PILOT SPIN
Pilot Zone => Accident Review/Never Again (I hope..) => Topic started by: acrogimp on January 03, 2016, 09:55:11 AM
-
So I alluded to this in my rec ride thread. I had an engine driven fuel pump failure recently, was very scary and could have been a literal disaster. Fortunately for me it was just after takeoff and was able to produce enough fuel flow for partial power operation for an immediate RTB.
I have had some time to go over the internal and external GoPro video and I think there is a real learning opportunity to review what I did right and what I could have done better that would be of help to everyone so I plan to put it all together with my self-critique notes and maybe share it as a private video.
I had a loss of power/backfire event while climbing out and made an immediate RTB. We had departed as an element #1 and #2, I was solo as #3 with SNAP as #4 on a delayed roll in his -50. Normal spool-up and takeoff, gear retracted fine and as I pulled the prop back to 82% passing 500 AGL I got a significant roughness that went into violent backfires/substantial loss of power when I advanced the throttle to try and clear it - I departed the formation, informed Tower I was making an immediate return, and recovered uneventfully.
I did not declare an emergency, I requested an immediate return, Tower asked if I needed assistance and I said no, they cleared for me short approach and landing and that's what I did. Since the plane was under power I was able to taxi back to the YakCave so I have not had nor do I anticipate any follow-up from the airport or FAA - if I had made it back to runway but engine stopped/unable to clear runway I know this would have been different.
I was initially convinced that the actual issue was a good slug of water due to the left wing tank sitting empty for a week during the recent cold snap as a result of the fuel leak I fixed previously. When it was filled from the truck on Sunday afternoon it took over 15 gallons on the Left meaning it was close to empty and had sat that way for a week during the temperature swings we have had.
After I recovered to the hangar, it was making plenty of power on the ground with no roughness or perceived imbalance when I did some high power runups with no sign of stumbles or misbehavior (~85% RPM and 7 potatoes) and no unusual smoke or other external evidence.
That it occurred on the reduction in prop speed on takeoff was most likely coincidental IMO. Only after I pulled the cowlings did I notice a fuel leak from the pump housing that indicated a failure of the internal diaphragm.
In thinking back, this was without a doubt the most challenging and threatening issue/malfunction I have had in 29 years of flying. We all like to think we will do it all right when the defecation hits the oscillation and while I did OK it was mostly reaction and in hindsight should have been a bit more deliberate. I did the whole aviate, navigate, communicate thing just fine, if I had been solo, but in formation and in a busy pattern it is a bit more complex and could have been, I think, done better.
One key takeaway is that practicing engine failures and the impossible turn did help a lot in that I made an immediate RTB decision. They say recognition/acknowledgement of an engine failure could take 5 seconds and that is how I have trained engine fail/impossible turn scenarios which I have done in all of the types I fly regularly, including the Yak.
I immediately leveled off at the first roughness (38 seconds after brake release), advanced the throttle to see if it would clear (made it worse actually) and made the RTB decision and initial turn when I got a series of large backfires - this was almost exactly 5 seconds after the initial roughness. I checked right to see if John was in position (he had called airborne about 2 seconds before the initial roughness and was well behind, no conflict) and seeing the path was clear I started a turn.
Had this been an element takeoff, I believe I would have pulled up and over if #4 were in position but can’t say for sure. I made my initial call to Tower as I started the turn/depart from formation but it was blocked – had it been an element takeoff I would like to think I would have made a call first to alert #4 but it was, in hindsight, pure reaction to depart the formation and RTB. Fortunately we were on Tower so I did not have the issue of tuning the radio to add to the rest of the situation.
I made a 45 deg bank turn (recommended in the POH) at about 500 AGL and just after passing west of Cuyamaca street - had the engine quit at any point during the turn I would have made it back to the airport grounds, barely – I did not climb much, and stayed close to best glide speed (190 KPH) using only enough power to stay airborne (this was not in my recollection a deliberate decision, I think I only actually looked at airspeed on the initial level off, and just before I went to set gear down – it was mostly head-out)– engine did not cough after the initial backfires and responded normally to power changes in the pattern, landing and taxi at obviously reduced power levels. Altitude should have been higher and in hindsight the engine was producing power, I should have climbed up to TPA or higher just in case.
Bailing out never crossed my mind due to altitude and the thought of the plane landing wherever it decided to go after I left, nor did landing on 17 which would have been, in hindsight, the better option. If the engine had quit or continued to stumble I would like to think I would have thought of 17 but am not sure I can say that definitively. Although I did get the gear and flaps, I did not make my customary callouts and video shows I did not check the gear indicator pins like normal. I did consider Highway 52 as an alternate since I was too far north for a safe return to 27 for a bit (at that point 17 would have been obvious since I would have turned into the airport immediately and would be looking right at it). Both 27’s were in use at the time, could have been nasty but I don’t recall Tower clearing any departures for 27L while I was on downwind or final.
A last point here is that again, in hindsight, I should have made a check flight following the fuel leak fix since this issue would likely have shown up when I was NOT in a formation departing the field (less risk to others).
I sumped the fuel system once back on the ground just in case but found no water or particulate contamination but am still changing my SOP from filling the tanks prior to flight to filling the tanks after flight if below about 45L per side to minimize the chance of a water in the fuel. The Yak fuel system is considered relatively impervious to fuel in water (we don't normally sump prior to flight) and even though water was not the cause for this issue it is much more on my mind than before.
Shared in an attempt to provide a review of events while still fresh in my head – hope it is useful.
'Gimp
-
Surprised the Yak have an auxiliary, electric fuel pump. Nice job Gimp!!! Very nice. That's what currency/recency will do for you.
-
Surprised the Yak have an auxiliary, electric fuel pump. Nice job Gimp!!! Very nice. That's what currency/recency will do for you.
No electric boost pump in the Yak, we do have a primer/aux pump on the instrument panel that is supposed to be able to keep the engine running but it would be very difficult to use in practice I think.
The Primer/Pump is on the right shaped like a handle, it would pumped in and out which would be a tough movement to do while also flying the plane.
(http://i11.photobucket.com/albums/a179/KahunaGrande/20150105_161120_zps259b4840.jpg)
'Gimp
-
Well done, Gimp.
But can you review the geometry for me? Did you take off on 27 and land on 09? Did you do a 225 degree turn back to the runway?
-
I made an immediate 180 and joined downwind for 27R (the runway I had just taken off from) because it smoothed out and was responding more or less normally at moderate power settings. Tower cleared me for a short approach and I basically just brought it in from a low pattern. The Yak actually only requires 1000-1500 feet on the ground to get down and stopped so I had options from the initial roughness to right base.
Had it continued to run rough or quit altogether I suspect I would have called for 17 or even 09L - in hindsight I allowed my comfort of the fact that it smoothed out to push me for a regular pattern with short approach - OK because it did clear up and run fine after the big backfire but on review I should have probably committed to 17 or 09L since I had no guarantee at that point that it would continue running.
It all worked out in the end but with the benefit of hindsight I see that I could have done better and that is why I am sharing.
'Gimp
-
Wow. Still surprised there is on electric boost pump, but it is Russian, and how often does an engine pump fail? Again, nice job!