I may not be the most appropiate person to do this summary, but unless someone has time to expand on the topic, this is the "simple" scenario.
-The front landing gear controls many of the ground/air logic through two sensors (right and left wheel).
-Each one of these two sensors "signals" (energizes) a circuit (left or right) depending on their ground/air position. When the wheels have weight and the strut is compressed=on the ground. Or the wheels don't have weight/strut decompressed=front wheels are off the ground. It can happen that a strut is "overinflated" (not properly regulated) and the sensors don't quite reach the closed position (strut is not compressed-down enough) even while on the ground. It could also happen i.e. if the airplane is out-of-balanced overloaded in the back and has no weight in the front. This doesn't seem to have been the case on this flight (read below). Also, the aircraft was almost at maximun takeoff weight, so certainly all wheels were supporting a heavy load.
-Each one of this left&right circuits "signals" (powers) aprox. between half a dozen (right wheel circuit) and a dozen (left wheel circuit) relays.
-Each one of these relays "signals" (powers) between 1 and 4 devices (TOWS and other warnings, heaters, ventilation, lights, cabin pressure, etc).
-Some of these devices receive "power" (signal) from two different relays at the same time from both, right and left wheel circuits. As a result, if the relay that powers it (or the whole circuit for the whole wheel) "fails", it would still work as long as the other wheel's circuit is ok. Those devices are "redundantly" serviced, and therefore would still work correctly even if one relay or one whole circuit (left/right) "fails" (i.e. is disconnected, etc) or one wheel sensor "fails".
-One relay in the left wheel circuit (R2-5) was responsable for both, the TOWS and the RAT probe air intake anti-ice heater. It also powered two other "devices" (AC x-tie/Radio vent), but those two were redundantly serviced from another relay in the other (right) wheel circuit.
-The only circuit servicing the RAT heater and the TOWS was the "left wheel" one. The only relay servicing the RAT and the TOWS was the R2-5. They were NOT redundantly serviced.
-In the same "left wheel" circuit, another relay, R2-12, MAY (someone please confirm) be the one responsable for signaling the Data Flight Recorder the ground/air mode state. The Data flight recorder DID detect and record a change in logic state between ground/air at the appropiate time (shortly after rotation was called). As a result, it is likely that the whole "left wheel" circuit was working (relays R2-308, R2-283, R2-5, R2-125, R2-58, R2-212, R2-3, R2-2, R2-240).
-But since the RAT heater could have only received power if the R2-5 relay thought it was in "flight mode", it was probably malfunctioning (i.e. stuck/blown, electrically shortcutted/opened to the sensor, otherwise malfunctioning). The same fact would've made the TOWS inoperative in all likehood.
-Assuming an R2-5 malfunction, no other systems would've been affected except for the TOWS and the RAT heater.
-The technicians pulled Z29 circuit breaker which only disconnects the RAT probe heater, w/o affecting other systems. They interpreted the MEL allowed for this. That would've "taken care" of the only problem reported by the pilots: "The RAT heater is on while on the ground", by turning it off completely and allowing the a/c to fly since there was no danger of ice formation in such hot weather and destinations.
The technician's actions, while not the smartest in the world, probably complies sufficiently with their jobs requirement. The interpretation to disconnect a "working" (but could be considered "erratic") device as being part of the MEL is up to semantics and interpretation. The only thing the MEL really says is that the airplane can fly with the heater inoperative if the weather and other conditions are met.
Also, nobody had told them anything about a non-working TOWS, that the crew presumably were required to check prior to all this.
Certainly maintenance guys could've put two-and-two together. Although I'm sure "nobody overly pressed them" to do a quick-and-dirty job (i.e. another airplane, transfer busses, gate, personnel, flight plan, etc were all ready for a new plane to be used for that service w/o an overly significant delay), certainly I'm sure they felt that for the comfort and interest of the passengers and the airline it was better just to quickly disconnect such a non-significant "problem" and take care of it later.
Should have not the pilots forgotten to deploy the flaps/slats exactly at that time, nothing would've come of it. I'm sure soon enough (although indeed probably not until another return flight at "night", potentially w/o TOWS unless the crew performed a test), it would've gotten noticed/fixed.
But it had to be on a hot day, with tail wind, with a virtually MAX gross that both issues aligned while a somewhat rookie (2 years, 1000h) copilot was in charge and random stall bad "luck" had them roll right a bit too much, enough to miss the runaway's course making recovery nearly impossible. The nose-up attitude tendency of MD-82 under stall didn't help, I'm sure.
So the short-short version would be: only the TOWS and the heater were "not working". The technicians disconnected only the heater. Nobody (pilots or technicians) noticed the TOWS were also not working, nobody tested them after all this, and nobody made the connection between both conditions being in all likehood closely related.
Last edited by justme69; 8th Nov 2008 at 18:50.