To me this report doesn’t hang together. The pilot sounds like totally out of his depth, not just with a lack of the most basic systems knowledge but the most basic VFR flying. Even with the servo torque limiting clutches mis-adjusted so over-riding it was easier than it should be, it is still unbelievable that one would sit there, pushing against the pitch servo continuously for some tens of seconds, with the system yelling TRIM IN MOTION at you, and then when you let go of the yoke the thing roars upwards and you just totally lose it.
I agree with Fuji that he was hand flying, but that is also uncharacteristic of a pilot who is behind the aircraft – they get airborne and engage the autopilot immediately. So something else was going on. I suspect he was flying just under an irregular cloudbase and dodging IMC. Bizzarely though, his only VFR map (on the tablet) was not running. I have flown with many pilots who didn’t know how the autopilot worked (or it did weird things when engaged) so they avoided using it. But this is not common among owners.
IME lots of pilots have those clutches mis-adjusted, but usually they are way too tight so that a runaway autopilot would kill you. They never realise it because they never do the preflight check (pressing AP and overpowering it in pitch and roll). The dire lack of half competent avionics shops in the UK doesn’t help; IME only one guy in the UK (Harry Lees) knows this stuff.
The AAIB did a very thorough investigation job on this one and it sounds like they concluded there was a significant “pilot factor”.
This is the area. Totally nuts to fly a DIY approach into something like OVC003-OVC008 (depending on what you believe) when Exeter with an ILS is just up the road and a £30 taxi away
Peter wrote:
it is still unbelievable that one would sit there, pushing against the pitch servo continuously for some tens of seconds, with the system yelling TRIM IN MOTION at you, and then when you let go of the yoke the thing roars upwards and you just totally lose it.
I think evidence shows it’s not. (AF447 coming to mind immediately, the guy landing at Meribel gear up with the horn running the whole approach, 3 pilots on board). I think it’s easier than you think to not hear/see an alert in a high workload environment, and also fairly easy to brush off as not important (you might even have some confirmation bias saying “oh, that’s normal”).
None of this would happen when sitting at a desk or simulator or even cruise when there is not much to do.
Agreed, but that suggests that the wx was considerably worse than the “basically VFR” conditions suggested by bookworm. Consequently I think this pilot was struggling all along, only just managing to fly the plane.
The Q, as with all accident reports, is what can one learn from it? That is why I posted it.
I think there is a widespread refusal to believe that anybody who can get a PPL (which intellectually is roughly on par with a MS Word evening course at your local adult education college) will automatically understand the somewhat nontrivial avionics and other aircraft systems. I think too many people have a few hundred k but can’t grasp this stuff. They have a duty to their unsuspecting passengers to either learn it, or get a C150 or some such and fly it alone.
Peter – I agree, but I also think that solid hand flying skills are equally important. Attempting an arrival without an instrument procedure, relying on the autopilot is, I think, a dubious business. As we know a cloud break over a known airfield potentially in conditions similiar to these is possible and safe with a low level circuit to land, or a preparedness to climb away on instruments and go somewhere else. Of course, if the it was a solid undercast and the base was as low as suggested then it would be very stupid. However, if the cloud was broken with some reasonable gaps it might have been possible, but I would far rather be doing that without the autopilot, leaving aside flying some sort of made up approach with the autopilot which is an entirely different matter.
If you are going to do a DIY IAP with an MDH of something like 400-600ft you need to design it properly, by reference to a topo map, and including designing a missed approach. I have done a few of these in the past and tested them in VMC but for various reasons never flew them in IMC.
And you need to know your avionics and then using the AP to fly it is the safest way to do it because you have more brain power left to monitor it.
You will never get vertical guidance so the best you will ever get is lateral guidance and setting a VS value (if your AP has a VS hold mode) that corresponds to the GS. Same as flying any LNAV GPS approach.
But reading the report between the lines, all of this was beyond this pilot’s understanding of his aircraft systems. He seemed to know only the heading-hold and altitude-hold modes.
He probably did that approach many times and – to use Ernest Gann’s favourite phrase – survived all but the last one.
Yes, I couldnt agree more with regard to a made up approach flown on the autopilot (or even without), and this being the only half sensible approach unless the weather conditions enabled a cloud break to be made without any serious IMC. I cant recall if the report suggested he was configured to land, or not? It is a relatively quick aircraft and you dont want to be eating up the terrain if he was attempting a visual approach, or, even for that matter dealing with configuration changes late in the approach.
Dave_Phillips wrote:
Bookworm, unusually for you, I feel you’re being bit selective with the information.
Perhaps I am. But the conversation with Dunkeswell was two hours before (and also unofficial). The weather passed by ATC was 20 mins before arrival, and indicated that a VFR arrival might have been possible. I think it’s highly likely that the weather at Dunkeswell did preclude a VFR arrival. But I don’t think the pilot, who (perfectly plausibly) reported being in VMC at 2000 ft amsl when 10 minutes west, realised that until he was quite close to Dunkeswell.
Peter wrote:
Agreed, but that suggests that the wx was considerably worse than the “basically VFR” conditions suggested by bookworm. …
Totally nuts to fly a DIY approach into something like OVC003-OVC008
But he didn’t, did he? Here’s the track flown.
From his 2000 ft cruising altitude, he descended to just 1700 ft before turning away from the approach path and subsequently losing control of the aircraft. That seems perfectly consistent with someone trying to maintain VMC, realising that it wasn’t going to work, and turning away from the weather.
Or I guess it coud be someone trying to establish on the extended centreline (of a made up approach), realising he hadnt done so to his satisfaction and attempting to reposition (or had some other reason to do so)?
Fuji_Abound wrote:
Or I guess it coud be someone trying to establish on the extended centreline (of a made up approach), realising he hadnt done so to his satisfaction and attempting to reposition (or had some other reason to do so)?
Indeed, there could be many possibilities. An airspeed of 150 kt in that position does suggest that whatever he was doing he was probably a long way behind the aircraft. Perhaps he realised that that wasn’t going to work.
Reading the pilot history section of the report, my impression is that he’d spent the 3-1/2 years ‘drinking from a fire hose’. I would think it’s better to experience a series of different planes versus C172, then straight to a pressurized turbo 350 HP aircraft. Military training is similar, but that’s for kids with virtually 100% dedication of time, a high wash-out rate, and virtually no ‘creativity’ in procedures. Is there a limit to how fast an average individual can learn over calendar time, regardless of 600 hours flown in that period? Is age (50s versus say 20s) a factor in that issue?
