dump it overboard, and that’s a considerable waste of fuel and a big performance hit. But a methane rich preburner is fine because the exhaust is clean.
Most of the Raptor’s ISP increase over the Merlin comes from using a high-chamber pressure fed by a full-flow staged combustion cycle, not from the switch from kerosene to methane.
The Raptor has a 4,400 psi chamber pressure, compared to Merlin’s 1,400 psi. Along with full-flow staged combustion that drives the pumps to deliver that pressure, this results in a reported ISP increase from the Merlin’s 282 second at SL to the Raptor’s 320 seconds at SL, and the Merlin’s 311 seconds vacuum to the Raptor’s 350 seconds vacuum. (Reported Raptor values might not be completely accurate yet). That’s about a 12 to 13 percent increase in ISP, whereas a simple switch to methane should only provide about a 3% increase. Nobody used methane because nobody thought the slight increase in ISP sufficiently offset the larger tanks for the lower density fuel.
The full-flow staged combustion cycle is a big win, and already working on the Raptor, which will be the first engine in history to actually fly with that cycle.
So what I’m suggesting is: Don’t touch the LOX pump. It’s fine.
Change the methane pump to provide roughly the same power output from a reduced methane flow rate, since we’re supplementing methane with kerosene in the combustion chamber. The methane turbopump will still be running fuel rich, and thermal limits and engine cooling requirements will determine how little methane can be burned (and how much kerosene can be substituted for methane in the combustion chamber) before the fuel pump’s turbine inlet temperatures or throat wall temperatures get too high.
Next, add a regular kerosene pump to the upper end of the shaft on the fuel turbopump, teeing in the methane infeed line to get around the shaft extension. Special seals to separate the kerosene from the methane aren’t critical because both fuels are hydrocarbons, so there’s no risk of combustion in the pumps. The kerosene doesn’t go anywhere but from the kerosene tank, through the kerosene pump, and straight to the injectors.
The RP-1/methane mixture ratio is going to be relatively fixed, which results in a new Oxygen/Fuel ratio that’s simple to calculate ratiometrically, because all we’re doing is burning a different cut of hydrocarbons. RP-1 was already a complex set of hydrocarbons, and the O:F ratio is just the average of all the different O:F ratios it would take to burn RP-1 as individual fractions.
What my idea doesn’t offer is a way to shift that ratio so it could be optimized like the Russian concept of switching fuels during different stages of flight. Having both fuel pumps on one shaft means it’s stuck with the mix it starts with.
However, if the methane flow is allowed to start a bit before the RP-1 flow, ignition should be a whole lot easier, and restart might be simple too.
Anyway, that’s what I’m thinking of. I’m not sure it’s worth it, but it would let SpaceX boost the pad weight and fuel fraction by a significant amount, which seems to be something they love doing.
Besides the proposal for bi-propellant of tri-propellant engines the Russians supposedly ran several tests with existing engines to prove the idea. This mostly happened in the early 1990s I think.
One thing they did was they ran the RD-0120 engine of the Energia with LOX/Methane, similar to the US tests with the RL-10, proving it could be used with either LH2 or Methane.
Besides changing the propellant, another thing which they looked at was engines with variable mixture ratios. Particularly in LOX/LH2 engines, LOX has a density close to 1.140 g/cc (similar to Kerosene at 1.02 g/cc), while LH2 has a density of 0.071 g/cc. So basically the more oxidizer rich the mixture is in a LOX/LH2 vehicle the better the overall propellant density will be. The downside with running oxidizer rich is that you decrease Isp, but if you run the engine on LOX/Kerosene you also decrease Isp. Only having two propellants simplifies the plumbing and tank structure. Something which is something typically underestimated as a complexity problem with rockets in general.
I have simplified my idea. The Raptor would probably run fine on alcohol almost as is, since alcohol doesn’t burn with a sooty residue, and the performance hit from a slightly lower ISP would be more than offset by the increased propellant density.
A few days ago I came up with an idea for a hopefully fairly simple upgrade to a Raptor engine to run on a mix of methane and kerosene. It keeps the methane turbopump running on a full-flow cycle, though with a lower flow rate, and adds a shaft extension to run a separate kerosene pump. By powering the fuel pump(s) with methane, fuel-rich preburner combustion doesn’t cause problems like it would with kerosene. Taken to the extreme, including changing the engine coolant from methane to kerosene, as on the Merlin, it would end up being a high-efficiency staged combustion engine with a methane powered kerosene pump.
This would allow an increase in the first stage propellant mass by 15 to 20%, with an almost corresponding increase in payload to the same staging velocity, without increasing the external size of the stage. The advantage gets back to why SpaceX decided on RP-1 for the Falcon 9 first stage. For a given chamber pressure and engine cycle, it offers more capability in the same size vehicle. Since Super Heavy is Earth based and won’t be launched from Mars, it doesn’t really need to rely on just methane.
It’s should be a fairly cheap and simple modification which doesn’t touch the LOX pump and adds the kerosene pump to an extension of the methane pump’s upper shaft, with the methane infeed line teeing in between the two pumps.
Sounds like added complexity needing another stage of control of combustion. Single fuel with single oxidizer usually gives linear thrust variation with added flow. Going to a second-order equation of thrust and flow gets to a point of complexity not worth the effort. Your example gives a secondary fuel supply directly proportional to the primary flow. Both fuels only give best burn at one ratio. Your two-stage pump has no variation available to compensate for the varying combustion conditions. A series-fed two-turbine, two-compressor pump might be needed to make it work. There is currently a lot more known and tested on single-fuel single-oxidizer turbo pump rocket engines. Dual-fuel engines have many more unstable possibilites than the single-fuel models.
Are there any models available for this combination of fuels you propose?
The Russians have suggested using dual-fuel engines before, but I don’t know if they’ve ever built a prototype. I think they were looking at RP-1/LH2, and the idea was that instead of using two stages, such as the common RP-1 first stage and an LH2 second stage, you just shift fuels, saving the weight of an engine and an interstage. But that’s a much bigger jump than I’m suggesting.
The Raptor is already a dual pump, dual turbine engine. The LOX pump and turbine is a full flow design, with all the LOX flowing through the pump, preburner, and turbine. But unlike the NK-33, RD-170, and RD-180 engines, the Raptor doesn’t use its oxygen rich preburner and turbine to drive the fuel pump. By only driving the LOX pump, the load on the Raptor’s LOX turbine is only about 55% of what it would take to drive both pumps, which should result in a less fuel consumed in the pre-burner for the large but fixed amount of oxygen, and thus lower turbine inlet temperatures.
The separate methane turbopump on the Raptor is also full flow, so all the methane flows through pump, preburner, and turbine, mirroring the oxidizer pump, with a small amount of oxygen burned in its preburner. Nobody tries using a kerosene rich pre-burner for staged combustion because the exhaust is horribly sooty, so US gas-generator engines like the Merlin just dump it overboard, and that’s a considerable waste of fuel and a big performance hit. But a methane rich preburner is fine because the exhaust is clean.
Most of the Raptor’s ISP increase over the Merlin comes from using a high-chamber pressure fed by a full-flow staged combustion cycle, not from the switch from kerosene to methane.
The Raptor has a 4,400 psi chamber pressure, compared to Merlin’s 1,400 psi. Along with full-flow staged combustion that drives the pumps to deliver that pressure, this results in a reported ISP increase from the Merlin’s 282 second at SL to the Raptor’s 320 seconds at SL, and the Merlin’s 311 seconds vacuum to the Raptor’s 350 seconds vacuum. (Reported Raptor values might not be completely accurate yet). That’s about a 12 to 13 percent increase in ISP, whereas a simple switch to methane should only provide about a 3% increase. Nobody used methane because nobody thought the slight increase in ISP sufficiently offset the larger tanks for the lower density fuel.
The full-flow staged combustion cycle is a big win, and already working on the Raptor, which will be the first engine in history to actually fly with that cycle.
So what I’m suggesting is: Don’t touch the LOX pump. It’s fine.
Change the methane pump to provide roughly the same power output from a reduced methane flow rate, since we’re supplementing methane with kerosene in the combustion chamber. The methane turbopump will still be running fuel rich, and thermal limits and engine cooling requirements will determine how little methane can be burned (and how much kerosene can be substituted for methane in the combustion chamber) before the fuel pump’s turbine inlet temperatures or throat wall temperatures get too high.
Next, add a regular kerosene pump to the upper end of the shaft on the fuel turbopump, teeing in the methane infeed line to get around the shaft extension. Special seals to separate the kerosene from the methane aren’t critical because both fuels are hydrocarbons, so there’s no risk of combustion in the pumps. The kerosene doesn’t go anywhere but from the kerosene tank, through the kerosene pump, and straight to the injectors.
The RP-1/methane mixture ratio is going to be relatively fixed, which results in a new Oxygen/Fuel ratio that’s simple to calculate ratiometrically, because all we’re doing is burning a different cut of hydrocarbons. RP-1 was already a complex set of hydrocarbons, and the O:F ratio is just the average of all the different O:F ratios it would take to burn RP-1 as individual fractions.
What my idea doesn’t offer is a way to shift that ratio so it could be optimized like the Russian concept of switching fuels during different stages of flight. Having both fuel pumps on one shaft means it’s stuck with the mix it starts with.
However, if the methane flow is allowed to start a bit before the RP-1 flow, ignition should be a whole lot easier, and restart might be simple too.
Anyway, that’s what I’m thinking of. I’m not sure it’s worth it, but it would let SpaceX boost the pad weight and fuel fraction by a significant amount, which seems to be something they love doing.
Besides the proposal for bi-propellant of tri-propellant engines the Russians supposedly ran several tests with existing engines to prove the idea. This mostly happened in the early 1990s I think.
One thing they did was they ran the RD-0120 engine of the Energia with LOX/Methane, similar to the US tests with the RL-10, proving it could be used with either LH2 or Methane.
Besides changing the propellant, another thing which they looked at was engines with variable mixture ratios. Particularly in LOX/LH2 engines, LOX has a density close to 1.140 g/cc (similar to Kerosene at 1.02 g/cc), while LH2 has a density of 0.071 g/cc. So basically the more oxidizer rich the mixture is in a LOX/LH2 vehicle the better the overall propellant density will be. The downside with running oxidizer rich is that you decrease Isp, but if you run the engine on LOX/Kerosene you also decrease Isp. Only having two propellants simplifies the plumbing and tank structure. Something which is something typically underestimated as a complexity problem with rockets in general.
I have simplified my idea. The Raptor would probably run fine on alcohol almost as is, since alcohol doesn’t burn with a sooty residue, and the performance hit from a slightly lower ISP would be more than offset by the increased propellant density.