alioth wrote:
The surging engine in a light twin would have lead to an aircraft that was difficult to control directionally as it would have gone from producing takeoff power to producing a tremendous amount of drag repeatedly, resulting in severe yaw oscillations (and as a consequence, roll as well), at 350 feet, although at least the gear would have been fully retracted at this stage. This would have made it harder to determine the correct engine to shut down as it’s no longer a matter of “dead foot dead engine”, you now have to examine the engine gauges and interpret them – which while not rocket science, does add to the time taken and workload, while in a suddenly stressful situation with an aircraft which now has terrible flying qualities thanks to an engine that keeps quitting and restarting. At this point it becomes much easier to mishandle the aircraft, which is likely to result in a high energy out of control impact which kills everyone on board.
Based on my experience of real world failures, I think that this paragraph contains some misconceptions that would benefit from examination.
Firstly, I have had quite a few engine failures and the famous John Deakin (with whom I have discussed this at length) far more than me, and neither of us has ever been assisted in any way by DLDE.
It works fine in the training situation where a throttle is pulled, but for a variety of reasons, in reality it does nothing.
When the cylinder detached on the Aztec, there was massive vibration (I was concerned that the engine would shake off entirely) but the engine kept running on the remaining five. It was immediately obvious from the RPM gauge (which was dancing between 1500 and 2200, as opposed to the other solidly at 2300) which engine was to blame, but my feet told me nothing.
On the fuel icing ones, one RPM danced again, and fuel pressure was almost zero.
On the broken MP pipe, one engine reduced power with throttle, the other remained at high power, so DLDE would have been reversed.
I would strongly suggest that people look at the gauges and forget DLDE; the gauges always tell you that something is wrong and the diagnosis is usually very obvious.
And the other point is about the surging. If it were to happen in EFATO, there would be no question but to shut down as soon as you are in control, cleaned up and at blue line. But in flight, surging is usually fuel starvation and can often be fixed by changing tanks, putting on an electric pump and maybe adjusting mixture, so you may not want your first reaction to be to shut down.
I actually think that instructors are doing a disservice by teaching DLDE. It might help get their students through flight tests, but not in the reality that we are training for.
Twin pilots I know say that the EGTs are the best instant clue as to which engine it is.
A big reason why it may not be readily apparent from other gauges is the constant speed system, whereby the governor maintains the rpm until it absolutely cannot, and this masks the issue.
The twin turbo props operate their rudder boost systems detecting differential above a certain value in the bleed air from both engines, and have good crew alert systems. I don’t think there has been a rudder boost system designed for piston aircraft.
The TP can, and have, bamboozled the rudder boost if the problem is in the FCU (fuel control unit), so not a panacea.
There are probably a lot of sweet looking twin TP for the price of a new Archer, but you would need to consider overhauling and replacing a lot of ancillary systems to ensure the safety of the twin TP is delivered as per design.
Peter wrote:
Twin pilots I know say that the EGTs are the best instant clue as to which engine it is.
Not really, it depends on the fault. APS do a whole weekend on EGT interpretation. You can have two perfectly functioning engines with completely different EGTs if the mixtures are different.
It’s always going to be a combination of potential sources, but a rapid instrument scan will tell you very quickly which one looks right and which wrong. It could be RPM, it could be MP, FF, FP or EGT
RobertL18C wrote:
The twin turbo props operate their rudder boost systems detecting differential above a certain value in the bleed air from both engines, and have good crew alert systems. I don’t think there has been a rudder boost system designed for piston aircraft.
In my experience rudder boost makes it harder to detect. It obviously helps the recovery but the rudden pedal of the dead engine side pushes your foot. In the sim you had to train yourself not to push it back.*
I find the heading bug the best solution. Step on the bug.
Then use engine instruments to detect the failure.
A point that is being missed here, is that in the case of the light twin one of the “unique” purposes of identifying the failed engine is to identify which propeller needs to be feathered. This doesn’t apply if you don’t have propellers or applies “less” if you have autofeather or a negative torque sensing system.
Ted wrote:
A point that is being missed here, is that in the case of the light twin one of the “unique” purposes of identifying the failed engine is to identify which propeller needs to be feathered. This doesn’t apply if you don’t have propellers or applies “less” if you have autofeather or a negative torque sensing system.
It isn’t being missed at all. The discussion has just broadened around how to identify the dead engine.
By mentioning how rudder boost operates, or alternatively auto feather using negative torque sensors, am trying to show that where engineers have sought automatic safety systems for asymmetric conditions were they not employing a form of dead leg dead engine?
I think the point about the rudder bias strut (at least on the only one I have flown, on the HS125) is that it is a direct analogue of engine power. If an engine were to surge, the pedals would follow and a sort of balance would be kept.
In the 125 the rudder bias was superb and following a V1 cut there was nothing to do but to maintain V2 and raise the gear.
It was a no go item, as the aircraft was thought to be uncontrollable in the event of an engine failure without it.
A more definitive answer to this perennial question (with due acknowledgement to BT) 
