Why did this jetliner crash

I just watched a program on the National Geographic channel that showed a Peruvian 757 jetliner crash into the ocean due to faulty information from their on board computer.

The pilots were in constant contact with the air traffic controller, but all this guy did was read the same bad data back to the pilot. My question is didn’t the air traffic controller have the plane on radar? And if so why didn’t he check the faulty and contradictory data with what the radar was showing?

Someone will be along directly to correct me if I’m wrong, but as I understand it, radar only tracks distance, not altitude. The airplane itself uses a transponder to send its own information to the air traffic control system, including the airplane’s identity and altitude.

If the transponder is disabled, all the control sees is a blip at X point on the radar screen. And if the transponder is getting faulty information from the onboard computer, that’s what it’s sending out.

It wasn’t bad navigation that caused the crash, it was an instrument problem. Some of the pitot tube static ports were left covered with tape after the plane was washed, so the pilot’s airspeed and altitude showed as too high (the co-pilot’s instruments were correct). The pilot reduced speed and altitude trying to correct his readings until the plane crashed.

I’m not sure what the status of the aircraft’s transponder was for this accident. The accident also happened within 5 mins of takeoff.

Here is some additional information regarding the crash:

http://www.airdisaster.com/cgi-bin/view_details.cgi?date=10021996&reg=N52AW&airline=Aeroperu

The problem wasn’t with the computer more so than the instrumentation itself. Even if the controller had altitude information, it wouldn’t have helped matters any, given the amount of instant and continuous information the pilot needs in order to fly the aircraft safely. The weather is not mentioned, but I’m guessing IFR conditions because the pilot obviously could not get visual cues necessary for “seat of the pants” flying.

Also, if he was flying over open water, he probably didn’t have many visual markers even assuming he was paying attention for them. I can imagine a pilot becoming too trusting in his instruments and just forgetting to look out the window from time to time. That said, water at 100 feet looks an aweful lot like water at 1000 feet. It’s flat and bluish. Kinda shiny too.

The radar data I used to work with contained range, azimuth and alititude.

It’s amazing to me that air traffic controllers can’t determine altitude from sources other than the planes transponder. These guys are stacking planes in holding patterns all the time and it would seem a problem like this could cause a major multi plane catastrophe. Plus I thought a top flight controller could actually talk an aircraft down if its instruments were malfunctioning.

Assuming the pilot can find his way to the airport, he’s often close enough for visual observation, which would make it easier for the ATC to talk him down. If he’s not near the airport, but high enough to be picked up on radar, they can guide him towards the airport.

But yeah, most airports (I don’t know if it’s all, but most) aren’t equipped with radar that can determine altitude. They get this from the transponder in the airplane, which reads off the instruments which are calibrated to whatever atmospheric pressure setting the ATC tells the pilot to set it to. Normally the ATC will tell the pilot to calibrate his instruments to whatever the local barometric pressure is, but any arbitrarily chosen calibration setting would work fine for purposes of maintaining safe seperation between planes as long as the same setting is given to every pilot.

In the absence of instructions from the ATC for instrument calibration, the pilot will usually use whatever setting is listed on his flight charts, adjusted for current weather (IIRC, IANAPY)

What kind of radar set did you get that from? It’s been ages since my radar courses in the navy but I’m not seeing how a search radar with a PPI sweep can generate azimuth and elevation data unless it is also sweeping or lobing vertically as well.

They were kind of busy figuring out what was happening at the time: http://www.avweb.com/other/peru603.html

At night it looks kinda black.

Altair, Alcore, Pave Paws and MMW.

Now that you mention it, these were tracking radars.

I’m curious (and hope this isn’t a hijack). Why would it be bad if a plane had too much altitude and airspeed on takeoff? Isn’t it really more of a problem of getting to the right airspeed and gaining altitude?

The plane wasn’t actually going as fast as the instruments indicated. It also was not as high as the instruments indicated. They thought they were moving too fast. They were actually moving too slow. When they further reduced speed (and altitude) they hit the water.

Yeah. Getting too slow is a bad thing. Obviously being too low is a bad thing too. The combination is often disastrous.

Not really. Water looks quite different low-level to high-level unless there is nil wind and it is glassy, then you have no depth perception. Quite possible for the pilots to be concentrating on instruments though and neglecting the real world*, outside. And as mentioned above, water on a dark night is just black, no matter how close to it you are.

*I once took a German airline pilot flying in a Tiger Moth, I let him fly it for a while and he complained that he couldn’t fly it straight and level because it didn’t have an artificial horizon. I asked why he would need an artificial horizon when there is a perfectly good one outside the cockpit. He was out of his element a little.

“Talking [them] Down” as you refer to it, is actually known as Precision Approach Radar (PAR). A PAR approach requires a properly trained (and current) controller, ditto for the pilot. PAR radar arrays are different from the sweeping-in-a-circle antenna you are accustomed to seeing at your local airport. It uses 2 antennas, one scanning in the vertical plane, and one scanning in the horizontal plane. Unlike the Surveillance Radars used by ATC, the PAR can only scan about 10 miles. Its azimuth is only 20 degrees, and elevation detection only thru an arc of about 10 degrees. Since these are used mainly as a last resort, the communications betwixt ground and pilot are kinda unusual. Instead of “Cessna 275, fly heading 220 degrees, descend and maintain 2100 ft”; It’s more along the lines of “Cessna 275, make all turns half-standard rate”, “Cessna 275, start turn”, “Cessna 275 stop turn”, “Cessna 275 increase rate of descent”, etc. etc. Generally the controller talks pretty much nonstop, with strategic gaps in the conversation to allow the pilot to talk. If the plane’s doing what’s expected, it’s assumed the pilot is hearing the controller OK. As you can tell, it pretty much requires a ratio of one pilot to one controller, so it’s only used in the case of an equipment (instrument) failure. It can also be used to guide a non-instrument rated pilot who’s blundered into low-visibility. IMO, it’s actually easier to fly than using the ILS instruments.
Eep. Sorry for the long post. I got into “instructor” mode and started pontificating. :slight_smile:

This accident was a full-meal-deal of confusion on everyone’s part.

To start with, the static ports were still taped over from an earlier paint job. These ports are necessary for the instrumentation onboard - without a static port the airplane can’t reliably determine airspeed or altitude. (The whys behind this aren’t important here - if someone wants to know more there are plenty of people on this board who can explain it!)

So the crew takes off into the Peruvian night and immediately turns out over the water (Lima has those nasty Andes mountains to the east). At night overwater with little or no moon means almost no outside visual references.

As they climb higher the taped-over static ports become more of a problem. Basically their airspeed indicators started varying with their altitude - the higher they climbed, the faster the airspeed indicators read. What really introduced confusion was that all the airpseed indicators were reading the same - usually a “bad” A/S indicator will read differently than the others, thus identifiying it as the bad one. These guys had all indications telling them that they were going too fast as they climbed. In fact they got the overspeed warning horn - a loud, annoying clacker that tells you to slow down. So the pilot reduces power and brings up the nose - proper things to do in case of an overspeed.

Except that they weren’t going too fast - they were close to being way too slow. As they pulled the nose farther up the airspeed decreased more, and the Angle of Attack (AoA) increased to the point where the stall-warning (stick-shaker) actuated. Now these poor guys have two warnings in the cockpit - one telling them they are going too fast, and another telling them they are going too slow.

At night, overwater with conflicting loud alarms in the cockpit: a recipe for confusion. In desperation the pilots asked ATC what altitude and airspeed he showed them at. As others have pointed out, ATC radar cannot determine this on its own. It needs data sent to it from the airplane’s transponder. In this case the airplane’s transponder was working fine and sending data - the same BAD data that was being displayed in the cockpit. So the controller reads his display and tells the pilots they are doing exactly what their instruments say they are. The panic and noise in the cockpit were surely heard by the ATC guy, making him wonder what was going on. With unreliable data being sent to him, though, he couldn’t help a whole lot.

Eventually we all know what happened - the pilots continued to slow the airplane until it stalled, and they spun into the ocean.

With the luxury of hindsight and sitting in a well-lit room, we can find ways to prevent this from happening again. Obviously not taking off with the static ports taped over is first, but what about airborne with the problem? If the guys were head’s up enough they could have looked at their FMS for a groundspeed readout - systems based on inertial nav or GPS can give you reliable groundspeed indication without relying on other aircraft instruments.

But what really would have saved these guys was basic pitch-and-power settings. The old “does this make sense?” test. Does 20 degrees nose high and idle power seem like something that could overspeed a 757? Not to me either.

This accident has been studied exhaustively, and basic pitch and power settings are part of training in airline courses (with renewed emphasis).

Sorry for the long post - I got on a roll!

I thought you couldn’t do that with a 757? :stuck_out_tongue:

Looking at the radio altimeter wouldn’t have hurt either.

I seem to recall there was a pilot who could barrel-roll a 747 (the BIG honking planes).