As I understand it during ignition the SSME TVC actuators had to be set to
allow the engine to flop around a but because the transient side loads were
enough to cause damage if the TVCs were holding rigidly. I'm sure that if
there was any real gain to be had by using a dual-expansion nozzle on the
booster of a two or more stage vehicle it would have been done by now, my
guess is the extra weight of the extension isn't paid for by the short time
before burnout that the extension is doing work but I haven't done any
calcs to show that.
Andrew
On Wed, Apr 26, 2017 at 4:41 AM, Redacted sender crogers168 for DMARC <
dmarc-noreply@xxxxxxxxxxxxx> wrote:
...Part 3 provides an equation for the nozzle static pressure where flow
separation will occur in a bell nozzle (Pages 6-8 of the pdf), that
being a bell nozzle without any special contouring...
< One caution: Cornelisse et al, "Rocket Propulsion & Spaceflight
< Dynamics", 1979, says the exact pressure depends somewhat on details
like
< wall roughness, pressure gradient, viscosity, and (especially) presence
of
< solid particles.
Agreed. The equations are meant to be rules of thumb, a good starting
point to avoid flow separation, if one has the design goal to avoid flow
separation.
Our simple solution was to remove the J-2 nozzle extension, leaving just
the regeneratively cooled nozzle inner expansion section.
< Um, as I understand it, the J-2 didn't have a nozzle extension. The
lower
< half had some help from turbine-exhaust film cooling, but unlike the
F-1,
< that section was still tube-wall regen, rather than being otherwise
< uncooled.
Yep, good catch. It wasn't just film cooling.
< However, shortening a regen nozzle isn't necessarily a big deal. The
< shortened RL10s for DC-X likewise had a shorter version of the original
< RL10's all-regen nozzle.
< Modern high-altitude engines sometimes deliberately put a manufacturing
< joint in the nozzle at a suitable point, making shortening easy.
You can see from J-2 drawings where the manufacturing break was, and where
we planned to remove the rest of the nozzle.
But as part of the study, we got to see a video of an early J-2 engine
test at Santa Susana without the nozzle lip, where the flow separation
at starting was very violent, and destroyed the engine. After watching
the video, the big take-away was avoid flow separation in liquid rocket
engine nozzles.
< I.e., "we tried it once and it failed badly, so no variation on it could
< possibly work". Hundreds of Atlases flew successfully with sustainer
< nozzles running separated at liftoff, although as I noted, this was
never
< widely advertised. Avoiding separation certainly would make things
< *easier*, and on a cost-constrained program that might be deemed
< mandatory, but one should be aware that there is a choice there and
< today's conventional wisdom is not the only way.
The SSME nozzle of course is the best example of sustained, stable flow
separation at sea level operation, to allow a high expansion ratio for high
altitude/vacuum operation. It can be done, and the dual bell nozzle
concept is based on a stable flow separation point at low altitude, and
then full nozzle flow at high altitude/vacuum operation.
The question is really is the payoff in performance worth the potential
development risk. For an SSME running from lift-off to orbit one can see
why they chose a nozzle with a stable flow separation point at low
altitude. Same with the sustainer motor on the Atlas, runs all the way to
burnout on an ICBM trajectory. For a small start-up company, with a two
stage liquid rocket, the advantages for the Stage-1 nozzle might not be
worth the development risk.
The performance benefit of a high expansion ratio nozzle with a stable low
altitude flow separation point can be quantified. You adjust the nozzle
expansion ratio during the trajectory simulations, typically in a couple
discrete step changes, low altitude with stable flow separation (low
expansion ratio), and then a step change to full flow in the nozzle (high
expansion ratio). That performance benefit should then be weighed relative
to the development risk.
Charles E. (Chuck) Rogers
CRogers168@xxxxxxx
