RS-68

[baldusi]

If the objective is a high-thrust upper stage engine for a heavy earth departure stage, then the most reasonable approach is to ditch the gas generator and go with some variant of the expander cycle or a tap-off cycle depending on how much thrust is required. This makes flight-start much easier and more reliable. So one could argue that BE-3U is the spiritual successor to RS-68.

Why not go crazy and use FFSC in a Raptor fashion of mass production and extreme reliability through heavy testing?

That's the origin story of Raptor. It was a hydrolox FFSC for three years before they pivoted to methalox. Mars ISRU was one reason for that, but also, pumping LH2 to high pressures is very difficult. SpaceX knew they were gonna be up against it to develop the ox-rich LOX pump. They probably wanted to focus their attention on overcoming that challenge and avoid biting off more than they could chew on the fuel side on the engine. If it was too big a leap for SpaceX, it's too big a leap for most engine suppliers.

It was hydrolox FRSC, nos FF. The problem to do FFSC on hydrolox is that you get something like 80% of the power on the H2 turbine, where you need 30% of your power, and 20% of your power on the O2 turbine where you need 70% of your power. Methalox has an almost 30%/70% ration on both turbines and pumps. That's why going FF means using methalox.

[envy887]

So one could argue that BE-3U is the spiritual successor to RS-68.

Except for thrust.  BE-4 (and Raptor too) is closer to RS-68 in that regard than BE-3U.   I still don't get it myself.  BE-4 still uses cryogenic fuel like RS-68, but has more pump-cycle complexity for less thrust and ISP than RS-68.  Less cost maybe, but we'll have to see about that.

 - Ed Kyle

But Vulcan is both higher-performing and cheaper than any single-stick Delta IV. This suggest that the higher engine complexity and count actually makes for a better overall system, by simplifying other components and systems. RS-68's fuel adds a lot of tankage and dry mass that adds expense and decreases performance compared to BE-4.

[envy887]

SpaceX expended a stage on STP-2. How much was that contract?

Not on purpose.  The core stage intended to land OCISLY, but missed.

 - Ed Kyle

They said before the launch that it was going to be an extremely difficult landing with a low probability of success. I highly doubt they were pricing the mission below normal profit margin, with the intent to make it up on later flights of that core. Doubly so because FH has a very low flight rate so such a reflight would not happen for a long time, and no FH customers have accepted used cores so far as we know, and SpaceX has never recovered and reused a FH core, and the USAF isn't very sensitive to price, and I could go on.

[Aeneas]

It was hydrolox FRSC, nos FF. The problem to do FFSC on hydrolox is that you get something like 80% of the power on the H2 turbine, where you need 30% of your power, and 20% of your power on the O2 turbine where you need 70% of your power. Methalox has an almost 30%/70% ration on both turbines and pumps. That's why going FF means using methalox.

I don't get it. What's the problem with having different sized turbo pumps on the hydrolox FFSC? You still have the advantage of gas in the combustion chamber, quite cool pumps and no complex bearings that need extra helium.

[edkyle99]

But Vulcan is both higher-performing and cheaper than any single-stick Delta IV. This suggest that the higher engine complexity and count actually makes for a better overall system, by simplifying other components and systems. RS-68's fuel adds a lot of tankage and dry mass that adds expense and decreases performance compared to BE-4.

Vulcan is called a "single core" design, different than the Delta 4 single core Medium and triple core Heavy alternatives, but those Vulcan "single cores" are different for the Medium and Heavy Vulcan versions.  The Heavy upper stage is stretched and may use more engines than the Medium upper stage.  Vulcan Heavy also leans heavily on those six big GEM63XL boosters, and requires two BE-4 engines per core rather than only one RS-68 for Delta 4 CBC.  Vulcan's first stage is fatter than Delta 4 CBC, though slightly shorter.  I think we'll see solids on most Vulcans, something not needed by Delta 4 Heavy. 

 - Ed Kyle

[baldusi]

It was hydrolox FRSC, nos FF. The problem to do FFSC on hydrolox is that you get something like 80% of the power on the H2 turbine, where you need 30% of your power, and 20% of your power on the O2 turbine where you need 70% of your power. Methalox has an almost 30%/70% ration on both turbines and pumps. That's why going FF means using methalox.

I don't get it. What's the problem with having different sized turbo pumps on the hydrolox FFSC? You still have the advantage of gas in the combustion chamber, quite cool pumps and no complex bearings that need extra helium.

  I thought you understood the cycle. Let me explain. In a full flow cycle, you take all the fuel, mix it with a tiny bit of oxidizer on the preburner to convert it to hot gas (actually it's critical fluid, but bear with me). The turbines are heat engines, so you run that hot gas through the turbine to power the fuel pump. Remember this: fuel preburner->fuel turbine->fuel pump. The oxidiezer is the same oxidizer preburner->oxidizer turbine->oxidizer pump.
  Now, the oxidizer to fuel ratio (O/F), is given by the main combustion chamber ratio, so that's a given. And how much heat is transfered to the turbines is a function of the amount of gas and it's elements. The formula is very correlated to:
 Power = Massflow x element specific heat.
 It just happens that for hydrolox engines, O/F ratio is somewhere between 5.5 and 6. And hydrogen has a specific heat around 15 times higher than oxygen. That means that you have 15 time more power (per unit of mass) on the hydrogen side, than on the oxygen side.
  So, you can not get higher pressure on the whole engine (and thus efficiency) than your oxygen side, which is really, really under powered. In other words, you are actually better off running the pumps with hydrogen, and still have a higher performing (and cooler turbines) than full flow. The SSME designers knew quite well what they were doing.
  Methalox, on the other side, have a very similar relation of power available to each turbine to power needed by each turbopump. So going full flow does allows you to extract a lot more work.

[Welsh Dragon]

But Vulcan is both higher-performing and cheaper than any single-stick Delta IV. This suggest that the higher engine complexity and count actually makes for a better overall system, by simplifying other components and systems. RS-68's fuel adds a lot of tankage and dry mass that adds expense and decreases performance compared to BE-4.

Vulcan is called a "single core" design, different than the Delta 4 single core Medium and triple core Heavy alternatives, but those Vulcan "single cores" are different for the Medium and Heavy Vulcan versions.  The Heavy upper stage is stretched and may use more engines than the Medium upper stage.  Vulcan Heavy also leans heavily on those six big GEM63XL boosters, and requires two BE-4 engines per core rather than only one RS-68 for Delta 4 CBC.  Vulcan's first stage is fatter than Delta 4 CBC, though slightly shorter.  I think we'll see solids on most Vulcans, something not needed by Delta 4 Heavy. 

 - Ed Kyle

But to be fair, every single Delta IV version also has different cores, every core of the Heavy is also different, so Vulcan is hardly at an advantage or disadvantage there. (Oh, if only we'd have had Atlas V Heavy!)

[su27k]

It was hydrolox FRSC, nos FF. The problem to do FFSC on hydrolox is that you get something like 80% of the power on the H2 turbine, where you need 30% of your power, and 20% of your power on the O2 turbine where you need 70% of your power. Methalox has an almost 30%/70% ration on both turbines and pumps. That's why going FF means using methalox.

There were proposals for hydrolox FFSC though, for example RS-2100. Also IPD is hydrolox.

[LouScheffer]

[...] hydrogen has a specific heat around 15 times higher than oxygen. That means that you have 15 time more power (per unit of mass) on the hydrogen side, than on the oxygen side.
  So, you can not get higher pressure on the whole engine (and thus efficiency) than your oxygen side, which is really, really under powered. In other words, you are actually better off running the pumps with hydrogen, and still have a higher performing (and cooler turbines) than full flow.

I believe that in theory you could just increase the temperature of the oxygen pre-burner until you get the power you need.  But in practice oxygen rich combustion is already a hideous materials problem.  Increasing the temperature enough to balance the pumps probably makes it a problem where no known materials suffice.

[Aeneas]

[...] hydrogen has a specific heat around 15 times higher than oxygen. That means that you have 15 time more power (per unit of mass) on the hydrogen side, than on the oxygen side.
  So, you can not get higher pressure on the whole engine (and thus efficiency) than your oxygen side, which is really, really under powered. In other words, you are actually better off running the pumps with hydrogen, and still have a higher performing (and cooler turbines) than full flow.

I believe that in theory you could just increase the temperature of the oxygen pre-burner until you get the power you need.  But in practice oxygen rich combustion is already a hideous materials problem.  Increasing the temperature enough to balance the pumps probably makes it a problem where no known materials suffice.

But with Raptor, it seems to work with a special alloy - whatever that may be. And RD-180 works with oxygen rich combustion anyway. So this given, why it's not possible to feed some hydrogen into the oxygen rich environment to enable full flow. The hydrogen side shouldn't be a problem anyway since RS-25 already prove it to work.

[Prettz]

It was hydrolox FRSC, nos FF. The problem to do FFSC on hydrolox is that you get something like 80% of the power on the H2 turbine, where you need 30% of your power, and 20% of your power on the O2 turbine where you need 70% of your power. Methalox has an almost 30%/70% ration on both turbines and pumps. That's why going FF means using methalox.

There were proposals for hydrolox FFSC though, for example RS-2100. Also IPD is hydrolox.

There's also more about converting SSME to FFSC here: https://eaglehill.us/spev-pdfs-articles/A005%20Knuth.pdf

[brickmack]

Vulcan is called a "single core" design, different than the Delta 4 single core Medium and triple core Heavy alternatives, but those Vulcan "single cores" are different for the Medium and Heavy Vulcan versions.  The Heavy upper stage is stretched and may use more engines than the Medium upper stage.  Vulcan Heavy also leans heavily on those six big GEM63XL boosters, and requires two BE-4 engines per core rather than only one RS-68 for Delta 4 CBC.  Vulcan's first stage is fatter than Delta 4 CBC, though slightly shorter.  I think we'll see solids on most Vulcans, something not needed by Delta 4 Heavy.

No, the core is identical for all Vulcan variants. Same tank size, and they always include the booster mounts whether they're used or not. Only the upper stage changes, and only the tank length (engines, thrust structure, etc are all identical). The 4 engine configuration was dropped a while ago, can get basically identical performance at drastically lower cost with 2 uprated RL10s (lower thrust, but also lower dry mass, and theres room for larger nozzles so higher ISP). I think the 4 engine option looked a lot more attractive when AR-1 was still on the table and the core stage would provide a lot less performance, needing the upper stage to have more thrust just to reach orbit

Not sure what point you're trying to make with the rest of that

[edkyle99]

Vulcan is called a "single core" design, different than the Delta 4 single core Medium and triple core Heavy alternatives, but those Vulcan "single cores" are different for the Medium and Heavy Vulcan versions.  The Heavy upper stage is stretched and may use more engines than the Medium upper stage.  Vulcan Heavy also leans heavily on those six big GEM63XL boosters, and requires two BE-4 engines per core rather than only one RS-68 for Delta 4 CBC.  Vulcan's first stage is fatter than Delta 4 CBC, though slightly shorter.  I think we'll see solids on most Vulcans, something not needed by Delta 4 Heavy.

No, the core is identical for all Vulcan variants. Same tank size, and they always include the booster mounts whether they're used or not. Only the upper stage changes, and only the tank length (engines, thrust structure, etc are all identical). The 4 engine configuration was dropped a while ago, can get basically identical performance at drastically lower cost with 2 uprated RL10s (lower thrust, but also lower dry mass, and theres room for larger nozzles so higher ISP). I think the 4 engine option looked a lot more attractive when AR-1 was still on the table and the core stage would provide a lot less performance, needing the upper stage to have more thrust just to reach orbit

Not sure what point you're trying to make with the rest of that

I was considering the second stage to be part of the "single core", so different designs for Medium versus Heavy though the first stage itself appears to be the same based on current information.

My point was that Delta 4 Heavy is not that easy to replace.  It requires for Vulcan, a stretched upper stage and full SRB complement, for Falcon Heavy at least a partly expendable flight and a new service tower, and for Omega bigger first and maybe third stages plus strap-on boosters, etc.  Thanks in large part to high-thrust LH2 propulsion of RS-68.

 - Ed Kyle   

[spacenut]

I thought Vulcan was the same diameter as Delta IV and used the same tooling.  5m.  Or did they go 5.5m diameter for Vulcan because of two engines? 

RS-68 if it didn't have ablative nozzle and had a cooled nozzle, may have been a really good engine for SLS core.  However, SLS would have required a decent sized upper stage.

[Aeneas]

I thought Vulcan was the same diameter as Delta IV and used the same tooling.  5m.  Or did they go 5.5m diameter for Vulcan because of two engines? 

It's 5.4 m diameter. Like the SLS Block 1 upper stage.

[baldusi]

[...] hydrogen has a specific heat around 15 times higher than oxygen. That means that you have 15 time more power (per unit of mass) on the hydrogen side, than on the oxygen side.
  So, you can not get higher pressure on the whole engine (and thus efficiency) than your oxygen side, which is really, really under powered. In other words, you are actually better off running the pumps with hydrogen, and still have a higher performing (and cooler turbines) than full flow.

I believe that in theory you could just increase the temperature of the oxygen pre-burner until you get the power you need.  But in practice oxygen rich combustion is already a hideous materials problem.  Increasing the temperature enough to balance the pumps probably makes it a problem where no known materials suffice.

Nobody like to leave free performance on the table. Staged engines are usually limited but the turbine blade materials. You can't really run them any hotter without sacrificing margin. And I'm not saying you can't get a bit better performance by going FFSC on H2 instead of FRSC. I'm saying that you don't get the astounding differences you get on methalox.

[Aeneas]

Nobody like to leave free performance on the table. Staged engines are usually limited but the turbine blade materials. You can't really run them any hotter without sacrificing margin. And I'm not saying you can't get a bit better performance by going FFSC on H2 instead of FRSC. I'm saying that you don't get the astounding differences you get on methalox.

Then I didn't understand it yet. Why is the difference between methalox FRSC vs methalox FFSC larger than hydrolox FRSC vs hydrolox FFSC? Yes, the LH2 pump is much bigger than the LCH4 pump but that's the case for both FRSC and FFSC.
So the only but major difference is spraying a little methane into the LOX stream behaves vastly different than spraying a little hydrogen in the LOX stream?

[Prettz]

[...] hydrogen has a specific heat around 15 times higher than oxygen. That means that you have 15 time more power (per unit of mass) on the hydrogen side, than on the oxygen side.
  So, you can not get higher pressure on the whole engine (and thus efficiency) than your oxygen side, which is really, really under powered. In other words, you are actually better off running the pumps with hydrogen, and still have a higher performing (and cooler turbines) than full flow.

I believe that in theory you could just increase the temperature of the oxygen pre-burner until you get the power you need.  But in practice oxygen rich combustion is already a hideous materials problem.  Increasing the temperature enough to balance the pumps probably makes it a problem where no known materials suffice.

Nobody like to leave free performance on the table. Staged engines are usually limited but the turbine blade materials. You can't really run them any hotter without sacrificing margin. And I'm not saying you can't get a bit better performance by going FFSC on H2 instead of FRSC. I'm saying that you don't get the astounding differences you get on methalox.

Isn't the main point of doing it with hydrolox improving reliability and reusability? (over what the SSME was ever capable of)

[LouScheffer]

So, you can not get higher pressure on the whole engine (and thus efficiency) than your oxygen side, which is really, really under powered. In other words, you are actually better off running the pumps with hydrogen, and still have a higher performing (and cooler turbines) than full flow. The SSME designers knew quite well what they were doing.

I don't see how this can be right.  Take the RD-180 oxygen side.  This uses only the oxygen flow to pump all oxygen up to record-breaking pressures.  Since it's using oxygen flow to pump oxygen, this is (almost) independent of the fuel used.

But the RD-180 is not only using the oxygen flow to pump oxygen, it's also using it to pump fuel up to even higher pressure (because of cooling losses).  Take away the need to pump fuel, and oxygen flow has the power to pump oxygen to even more eye-popping pressures.   So it's really hard for me to see how the oxygen side could limit the attainable pressure of the engine.

[baldusi]

Nobody like to leave free performance on the table. Staged engines are usually limited but the turbine blade materials. You can't really run them any hotter without sacrificing margin. And I'm not saying you can't get a bit better performance by going FFSC on H2 instead of FRSC. I'm saying that you don't get the astounding differences you get on methalox.

Then I didn't understand it yet. Why is the difference between methalox FRSC vs methalox FFSC larger than hydrolox FRSC vs hydrolox FFSC? Yes, the LH2 pump is much bigger than the LCH4 pump but that's the case for both FRSC and FFSC.
So the only but major difference is spraying a little methane into the LOX stream behaves vastly different than spraying a little hydrogen in the LOX stream?

I don't think I was clear enough. You have two sides for each propellant:
1) How much power they can PROVIDE to the turbines (massflow*specific heat).
2) How much power the pumps NEED to increase the pressure to the desired level.

Now, material science is what limits turbines (usually to a max inlet temperature of 750K to 850K). So, you can not get more power than what your masswflow*specific heat provides.
Hydrogen provides so much power, that you might end with actually less power on the LOX side (because you can only use O2 to run the LOX turbine) than splitting the Hydrogen flow (in a Fuel Rich SC). Not really a loss, but your gain is relatively low. Your H2 side will run really cool. But the LOX side will be working well to the material limit to keep up.
In the methalox case, the power GENERATED by the CH4 side is almost what the CH4 pumps need for matching the LOX side pressure (all at the same turbine temperature). So you can go with low performance and super life (low temperature turbines) or dial up the performance as much as you dare, that both sides will be evenly matched.
Again, your LOX side will limit you very fast while the H2 side will basically idle in a FFSC. Was I a bit clearer?