Even without being in a closed-loop control circuit, or faulty, sensors
can react to inputs other than what you'd expect - pressure sensors, for
instance, can also act to varying extents as microphones picking up
shocks in structures they're attached to.
More subtly, in looking at data traces, it can be hard sometimes to
determine definitively what's cause and what's effect. EG, a tank
structural failure under load may well have some interesting pressure
spikes closely associated with it, and absent other data resolving the
ambiguity, which came first?
On 6/30/2015 8:58 PM, Troy Prideaux wrote:
In a purely general context: a “counterintuitive” explosion can also
mean a faulty sensor input, especially if the
flow/pressure/vent/relief/control mechanisms were governed by such
readings. Of course, active (only) controls are probably an extremely
rareoccurrencefor critical systems and a bad sensor reading can be
generally correlated with other sensor readings (in complex systems with
lots of sensors) to identify possible issues with its readings.
Troy
The "counterintuitive" comment may mean that we're all up in the night
here and anyone from SpaceX reading this is not getting any useful
information. Of course, something that seems "counterintuitive" to those
not familiar with cryogenics may be more intuitive to those who have
experience with weird things like geysering or film boiling...or someone
who really understands BLEVEs (not me).