Citation down North of Kelowna

[ChrisEvans]

Just for information;

N102PT Single Pilot CJ crash in 2008 - 2 fatal - owner and her son - 6 minutes after departure

http://www.ntsb.gov/_layouts/ntsb.aviat ... 0172&key=1

[co-joe]

Just for information;

N102PT Single Pilot CJ crash in 2008 - 2 fatal - owner and her son - 6 minutes after departure

http://www.ntsb.gov/_layouts/ntsb.aviat ... 0172&key=1

I wouldn't have believed that could bring a plane down, especially with 2 other AI's in the cockpit but then I watched a guy do it in his first time in sim and despite my best efforts I couldn't get him to give me control and the instructor had to freeze the sim. That's with 2 pilots and SOP's. You can talk about AI failures all you want but they are so hard to identify at first if you are not following a proper scan or if something else distracts you. Once they build it's pretty easy to see something's wrong but by then it can be too late.

[pdw]

7200 and the problems started. The 120 knot speed from FlightAware was not accurate. It was higher than that. The -2200 fpm was accurate.

Didn't see "7200" "Flightaware" data. It seemed the "-2200 fpm" was simply difference between the two last hits: the "7000" minus "4800". In that last minute though, it appears possible the AC finishes very the steep 3000'/min climb and then descent a few seconds into it (if the first ~20 secs in that last minute of data is steep climb) ... yet the last 40sec or so is a very steepening descent then ? Not impossible there to have a total altitude change between hits of 4K, including both the up&down. Just my thoughts from the data at hand, ... the evidence for "higher" speed. (One early news article quoted a source for "8000" and another suggested 'down from 4000 in 20 sec')

[cncpc]

7200 and the problems started. The 120 knot speed from FlightAware was not accurate. It was higher than that. The -2200 fpm was accurate.

Didn't see "7200" "Flightaware" data. It seemed the "-2200 fpm" was simply difference between the two last hits: the "7000" minus "4800". In that last minute though, it appears possible the AC finishes very the steep 3000'/min climb and then descent a few seconds into it (if the first ~20 secs in that last minute of data is steep climb) ... yet the last 40sec or so is a very steepening descent then ? Not impossible there to have a total altitude change between hits of 4K, including both the up&down. Just my thoughts from the data at hand, ... the evidence for "higher" speed. (One early news article quoted a source for "8000" and another suggested 'down from 4000 in 20 sec')

Some useful information came from ATC radar data.

[RatherBeFlying]

I wouldn't have believed that could bring a plane down, especially with 2 other AI's in the cockpit but then I watched a guy do it in his first time in sim and despite my best efforts I couldn't get him to give me control and the instructor had to freeze the sim. That's with 2 pilots and SOP's. You can talk about AI failures all you want but they are so hard to identify at first if you are not following a proper scan or if something else distracts you. Once they build it's pretty easy to see something's wrong but by then it can be too late.

In my case the left seat turned us inverted at 16,000' and sat there frozen while I was shouting in his ear that he was upside down and had to roll right side up until we hit the "ground".

In a real airplane you may have to grab the controls and possibly whack your buddy in the head. Given the small likelihood of recovering from a bank past 90°, a bank going past 45° is time for drastic measures.

[ChrisEvans]

Incidentally, the investigation by NTSB on N102PT took just over three years to complete, longest I've seen in recent history.

[cncpc]

It's not clear that there ever was an AI failure.

The fact that the airplane was operating at night in instrument meteorological conditions and the departure was an accelerating climbing turn, along with the pilot's demonstrated complacency, created an environment conducive to spatial disorientation. Given the altitude and speed of the airplane, the pilot would have only had seconds to identify, overcome, and respond to the effects of spatial disorientation.

[pdw]

The Maine accident climbed on track for 38 sec and then radioed the failure ... and shortly after radioed again not knowing where to turn. So it gave the impression it's understood there was a failure of the information in very rapid forward speed in clear ice, and plane was outside in FZRA. What else can an investigation conclude about it if no proof left of a pitot system obstruction. How does the investigation differentiate from "spacial disorientation" with attitude indictor O.k. ... or disoriented from failure only ?

[cncpc]

The Maine accident climbed on track for 38 sec and then radioed the failure ... and shortly after radioed again not knowing where to turn. So it gave the impression it's understood there was a failure of the information in very rapid forward speed in clear ice, and plane was outside in FZRA. What else can an investigation conclude about it if no proof left of a pitot system obstruction. How does the investigation differentiate from "spacial disorientation" with attitude indictor O.k. ... or disoriented from failure only ?

Because there were two other attitude indicators.

[pdw]

In total blockage?

By "failure" (in Maine report) the pilot's radio message also can imply 'not useable' (all of them), so pitot blockage may or may not be realized at that point .. if in fact was still possible. The comparison looks right; in Kelowna also in climb although still in process of the light right turning in the direction of an enroute heading.

[CpnCrunch]

In total blockage?

By "failure" (in Maine report) the pilot's radio message also can imply 'not useable' (all of them), so pitot blockage may or may not be realized at that point .. if in fact was still possible. The comparison looks right; in Kelowna also in climb although still in process of the light right turning in the direction of an enroute heading.

In a lot of spatial disorientation accidents, the pilot blames the AI even though it's working fine. When the AI doesn't match up with what your body senses, a natural instinct is to think the AI is broken. The investigation didn't find any possible sources of failure for the AI.

[cncpc]

The comparison looks right; in Kelowna also in climb although still in process of the light right turning in the direction of an enroute heading.

Where do you see the right turn? In Flightaware?

[pdw]

ASN has the flightaware link for the flight.

The first 3 radar hits on there are for two minutes of flight: 360deg, 12deg and 33deg (for 4000', 7000' and 4800' respectively). For the fourth hit at the end of the 3rd minute of data, which shows already off radar, is "estimated" for "74deg". At that point of the flight it does appear (from this data) the AC was approx at the halfway point thru a gradual turning required to get onto an intended easterly heading laid out in the flightaware estimates.

[cncpc]

ASN has the flightaware link for the flight.

The first 3 radar hits on there are for two minutes of flight: 360deg, 12deg and 33deg (for 4000', 7000' and 4800' respectively). For the fourth hit at the end of the 3rd minute of data, which shows already off radar, is "estimated" for "74deg". At that point of the flight it does appear (from this data) the AC was approx at the halfway point thru a gradual turning required to get onto an intended easterly heading laid out in the flightaware estimates.

Ok. Flight Aware. Those aren't all the radar hits that took place between 7200 and 4800. The loss of control was at 7200 after a steady climb at 160 knots with everything normal and the beginnings of a slight right turn onto course for Menbo, I assume. The 33 degree heading was the last heading. It wasn't the only heading on the way down from 7200.

If the assumption is that the climb to 7200 was on autopilot, then what subsequently happened to the autopilot? It doesn't get spatially disoriented.

In a 604 single pilot Citation, do you need a functioning autopilot?

[pdw]

Can't imagine the autopilot not "functioning" upon departure in this aircraft.

[cncpc]

Can't imagine the autopilot not "functioning" upon departure in this aircraft.

Then if it was, it stopped functioning.

[pdw]

There's ones without pitch axis. In earlier discussion was surprised no-one mentioned sticky snow, as there's mention 5ft snow fell somewhere in the mountians at the time ...

The Maine accident had freezing precip (including sticky snow / coarse texture). In rereading the report, sure enough, no mention of static source-intake blockage of any kind except "parked outside in FZRA". That is saying (in report) it was spacial disorientation and verifying it was 'with perfectly functioning instruments'. If the static source is iced shut ... can say instruments are fine ... OK.

Then let's refrase: How would destroyed instruments save evidence to indicate to investigators their static source was iced or not,... seeing that any ice-blockage melts away if the crash site gets above freezing before investigative teams get there.

[BGH]

The impact of the indicator on the instrument face may leave a legible mark,thus giving a clue to what it was indicating when the aircraft impacted the ground.

Daryl

[cncpc]

The airplane climbed on track to 7200 feet and control was lost after a small turn to the right. It was never doing 120 knots after control was lost. That is a misinterpretation from the rough Flight Aware points.

It impacted at very high speed, vertically.

Any Citation pilots with an informed view on whether this climb would have been on autopilot or hand flying?

And the effect on the autopilot of pitot-static icing?

[Zaibatsu]

Likely spatial disorientation likely due lack of proficiency and recency.

http://www.cbc.ca/news/canada/calgary/t ... -1.4635541