NASA-speak on sensor glitch

Can somebody put this 'NASA-speak' into English for me?


From NASA Watch
(http://www.nasawatch.com/archives/20...t_into_lo.html)

Insight into Low Flow Sensor Anomaly
Note from : "The level sensor itself is a platinum wire.
The wet/dry situation is a measure of resistance. After the tank was
drained, one sensor showed wet but 3 hours later showed dry. In a later test
the sensor again disagreed with the others but 5 minutes later agreed.
Current thinking is that the vast time difference in the reading returning
to nominal indicates that this is a mechanical problem where vibration (wind
gusts, etc.) that can produce instant effects would be causal, rather than
electrical where environmental (temperature, humidity) changes that are slow
to build and slow to dissipate would be the culprit."

Sounds like a case of vapor lock. When did they start buying shuttle
engines from Chevy?


"Jim Oberg" wrote in message
.. .
Can somebody put this 'NASA-speak' into English for me?


From NASA Watch
(http://www.nasawatch.com/archives/20...t_into_lo.html)

Insight into Low Flow Sensor Anomaly
Note from : "The level sensor itself is a platinum wire.
The wet/dry situation is a measure of resistance. After the tank was
drained, one sensor showed wet but 3 hours later showed dry. In a later
test
the sensor again disagreed with the others but 5 minutes later agreed.
Current thinking is that the vast time difference in the reading returning
to nominal indicates that this is a mechanical problem where vibration
(wind
gusts, etc.) that can produce instant effects would be causal, rather than
electrical where environmental (temperature, humidity) changes that are
slow
to build and slow to dissipate would be the culprit."

It is badly written, isn't it?

If I read it right, they're saying that the sensor in question does
eventually read correctly, but takes a short period of time to do so. If
the malfunction was an electrical problem caused by temperature or humidity
issues inside the tank, the sensor would take much longer to read correctly
than just a few minutes or hours. So, their best guess is that the
malfunction is a mechanical issue where the sensor suddenly gives a correct
reading after experiencing some sort of vibration.

So, if I'm reading it right, they're saying it's a loose connection
somewhere between the physical sensor and the computers that read it, and
not a problem with the sensor itself. Whack it with a stick and it'll work
right for a while.

I don't understand the description of the sensor itself. I'm not a EE (I'm
software, systems, and control electronics), but I don't see how a single
wire can be used as a fuel level sensor. I'd have expected the sensor to
have two electrical points of contact that are physically separated inside
the tank (if you measure the electrical resistance between them you'd get a
different reading if there's fuel in the tank versus when there's not).
I've no idea if that's how the system in question actually works. Anyone
got a schematic I can look at?

El Doctoro


"Jim Oberg" wrote in message
.. .
Can somebody put this 'NASA-speak' into English for me?


From NASA Watch
(http://www.nasawatch.com/archives/20...t_into_lo.html)

Insight into Low Flow Sensor Anomaly
Note from : "The level sensor itself is a platinum wire.
The wet/dry situation is a measure of resistance. After the tank was
drained, one sensor showed wet but 3 hours later showed dry. In a later
test
the sensor again disagreed with the others but 5 minutes later agreed.
Current thinking is that the vast time difference in the reading returning
to nominal indicates that this is a mechanical problem where vibration
(wind
gusts, etc.) that can produce instant effects would be causal, rather than
electrical where environmental (temperature, humidity) changes that are
slow
to build and slow to dissipate would be the culprit."

Note from : "The level sensor itself is a platinum wire.
The wet/dry situation is a measure of resistance. After the tank was
drained, one sensor showed wet but 3 hours later showed dry.


It is easy to imagine a sensor showing wet for a while after the tank was
drained. It probably would be cold enough to still allow the H2 to remain
liquid on it for a while and not evaporate.

However, going from dry to wet should be instant.

And at lauch time, the tank had already been filled for a while right ? No
excuse for an erratic sensor if the feeling is that it takes a while for
sensor to acclimatise to the new environment.

How much of the ET's insulating properties changed with the changes to the
foam ? Perhaps changes in how slowly heat can penetrate the ET in certain
areas would change how quickly the sensors could dry following the tank being emptied.

In any event, since in real flight, the tank is emptied in 8 minutes, and the
safety requirement of having engines shutdown in case of fuel starvation would
require a sensor able to react in seconds (if not less), not minutes or hours,
there is something wrong with the sensor.

"John Doe" wrote in message ...
Note from : "The level sensor itself is a platinum
wire.
The wet/dry situation is a measure of resistance. After the tank was
drained, one sensor showed wet but 3 hours later showed dry.


In any event, since in real flight, the tank is emptied in 8 minutes, and
the
safety requirement of having engines shutdown in case of fuel starvation
would
require a sensor able to react in seconds (if not less), not minutes or
hours,
there is something wrong with the sensor.


You're making a serious logic error. There's all sorts of connections
between the sensor and the computers that read it. The fault could be
anywhere along that chain.

"Dr. P. Quackenbush" wrote:
You're making a serious logic error. There's all sorts of connections
between the sensor and the computers that read it. The fault could be
anywhere along that chain.

Of course. And the glitch might even be in software way up at the top.

But the premise for this discussion was that the sensor seemed to get back to
normal a few hours after the tank was emptied, with mention of wet/dry.

With Columbia, NASA discussed a lot the differences between off-scale values
and within range values. But I haven't seen much discussion here about whether
the sensors are on-off values, or if they are gradual values with some
off-scale readings possible during malfunctions.

(for instance, if a connection was broken, it would read offscale , wheras if
the sensor though it was still wet/cold, it would render a value within a
valid range, ecven though the value is unexpected).

Sounds like a broken pin keeper in a plug. A gust one way pulls the pins
apart. A gust in another way pushes them back together. Found that type of
problem when I worked on C-5As. Hard to find because the plug looks good
when taken apart.

Suzanne Rathburn

"Dr. P. Quackenbush" wrote:
I don't understand the description of the sensor itself. I'm not a EE (I'm
software, systems, and control electronics), but I don't see how a single
wire can be used as a fuel level sensor.

I see no indication that it's a single wire - all we have to work with
is pretty high level block diagrams, impossible to get a wire count
from.

I'd have expected the sensor to
have two electrical points of contact that are physically separated inside
the tank (if you measure the electrical resistance between them you'd get a
different reading if there's fuel in the tank versus when there's not).

That's one way to do it. It's not the only way.

D.
--
Touch-twice life. Eat. Drink. Laugh.

-Resolved: To be more temperate in my postings.
Oct 5th, 2004 JDL

Dr. P. Quackenbush wrote:
I don't understand the description of the sensor itself. I'm not a EE (I'm
software, systems, and control electronics), but I don't see how a single
wire can be used as a fuel level sensor. I'd have expected the sensor to
have two electrical points of contact that are physically separated inside
the tank (if you measure the electrical resistance between them you'd get a
different reading if there's fuel in the tank versus when there's not).

I'm sure they just 'Earth' (or is it spacecraft? :-) ) one end by
connecting it to the tank wall, to provide the return path. Then they
provide current to it down the wire, and measure the current. The tank
wall is fairly thick aluminum and a very good conductor.

"Derek Lyons" wrote in message
...
"Dr. P. Quackenbush" wrote:
I don't understand the description of the sensor itself. I'm not a EE
(I'm
software, systems, and control electronics), but I don't see how a single
wire can be used as a fuel level sensor.

I see no indication that it's a single wire - all we have to work with
is pretty high level block diagrams, impossible to get a wire count
from.


See the original Oberg post. He quoted from Nasa Watch, where it was said
that the sensor was "a platinum wire".