After yesterday’s technical troubles, today’s problems were entirely down to pilot error. Quickly rescheduling from yesterday, I rushed off after work to Lyneham and was sitting in the clubroom when Roger returned from his previous sortie. He’d also had a busy day.
I booked us out with ATC, preflighted and started up – giving the engine enough time to warm up before power checks as well as tune and ident all the instruments. Roger joined me in the aircraft and we taxied off straightaway, departing off blocks at 19:20. I had asked for our clearance before taxiing, which isn’t standard procedure, and there was also some confusion because I had mixed a SID IFR departure (which involves climbing straight ahead to 1600 feet) with a VFR clearance as we departed the Zone to the North. This was resolved and we were given a squawk and departure clearance after the power checks as usual.
Errors as soon as we left the ground
With no other aircraft around, we were soon airborne. Definitely a step ahead from yesterday where the aircraft had failed prior to departure, but from that point it all started to go horribly wrong. With a crosswind breeze of some 15 knots, I hadn’t allowed for the drift and as we climbed out we were slowly drifting north rather than continuing on the runway track. I must check this point again – unlike VFR departures, I had thought you need to maintain runway heading rather than runway track on a SID departure. Perhaps I had been misled by reading some stuff recently from the FAA website about how they handle multiple parallel runways, requiring all aircraft to fly the heading so that all are equally affected by the drift.
[Update: I researched this point and determined that it depends on which country you are flying in. In US, Canada and Australia, IFR departures must keep the same heading (ie not allowing for drift). In all other countries including the UK, IFR departures should maintain track. This pprune thread clarifies with references.]
As we departed the Zone to the North, I climbed up to 4000 feet. I thought we had already agreed a clearance with ATC to start climbing before we left the Zone, but that wasn’t correct – luckily I didn’t start the climb until we had crossed the Zone boundary so this wouldn’t have been a failure point, but it was close for comfort. Roger had given me a written brief the previous day which involved intercepting and tracking the 300 degree radial on the ADF. I’d done similar tasks before and wasn’t expecting this to be a problem. He had to remind me that this is what he wanted, and I rotated the RBI to determine what angle to make the intercept at.
I got the intercept wrong – I can’t recall exactly what I did – but I think I mixed up which way to go and then overreacted. It’s annoying because I’ve done this successfully several times before. I’d been taught by different instructors, some of whom recommend moving the indicator card on the RBI to your current heading, and others who don’t. It took me a while to realize that I was not on an intercept track, correct the error and plan a 90 degree intercept. At the debrief, I was reminded that depending on the distance from the NDB you should select a different angle of intercept (basically the closer you are the smaller intercept is required). After a while (too long really), I made the intercept and tracked the radial with due allowance for drift (about 10 degrees). We then made a procedure turn and intercepted the reciprocal which went quite well.
Climbs, descents and recoveries under instruments
At that point, we moved on to various climbs, descents, turns and then recovery with full panel – all of that went pretty well, although I perhaps my height keeping could have been a little better. I was trying to remember to do FREDA checks fairly regularly, and turn on the carb heat when descending – we did change tanks once during the sortie but probably should have done so twice. We repeated this under partial panel, which again I was fairly happy with. Roger made me close my eyes while he put it into unusual attitudes (and put on quite a lot of trim too), so you really had to read and trust the instruments to do this properly.
A quick fix
Roger asked me to fix my position. I looked at the ADF and DME, stupidly using the bearing on the rotating card rather than checking with the compass – this put me South of the field. It didn’t seem right, so after reconsidering I figured out we were in fact about 10 miles North of the field – just slightly to the North of Kemble. It always pays to do a sanity check.
Return to base for the NDB/ILS procedure
Returning back to base, I was asked to home to the NDB Beacon and perform the NDB/ILS approach followed by a low level circuit to land. I’d done this a few times before and had thought I understood what was needed. My biggest mistake was intercepting the NDB, knowing that I should approach it within 30 degrees either side of the outbound track. Since we were coming from the North, I’d need to turn right and then approach it from the West. I completely messed this up, and ended up going around in a circle before departing to the East – Roger later suggested that if I had done a hold to sort this out, then that would have been acceptable, similar to a go-around when landing. After departing the beacon and calling “Beacon outbound” with the QFE, I checked the DI against the compass to find we were some 20 degrees out. It seems my circling had probably caused the DI to drift more quickly. Using the ADF and my corrected DI, I adjusted for my error and probably intercepted the correct outbound radial at around the 8 mile distance required to start the procedure turn.
The procedure turn
Starting the descent from 2600 to 2100 feet, I also commenced a Rate 1 turn onto North (with a few degrees more to allow for drift), then straight and level for 30 seconds, then turned another 45 degrees to intercept the ILS. The glide slope indicator appeared to show (before the localizer), but the error flag also popped up. I mentioned that the glide slope was showing low, which suggested we might be closer in than we should be, but the DME was still at 7 or 8 miles so that wouldn’t be correct. Roger picked this up later, warning me to disregard any instruments when they show warning flags whether they be VOR, ILS or anything else.
I was very relieved to see the localizer start to come in but reacted too slowly and by the time I turned onto it, we had overshot, were off to the right and required further correction. I called localizer established and saw the glideslope come in nicely, starting my descent at a target rate of 480 feet (ground speed times 5). Roger pointed out that the DME can be switched to tell you current ground speed, but when we did this it seemed stuck at 73 knots. I thought I did a pretty good job of tracking both glideslope and localizer during the descent (and even managed to ensure carb heat was on). It was reassuring that when we got to Decision Height (500 feet) and I took off the foggles that the runway was clearly where it should be.
VFR low level circuit to land
My next task was for a low level circuit to land. The brief was that the cloudbase was around 700 feet with visibility down to about a mile. Roger asked me what the minimum circling height was and I said I thought it was about 300 feet (I remembered it was much less than I had expected, and knew it was less than 500 feet). I didn’t make it clear that I planned to make the circuit at 500 feet (which I did). I shouldn’t have guessed this answer – the actual figure is 410 and is on the plate – and/or made my intentions clearer.
I was a bit tired and probably a bit disorientated by this stage (which is probably how it would be after a difficult IMC recovery), so it took me a while to slow the aircraft down and put in 2 stages of flaps. The DME was showing about 0.9 to 1 mile as I made a circular low level track – so we would have been on the edge of visual range – descending a bit late and not compensating properly for the cross wind into what was a pretty flat but at least not a bouncy landing.
There were quite a few points which came up at the debriefing ranging from lack of FREDA checks, lack of trimming, through to the poor ADF tracking. Roger would have expected me to ident all the navigation aids immediately prior to use – I had done this prior to his boarding the plane, but with a gap of 90 minutes and our departing the area, it’s a fair point that these may no longer be correct. He also pointed out that the DME frequency for the outbound leg from the beacon is different from the ILS on the inbound leg. I had been very particular about tuning into the correct one when approaching the beacon, and had overlooked that it needs to be changed midway during the descent. I didn’t recall having to make this change on previous approaches.
The bottom line is there is still quite a bit of refinement required before I could successfully pass an IMC rating test. Time in running out for me – the few weeks that the club has left at Lyneham no longer permit weekend flying and my work commitments will restrict further evening sorties after this week. I have one more flight with Roger booked later this week and we’ll have to see whether this is enough to take the test or simply address the various errors I made today. I don’t think the instrument flying itself is the issue, so will try to have some further practice on the simulator in the meantime.
At the end of the day, it doesn’t prevent me from VFR flights but I would be more comfortable knowing that I have the option of making a (legal) instrument landing procedure should the weather unexpectedly catch me out.