Why doesn't the Merlin have a nozzle like SSME?

[IainMcClatchie]

As near as I can tell, the Merlin 1D has a bell nozzle.  Specifically, the nozzle is diverging at the base.

The very end of the SSME nozzle had greatly reduced divergence at the base.  This feature causes a local high pressure area around the lip, which reduces the tendency of flow to separate.

If I understand correctly, when the Merlin throttles down its chamber pressure drops, and as a result all downstream absolute pressures drop as well.  So unlike most booster engines, the amount of expansion the nozzle can do is limited not by the base pressure at liftoff, but the base pressure at landing, which is significantly less.  I'd expect that compared with other booster engines, the Merlin 1D probably has a higher base pressure at 100% thrust.  I'd expect it could eke out a bit more Isp by increasing expansion, using the SSME nozzle trick.

I can think of two, maybe three potential reasons why the nozzle doesn't use the SSME trick:
  * The local pressure increase causes a hot spot on the nozzle wall.  The SSME nozzle is actively cooled all the way down to the base.  The Merlin 1D nozzle active cooling does not extend down to the base, and so maybe it can't handle the increased heat load down there.
  * The diameter of the nozzle is constrained by fitting nine of them on the bottom of the Falcon 9, which is itself diameter constrained.  To take advantage of lower base pressure, the nozzle would have to get larger and it can't.
  * Something to do in midair restarts while going backward, supersonic?

Can anyone give me the real scoop?

[OneSpeed]

... The diameter of the nozzle is constrained by fitting nine of them on the bottom of the Falcon 9, which is itself diameter constrained.  To take advantage of lower base pressure, the nozzle would have to get larger and it can't. ...

If the figure of the Octaweb on the SpaceX site is accurate, for a 3.66m fuselage Ø, a Merlin 1D is 0.89m Ø. If the engines are going to gimbal, they can't be much larger.

[envy887]

Available expansion is limited by Merlin throat size and F9 diameter: they can't make the nozzle bigger and still fit it on Falcon, and they can't make the throat smaller without sacrificing thrust.

With fixed expansion ratio, the only reason to try to limit flow separation is if they want to throttle lower at landing AND the throttle limit is due to flow separation.

But they might not want to throttle lower, since a slower landing increases gravity losses (they don't want to sacrifice thrust for the same reason).

Merlin is very well optimized for what it does. It has a very different flight profile from the SSME, so it's not surprising it's designed differently.

[TomH]

Remember that RS-25 is a sustainer, burning from launch to disposal orbit. Extreme efficiency is required to squeeze out every second of ISP. And this costs money. One of Merlin's efficiencies lies in cost. (Another is T/W.)

[Jim]

Remember that RS-25 is a sustainer, burning from launch to disposal orbit. Extreme efficiency is required to squeeze out every second of ISP. And this costs money. One of Merlin's efficiencies lies in cost. (Another is T/W.)

That is it

[Nilof]

On a first stage, fitting more engines at the bottom of your rocket is more important than optimizing the Isp with bigger nozzles. If your TWR is 1.3 and you are accelerating upwards, a 10% increase in thrust helps more during the early phase of flight than a 30% improvement in Isp.

[envy887]

Remember that RS-25 is a sustainer, burning from launch to disposal orbit. Extreme efficiency is required to squeeze out every second of ISP. And this costs money. One of Merlin's efficiencies lies in cost. (Another is T/W.)

I doubt reducing the nozzle divergence would add much, if any, cost. It simply doesn't improve performance much for Merlin's flight profile on F9 unless they want to throttle lower, which I doubt they do.

[livingjw]

I believe the Merlin is actively cooled all the way down to the end. Pretty sure they use alternating coolant tubes, down and up.

John

[IainMcClatchie]

Oh, right.  The MVac has a radiatively cooled nozzle extension, and the M1D does not.  Thanks.