As my non-evil twin has pointed out, the problem here wasn't fuel and
oxidizer meeting in the He press system, it was oxidizer alone (plus a
fast pulse of high-pressure He) making fuel of a check valve that wasn't
supposed to be fuel...
But yeah, AIUI some systems do use a common He bottle to pressurize both
fuel and oxidizer, and depend on check valves to keep the two from ever
meeting back in the common parts of the press plumbing. This strikes me
as workable for a short-life expendable system that starts from a
clean-parts-assembled known state. But yes, very sporty for any sort of
long life reusable system - because given enough time and operating
cycles fluids ALWAYS migrate to places they're not supposed to.
As for the penalty for completely separate press systems, well,
obviously some amount of additional mass, parts-count, and cost. Parts
count and cost should be easily figureable for a given design. The
exact amount of extra mass depends on how precisely it's affordable to
size things to their tasks. In the real world with a limited range of
tubing, press bottle, valve, and fitting sizes in the catalog, probably
some noticeable chunk of extra mass.
Henry
On 7/15/2019 7:19 PM, Paul Breed wrote:
How big is the penalty for completely separate pressurization systems.....
On Mon, Jul 15, 2019 at 5:30 PM Henry Vanderbilt <hvanderbilt@xxxxxxxxxxxxxx <mailto:hvanderbilt@xxxxxxxxxxxxxx>> wrote:
And if that check valve was already leaking, some sort of grit or
contamination as the original cause is a pretty good bet.
Guessing at "what were they thinking" here... The SuperDraco
thrusters were originally going to be used routinely to land the
capsule. Hence, design them for easy turnaround - check valves to
keep propellant out of the He press system, rather than burst
discs that'd need replacing after every flight. Now, of course,
the SuperDracos will only be used for a presumably rare abort, so
burst discs are tolerable.
As for titanium rather than, say, stainless, presumably that's to
save mass.
On a pressure-fed thruster system that is to be routinely used,
possibly pressurized more than once per flight, any thoughts on
how to avoid such issues? Off the top of my head, doing sensitive
fuel/oxidizer detection on the He side of the press system during
servicing is one possibility. If you find traces, inspect the
check valves. Maybe pull vac on the He system to detect any leaky
check valves. Maybe combine the two, doing propellant detection
on the evacuated stream as you pull the He system down to vac.
Or, in-flight, propellant detection upstream of the He check
valves - or even simple liquid detection - could be enough for a
no-go warning. Or a "slow press only, if you MUST use these
thrusters" warning.
Henry
On 7/15/2019 3:58 PM, Ben Brockert wrote:
Titanium is dramatically less compatible with oxidizers than
aluminum is. It's sensitive to impact with N2O4, especially when
there's grit or filings in the impact, and check valves are
really effective at creating impacts and finding grit.
On Mon, Jul 15, 2019 at 3:28 PM Henry Vanderbilt
<hvanderbilt@xxxxxxxxxxxxxx <mailto:hvanderbilt@xxxxxxxxxxxxxx>>
wrote:
On 7/15/2019 3:13 PM, Henry Vanderbilt wrote:
On 7/15/2019 1:05 PM, George Herbert wrote:<Emily Litella> Never mind! I went on to read the SpaceX
*Lars Osborne (@lars_0 <https://twitter.com/lars_0?s=11>)*
7/15/19, 12:51 PM
<https://twitter.com/lars_0/status/1150855437765316608?s=11>
I work with hypergolic propellants now. I feel pretty
satisfied with this explanation and think it has lessons
for the industry.
What is missing is that the destruction of the check valve
exposed bare Ti and accelerated the metal chips into the
MON tank at high velocity. pic.twitter.com/09ud6OZFMv
<https://t.co/09ud6OZFMv>
Lars posted this a bit ago as the SpaceX failure analysis
went public. Basically, slug of NTO migrated upstream past
a He check valve, pushed back downstream on subsequent
pressurization to fracture the check valve and throw
Titanium metal fragments into the NTO tank...
Fix is replacing with burst discs.
Replacing He propellant pressurization check valves with
burst discs? I'm not following your thinking here... Or
are you saying, adding burst discs by the check valves?
piece, and their cure is to make the high-thrust escape
rockets one-time activation by subbing burst discs for the
check valves in the high-speed He press system, thus
eliminating this particular failure probability. </Emily Litella>
Interesting that high-velocity NTO igniting the inside of the
Ti check valve is described as a surprise, and something not
seen before. I'd assume Ti is like Al in inherently forming
a surface oxide layer that makes it compatible with strong
oxidizers, unless something is actively scrubbing off the
oxide layer. So, was it the check valve slamming open too
hard with NTO present? Or could high enough velocity NTO
fluid perhaps directly scrub the Ti down to bare metal?
Henry