Kat,
I remember reading that amazing story on some other site, can`t remember where.
My question: what caused this abrupt engine failure?
Ultra Long Hauler
Well, it's pleasing to hear that I am now known as "The Kat" rather than just plain old "Kat" ;)
I shall copy and paste my engine failure at night experience, and maybe tweak it a bit. Here goes:
Last night my instructor and I were doing the final night requirement for my US PPL. We decided to combine it with some IR work, and filed IFR down from Sarasota to Naples and back. Upon taxiing out in our 172S we discovered that the right brake had failed, through the medium of braking, and violently veering left towards parked aircraft. We missed all of them, the last one in line only because the pilot pre-flighting it saw us coming, leapt out, and pushed it backwards! We returned gingerly to parking and swapped into the other 172.
An hour's flight saw us shooting the VOR approach to Naples, and then departing back to Sarasota. Just north of Fort Myers, out over the bay, the engine hiccupped once. We glanced at each other, sharing a "that's not good" expression, before all hell broke loose; the engine started running very rough, and we started to lose altitude from our assigned 4000'. I was flying, and stayed on the controls while the instructor called up the nearest airport in the Garmin 430; a private strip called Coral Creek 6nm away. There was no info on the charts about it but Ft Myers approach gave us the frequency, and informed us that it had pilot controlled lighting.
Turns out, the lighting was inop (see post-script), and the airfield is in the middle of a swamp; no ground lighting for miles around. We poured on full power and turned for Punta Gorda, 15 miles across the bay; but lighted. After about a mile, and topping out at 3500' it was clear that the engine was getting worse and would not maintain altitude, and an invisible airport was preferable to a swim so we turned back.
Coral Creek during the day:
Coral Creek at night:
The instructor brought Coral Creek back up onto the map, and switched to OBS mode; this gave us a line centred on the airport that we set to the runway heading and used as a guide. On a tight downwind we were still at 2500' so I put on full flap and extended my downwind; the instructor wanted to turn, but I was sure we were still too high and needed more space. I eventually turned a single 180 degree turn to finals and lined up using the GPS; we went down a little faster than best glide to lose the rest of the excess height, and flared when the altimeter read just above field elevation. The GPS at full zoom showed us over the runway and finally the rather weedy landing light picked out the centreline; right as our main wheels settled onto it. We had enough power to taxi off and shut down.
Things that we should have done better:
1) We should have declared a Mayday or at least a Pan. Approach control helped us all they could, but there's no point hesitating. I wanted to call a Pan but the instructor did not, which brings me on to point 2.
2) We did not brief what to do in emergencies; who would have control, etc. In the event it worked ok, with me doing all the flying and him handling the avionics, but things like the Pan decision were left unclear. In future I will do an emergency brief before departure and male it clear who will be PIC in event of an emergency.
3) We should have turned on the taxi light, in addition to the landing light. This would probably have helped us see the runway before we actually landed on it!
4) I need to familiarise myself more with the 430. I can run the basics, but a couple of times the instructor had navigated to another page (to get a frequency for example) and had then become distracted talking on the radio, and I could not navigate back to the map. I suggested fairly certainly that he should change it back and he did, but if I could have done it myself it would have helped.
A post-script: I flew back past the Coral Creek airport a few nights later. I decided to try the lighting again/ Where-as the instructor had been the one keying it before (in a fast, frantic, way), this time I tried it his way to start with (no lights came on), and then in a slow, measured way. The lights came on...if we'd tried this at the time, life might have been a whole lot easier...
I think the answer would be Yes i.e. you are much more likely to survive.
I can think of at least two cases straight off, of a CFIT, where they survived.
One was into a hill in a TB20, where the pilot said it was at about 120kt.
The other was at an unknown speed but in a C150 (?) so perhaps not all that fast, on a night PPL training flight. Badly injured but apparently now recovered.
I think both hit the ground on a gentle slope.
This is why, when I cross the Alps above an overcast (thankfully not often) I am running a topo map on the GPS. Better than nothing...
Actually I was thinking of descending at say 150kt, which would give you a whole load more options - providing you do actually become visual at all
Putting that through the calculator, I make it as the equivalent of 230 extra metres of altitude, which is certainly worthwhile.
If you meet the hillside at 65 knots, are you much more likely to survive than if you meet it at 150 knots? If the answer is 'no' then you could argue that it's a good bet to keep the extra speed.
Do you mean the one who used to fly the DA40TDi? Yes I think so.
Didn't you invite the Kat to the forum, Peter? He has an interesting tale to tell on engine failure at night.
200ft is quite a lot 
Actually I was thinking of descending at say 150kt, which would give you a whole load more options - providing you do actually become visual at all 
Doesn't sound very wise, unless you have a very definite idea of the cloudbase, and if the cloudbase is high then there is little point in having too much speed anyway.
If you want a fun debate, consider an engine failure above an overcast. Should you glide "fast" so that when you break out below you have the most options, or should you glide "slow" so that if you don't break out you have a (slightly) improved chance of survival?
One for the physicists to correct if required:
potential energy = mass x height x g
kinetic energy = 1/2 x m x v^2
70 knots = 36 metres per second 100 knots = 51 metres per seecond
(1/2 x v^2) / g = height
at 70 knots, you have the same energy as if you were dropped from 64 metres at 100 knots, you have the same energy as if you were dropped from 130 metres
The difference is 66 metres, or about 200 feet. So if you come out of the cloud at 1000 feet at 100 knots rather than 70, you have approximately the equivalent of 200 feet of extra altitude with which to find a landing place. How well you can use this extra energy will depend on the plane and pilot.
It was something I was considering the other day, flying along the coast at about 700 feet altitude, and 500 feet away; parameters chosen in order to remain legal. I kept my speed up so that if I did have an engine failure, I could turn in and have enough energy to turn and land on the fields just past the coastline. There were very few seagulls about, which I guess is the other consideration there. It felt as if the coast was within reach, though I'm tempted to do some tests at altitude.
To get minimum descent rate with a windmilling prop you need to fully open the throttle and pull the prop lever to full coarse, it makes a big difference.
I agree that too much emphasis is put in training on flying at Vbg. If you have an engine failure, your aim is not to break a gliding record. Your aim is to arrive on the ground safely, and preferably so that the airframe can be used again. The exceptions to this would be a glide to reach land, or Peters consideration of an engine failure above an overcast or in IMC. But in those situations you are generally flying relatively high and that gives you time to think, and to trim the aircraft properly.
But in a normal VFR (relatively low level) engine failure, the speed at which you glide should be dedicated by safety interest first, and will lie somewhere between your normal cruise speed and Vbg. But the exact speed is not all that important. Looking out for a suitable field, and flying an approach that allows you to reach that field with energy to spare (so you can use flaps, sideslip, S-turns and whatnot in the final stages of the approach) is far more important.
One of my instructors got that idea right. Her matra (in a PA28) was simple: From the cruise setting, turn two notches backwards on the trim and let the aircraft settle on that new speed. That will give you an excellent compromise speed with very minimal effort, and without having to look in.
(And furthermore, as a licensed glider pilot, I agree that there is no single Vbg. Your best glide speed depends both on the wind and the up/downdraughts you experience and expect, and even on the actual aircraft weight. There's a lot of knowledge about this in the gliding world, and it will be encoded in things like polar diagrams, McCready rings and such.)
