[AR] Re: fatigue life (was Re: Re: SpaceX F9 Launch/Update...)
- From: Henry Spencer <hspencer@xxxxxxxxxxxxx>
- To: "arocket@xxxxxxxxxxxxx" <arocket@xxxxxxxxxxxxx>
- Date: Wed, 30 Dec 2015 23:54:47 -0500 (EST)
On Thu, 31 Dec 2015, Richard Garcia wrote:
Allowing the inner wall to expand may ameliorate the problem but does
not eliminate it...
That's why I said "much of the problem". :-) How much remains, due to
differential expansion in the wall itself, depends on details of materials
and design.
... What would really help is reducing the thermal gradient across the
wall. The best way to do that would be to make it out of something very
thermally conductive (i.e. copper) and to make it as thin as possible...
I would say the *best* way to do that would be to keep hot gas away from
the wall, reducing heat flow through it, since the gradient is more or
less thermal resistance (fixed by materials and dimensions) times heat
flux. Of course, the ways to fend off hot gas, like film cooling, all
come at a price; it may be worth it. For example, P&W's SSME proposal
used transpiration cooling to give its inner walls essentially unlimited
life -- the SSME spec called for 400 cycles and P&W felt that this was
impossible with just milled channels.
That thinness is where you start to run into trouble.
Making the thrust chamber walls thinner than some copper milled wall
chambers already are (less that 0.075") will limit what chamber pressure
you can run at, sacrificing performance.
One caveat: if I'm not mistaken, there's a hidden assumption here, that
there's a significant pressure difference (proportional to the chamber
pressure) across the inner wall. That's true of orthodox designs, but
it's not absolutely inevitable. Put a final pump stage between the
cooling jacket and the chamber, or use a separate fluid loop and a
propellant heat exchanger for cooling, or do dump cooling with LH2 -- just
exhaust it through an auxiliary nozzle rather than trying to burn it.
All of these concepts have their own problems, but they're not ridiculous
(dump cooling in particular has been flown), and all permit nearly zero
pressure difference across the wall (perhaps only at one particular point
along its length).
State of the art, high performance, milled channel , nickle backed
thrust chambers can usually get around 100 runs before cracks start
showing up from thermal cycling induced fatigue.
Of course, there's always the option of backing off on performance. As of
1990 -- don't have more recent data -- there were RL10s at P&W with over
200 cycles on them, still in good shape, with stainless-steel tube-bundle
walls. They thought then that qualification for 300 cycles of operational
use was feasible at the original 15klb thrust, although fatigue would
become an issue at higher thrusts (as in the RL10B) -- that engine was
somewhat overbuilt for 15klbf.
In aviation, it's routine to back off on performance to improve operating
life and safety. Nobody's delighted about doing it, but it's just
necessary, so nobody blinks at it. Rocketry has some way to go yet. :-)
Henry
Other related posts: