What makes Centaur such a “good” upper stage?

[Skye]

Centaur is renowned as a phenomenal upper stage (family), good for LEO, GTO, & beyond. Utterly amazing all round. My question is what makes it so good? Does it tick all the right boxes? (If so, what are those boxes & what does it do that others don’t?), does it have some kind of secret sauce, & can it be replicated in other stages?

Note: I am unsure as to whether or not I should put this in the QnA section, but I decided to put it in the ULA section since it discusses a ULA upper stage (family) used on 2 of ULA’s rockets (I am aware Atlas is a family, just said rockets cause I could find a good way to integrate rocket & rocket family well into a sentence)

[lightleviathan]

It’s very light and very efficient. That’s basically why.

[Orbiter]

Big hydrolox upper-stage tank. Very efficient RL10 engines.

[John-H]

Is the stainless steel tank lighter than any other kind? I thought that was the secret sauce  that put it ahead.

[DeimosDream]

Centaur is renowned as a phenomenal upper stage (family), good for LEO, GTO, & beyond. Utterly amazing all round. My question is what makes it so good? Does it tick all the right boxes? (If so, what are those boxes & what does it do that others don’t?), does it have some kind of secret sauce, & can it be replicated in other stages?

Centaur is excellent for GTO and beyond but is actually rather lackluster for LEO.

As others have mentioned it is hydrolox (best practical specific impulse) and remarkably light weight and efficient for its volume (good mass fraction). Combined that provides incredible delta-v potential for reaching high-energy GTO, GEO direct, and beyond-earth trajectories.

However small efficient light weight engines with high specific impulse comes with a trade off: low thrust. To compensate the first stage must deliver centaur into a higher energy starting trajectory. That isn't optimal for moving a heavy payload into LEO and is also problematic if you wanted to try and make a new partially reusable rocket with a centaur-like upper stage and a reusable first stage.

Is the stainless steel tank lighter than any other kind? I thought that was the secret sauce  that put it ahead.

The pressure stabilized ballon tanks are another ingredient. Ultra light weight which helps compensate for low-density propellant volume, but I think those also require special ground handling as a tradeoff.

[envy887]

Centaur has a lot good points, several of which have been mentioned, including its high specific impulse and good mass ratio.

Some other reasons that it's highly regarded are:

1) Heritage. Centaur has been around since 1962 and launched over 270 times.
2) Reliability. See #1. Centaur has been a very reliable stage.
3) Restartability and endurance. These enable complex mission trajectories with multiple burns with long wait periods between them. Not all upper stages can restart or wait long periods between burns.
4) Low thrust. This is a downside for performance, but has other advantages. G-loads at burnout are low, as are vibrations. It may also help improve insertion accuracy, saving the payload fuel on course corrections.

On the other hand, Centaur is very expensive, and its light weight and low thrust requires a high staging velocity which makes booster recovery difficult.

[sdsds]

Everything said above, for sure.

Also Centaur has been remarkably flexible in part due to its somewhat modular design.
- Optional Mission Extension Kit, for missions requiring it
- Two intrinsically different ways of handling fairing loads
- Single or dual engine configurations
- Optional non-deployed payloads on the aft bulkhead.
- Option to be carried as a payload by the spacecraft (LCROSS)

[Proponent]

Does Centaur use hydrogen boil-off to cool the lox? If do, has it always done so?

[Jim]

Does Centaur use hydrogen boil-off to cool the lox? If do, has it always done so?

No, it was vented overboard.  LOX did need to be cooled.

[Proponent]

The Atlas V PUG, Rev. 11 (PDF page 39, para. 2.3.1.5) mentions a capability for 3-burn missions involving a 5.2-hour coast between burns 2 and 3. Centaur's viability, then, would need to extend to something like 6 hours. That's maybe not quite as long as the viability of the much larger Saturn S-IVB, which did cool its lox, but given the square-cube law, I'd have thought the advantage of lox cooling would be, if anything, larger for Centaur.

[Jim]

The Atlas V PUG, Rev. 11 (PDF page 39, para. 2.3.1.5) mentions a capability for 3-burn missions involving a 5.2-hour coast between burns 2 and 3. Centaur's viability, then, would need to extend to something like 6 hours. That's maybe not quite as long as the viability of the much larger Saturn S-IVB, which did cool its lox, but given the square-cube law, I'd have thought the advantage of lox cooling would be, if anything, larger for Centaur.

Not really needed, the insulation was enough.

[Sam Ho]

The Atlas V PUG, Rev. 11 (PDF page 39, para. 2.3.1.5) mentions a capability for 3-burn missions involving a 5.2-hour coast between burns 2 and 3. Centaur's viability, then, would need to extend to something like 6 hours. That's maybe not quite as long as the viability of the much larger Saturn S-IVB, which did cool its lox, but given the square-cube law, I'd have thought the advantage of lox cooling would be, if anything, larger for Centaur.

Not really needed, the insulation was enough.

ULA lists Centaur III mission duration as 8 hours and Centaur V duration at 12 hours, with the possibility of a multi-month mission extension kit:
https://www.flickr.com/photos/ulalaunch/49720321573

The longest Centaur mission to date was STP-3 at 7 hours 10 minutes:
https://blog.ulalaunch.com/blog/sbirs-geo-flight-5-ula-successfully-launches-atlas-v-0-0-0

ULA published several studies on extended-duration (weeks to months) Centaur derivatives as part of what became ACES and eventually the multi-month mission extension kit:
Szatkowski et al, (2006) Centaur Extensibility For Long Duration
https://www.ulalaunch.com/docs/default-source/extended-duration/centaur-extensibility-for-long-duration-2006-7270.pdf

Kutter, et al, (2005) Atlas Centaur Extensibility to Long-Duration In-Space Applications
https://www.ulalaunch.com/docs/default-source/extended-duration/atlas-centaur-extensibility-to-long-duration-in-space-applications.pdf

Kutter, et al, (2008) A Practical, Affordable Cryogenic Propellant Depot Based on ULA’s Flight Experience
https://www.ulalaunch.com/docs/default-source/extended-duration/a-practical-affordable-cryogenic-propellant-depot-based-on-ula's-flight-experience.pdf

With respect to using LH₂ boiloff to cool the LO₂, it can be done passively; see this quote from Kutter 2005:

Perhaps most important is the common bulkhead; a feature of all Centaur tanks and carried over to the ICES. Due to inherent thermodynamic properties, it is two to 10 times more efficient to vent hydrogen in terms of amount of heat removed per pound than oxygen, (table 1). The common bulkhead provides an extremely efficient and reliable method to direct all stage heating to the LH2 tank, where the energy can be efficiently removed via H2 venting to allow zero O₂ boil off.

This inherently structurally and thermally efficient design will reduce the system boil off to 0.1%/day for the baseline ICES compared to the current Titan Centaur ~2%/day. This low boil off provides a common stage that is ideally suited for standard Low Earth Orbit (LEO), Geosynchronous Transfer Orbit (GTO), and Geo-stationary Orbit (GSO) missions with mission durations up to 24 hours and provides the foundation for much longer missions.

[sdsds]

Yes, another advantage of the current Centaur is that it provides a platform for an evolved upper stage like ACES. Even the first step in that (using an internal combustion engine for integrated vehicle fluids) seems like it would be a major advancement.

[Proponent]

Kutter, et al, (2008) A Practical, Affordable Cryogenic Propellant Depot Based on ULA’s Flight Experience
https://www.ulalaunch.com/docs/default-source/extended-duration/a-practical-affordable-cryogenic-propellant-depot-based-on-ula's-flight-experience.pdf

With respect to using LH₂ boiloff to cool the LO₂, it can be done passively; see this quote from Kutter 2005:

Perhaps most important is the common bulkhead; a feature of all Centaur tanks and carried over to the ICES. Due to inherent thermodynamic properties, it is two to 10 times more efficient to vent hydrogen in terms of amount of heat removed per pound than oxygen, (table 1). The common bulkhead provides an extremely efficient and reliable method to direct all stage heating to the LH2 tank, where the energy can be efficiently removed via H2 venting to allow zero O₂ boil off.

This inherently structurally and thermally efficient design will reduce the system boil off to 0.1%/day for the baseline ICES compared to the current Titan Centaur ~2%/day. This low boil off provides a common stage that is ideally suited for standard Low Earth Orbit (LEO), Geosynchronous Transfer Orbit (GTO), and Geo-stationary Orbit (GSO) missions with mission durations up to 24 hours and provides the foundation for much longer missions.

Thank you for all that information, especially Kutter & al. (2008). That makes clear heat transfer from lox to LH2 was a design feature. Now that I think about it, I don't really know that the S-IVB's means of cooling lox was any different. I had interpreted the use of LH2 boil-off to mean that GH2 was ducted to the lox tank (though not into the lox itself, of course), but maybe it was just a matter of heat transfer through the bulkhead. Does anybody know?