The LOX-Kerosene data does come from a bunch of testing in the 1960s;
I’m only partway through the documentation. Lots of fun stuff, I’m
certain George would approve. Tanks of LOX next to tanks of RP-1,
linear shaped charges, mechanical cutters, deliberate spills into a
shallow depression, even a deliberate overpressurization of a
common-bulkhead tank. Real attempts to simulate credible events.
Almost certainly more expensive to repeat now than it was the first time.
And yes, if there’s to be any solution short of throwing up our hands
and designing around 100% theoretical yield, it’s going to have to
involve active mitigation. Because, “a big pit for the rocket to fall
back to”? Flame trenches come to mind. Have an inadvertent MECO at say
T+3s, and watch the LOX tank pancake into the methane and spill the
whole load into the flame trench. Confined mixing and even some
tamping. Now do the cratering calculation for half a kiloton of TNT
equivalent and see if we have to rename “Cape Canaveral” to “Canaveral
Island” :-)
So, make sure there’s always an ignition source to start a conflagration
before we can get enough mixing for a detonation. There’s still the
scenario where the engine runs a bit longer and drops the rocket a
kilometer downrange, but that likely involve falling sideways onto a
flat surface rather than dropping tanks directly on top of each other
and into a pit. Less mixing, less potential yield, but how much less?
Also, as the Other Henry pointed out, no common bulkheads please. Maybe
not run the oxidizer standpipe right through the propellant tank, or a
leak in the wrong place could lead to propellant loading inadvertently
pumping both propellants into the lower tank.
Going to be fun just thinking about all this. If we can get an actual
test budget…
John Schilling
John.schilling@xxxxxxxxxxxxxx <mailto:John.schilling@xxxxxxxxxxxxxx>
(661) 718-0955
On 5/24/2016 11:19 AM, Henry Vanderbilt wrote:
Seems to me it might be worthwhile to come at this from the opposite end: What conditions would be necessary to produce thorough mixing of the majority of a vehicle's LOX-CH4 propellants, followed by detonation? And, how unlikely can they be affordably rendered?
You'd need major structural failure with LACK of immediate ignition, followed by release of the propellants into some sort of containment (a big pit for the rocket to fall back into?) so they mix in bulk rather than boil away furiously and disperse (how long might this take in this theoretical worst-case?) followed by ignition.
So, how much of this worst-case would be rendered implausible by existing launch-site conditions and launch procedures? How much additional mitigation would be relatively cheap & easy? (EG, leave the engine(s) idling in an abort case to guarantee an immediate ignition source? Or just add a bit of thermite to the det cord...)
And at that point, how much propellant might plausibly still thoroughly mix before igniting, worst-case?
I'd look at where that 10% figure for LOX-kero came from. How much was BOTE guesstimate, how much was real-world experience? A similar logic chain, adjusted plausibly for the increased miscibility minus the easy mitigations, could be salable.
Absent funding for field experiments (if you did get funding, I know someone who'd bid low because he REALLY wants to make that earth-shattering Ka-Boom happen - hi, George!) it's going to be a plausible best guess. The goal then is to see how much plausibility can be assembled for the existing (near-zero, I'd guess) budget...
Henry
On 5/23/2016 7:24 PM, John Schilling wrote:
Question for the group mind: Has anyone out there come across
a formal methodology and/or experimental data for determining
the TNT equivalence of a LOX-Methane propellant explosion?
Not, I hasten to add, a deliberately premixed one - I think I
classified that potential blend as an "insanely dangerous" mixed
monopropellant in my Space Access talk. But we've now got
multiple launch providers talking about flying LOX-Methane
bipropellant out of the major ranges, and people are starting
to ask for specifics about how big an explosion you might get
in a worst-case accident.
Which presumably would be a low-velocity crash right next to
(or back onto!) the pad after an early engine cutoff, tank
rupture and no immediate ignition so the potential for
extensive mixing followed by detonation. LOX and methane
are fully miscible as liquids at typical storage temperatures,
and not hypergolic, so the usual explanations of why we can't
get 100% mixing and full theoretical energy release don't hold.
With LOX-Kerosene, for example, the "maximum credible event"
is assessed at about 10% of theoretical yield partly on the
grounds that the kerosene will freeze as it mixes.
I would expect that LOX-Methane would be somewhere between 10%
and 100% of theoretical yield in a worst-case crash, but that
covers a fairly broad range - and at 100% theoretical yield,
it probably covers an impractical amount of the Cape in the
blast zone. So, looking for the least troublesome way to
narrow that estimate, and if that comes down to "Yeah, the
experiments are a bitch but the guys at White Sands did that
in 1963...", this might be a place to ask.
John Schilling
john.schilling@xxxxxxxxxxxxxx
(661) 718-0955