RealAir Duke Turbine Engine Failure Frequency

[superglide17 11]

Let me start by saying that I think the Turbine Duke may be the finest GA simulation on the market today. (But I do love my F1 Mustang.) I have flown the Duke B60 just about since day one. When RealAir introduced the Turbine, I jumped on it. This may be the answer to the notorious lack of an FSX B200 (or not).Now my question/comment. I have the Random Engine Failure probability set to "Very Low". According to the Flying Guide, "When set to ‘Very Low’ there is a 1/500 chance of an engine failure occurring once every minute...On average you should see an engine failure about once every 500 minutes (8.3 hours)." I have 11.5 hours on my Turbine Duke and have had 2 engine failures, both on the right engine. The first failure, I had a precursor high oil temp on the rt engine. I had the oil door open. It first showed up on the cruise climb. It did not quite go red. Later in the flight at cruise altitude, it went red and I was unable to bring it down before the engine stopped. I feathered the prop and was able to bring it in on one engine. The second failure was again on the right engine but, there was no high oil temp. It just quit in cruise flight at FL160.I have been very careful in the startup, not let the ITT temp get out of line. Flew everything in the "green" and by the numbers. It just seems that 2 failures in 11.5 hours is kind of extreme. I don't want to set the probability to "Never". I kind of enjoy the underlying anticipation of a possible failure. Any comments from you other pilots?SuperglideWin7 P64b| FSX ACCEL| ASUS P6TDV2| i7 920 @4.2| EVGA GTX470| MUSH RL 6G DDR3

[zfehr 1]

I set mine to never after having it on very low and experiencing an engine failure in cruise. I do have it set so that abuse of the engine parameters can cause it to fail. Given the extremely low incidence of failures in these engines that is quite realistic for my uses. I prefer flying approach procedures more anyway and with the RXP Garmin coupled to the flight director this is a really fun platform to fly from.

[briansommers 74]

I'm wondering how this would work with FSPassengers... Do any of you have this? I would think to set it to never then use fsp.

[Andydigital 93]

I had a similar experience in the first 30 minutes of engine run time, I then went into the configuration program and toggled it to very high and then back to very low just to assure myself it really was at very low, Ive not had anymore failures since and there is about 20 hours on the clocks now.

[robert young 2,897]

Answering the original poster's query:It is important to understand there are TWO DISTINCT failure modes. One is random and the other is not. Even if you choose a low chance of random failure you can still have one, but generally that would be signalled in advance by the annunciator panel.....not always, but mostly. But you can choose no failures whatsoever by selecting that option. This means that you can fly all day with the throttles fully forward and no matter what the gauges tell you, the Duke will not fail.If you switch random failures off, but still enable user provoked failure (excessive torque, ITT or turbine N2 overspeed), then you might have a failure, and in most cases the annuciator red lights will appear before the failure, so you have a chance to react. The engines can also fail if you do not follow start procedures.If you want a trouble free flight with no failures, so you can concentrate on navigation and getting to your destination, then you can always switch off every possible failure. (Please read the Flying Guide!!)Random failure modes are very much reliant on us trusting the parameters we are given by the FSX SDK to "play" with. At this stage it is probably far from perfect, and time will tell whether we need to look at changing the "chance" params. But it is not clear yet whether these random failures really are random since we have no objective evidence that all users are flying in a manner that guarantees all failures will be solely random and not "provoked".We are keeping a careful eye on feedback and if there are adjustments to be made, we will respond. Meantime, I have never had a random failure in hours and hours of testing, when random failures are set to "low", but I have had failures when deliberately provoking one by not opening the oil doors on a slow climb, or exceeding limits. It's quite possible I've been lucky!All the best,Rob - RealAir

[ubersu 78]

Let me start by saying that I think the Turbine Duke may be the finest GA simulation on the market today. (But I do love my F1 Mustang.) I have flown the Duke B60 just about since day one. When RealAir introduced the Turbine, I jumped on it. This may be the answer to the notorious lack of an FSX B200 (or not).Now my question/comment. I have the Random Engine Failure probability set to "Very Low". According to the Flying Guide, "When set to ‘Very Low’ there is a 1/500 chance of an engine failure occurring once every minute...On average you should see an engine failure about once every 500 minutes (8.3 hours)." I have 11.5 hours on my Turbine Duke and have had 2 engine failures, both on the right engine. The first failure, I had a precursor high oil temp on the rt engine. I had the oil door open. It first showed up on the cruise climb. It did not quite go red. Later in the flight at cruise altitude, it went red and I was unable to bring it down before the engine stopped. I feathered the prop and was able to bring it in on one engine. The second failure was again on the right engine but, there was no high oil temp. It just quit in cruise flight at FL160.I have been very careful in the startup, not let the ITT temp get out of line. Flew everything in the "green" and by the numbers. It just seems that 2 failures in 11.5 hours is kind of extreme. I don't want to set the probability to "Never". I kind of enjoy the underlying anticipation of a possible failure. Any comments from you other pilots?SuperglideWin7 P64b| FSX ACCEL| ASUS P6TDV2| i7 920 @4.2| EVGA GTX470| MUSH RL 6G DDR3

According to my calculations the probability of at least one failure in an hour is approximately 1 in 10 per hour given a 1 in 500 chance per minute and increases to 1 in 5 over a two hour period. That seems excessive. Are you sure of the 1 in 500 figure?DJ

[robert young 2,897]

According to my calculations the probability of at least one failure in an hour is approximately 1 in 10 per hour given a 1 in 500 chance per minute and increases to 1 in 5 over a two hour period. That seems excessive. Are you sure of the 1 in 500 figure?DJ

Well, "Superglide" reports that he had two failures, but from his description one of them was probably not random, but due to excessive oil temperature. The annunciator normally picks this up, but not always, especially if the temp rises quickly. So it is possible "Superpilot" had one failure that was random, not two, and that's well in keeping with the odds we publish.But as I said, it's early days yet and we are relying on the mathematical probabilities set; it's quite rare for FSX to always reflect mathematical probabilities accurately!Rob - Realair

[ubersu 78]

Well, "Superglide" reports that he had two failures, but from his description one of them was probably not random, but due to excessive oil temperature. The annunciator normally picks this up, but not always, especially if the temp rises quickly. So it is possible "Superpilot" had one failure that was random, not two, and that's well in keeping with the odds we publish.But as I said, it's early days yet and we are relying on the mathematical probabilities set; it's quite rare for FSX to always reflect mathematical probabilities accurately!Rob - Realair

My guess would be that the failure rate should be somewhere in the area of 1 in 500 hours rather than minutes... that gives a much more acceptable rate at the lowest setting. That's why I questioned the 1 in 500 per minute figure.DJ

[robert young 2,897]

My guess would be that the failure rate should be somewhere in the area of 1 in 500 hours rather than minutes... that gives a much more acceptable rate at the lowest setting. That's why I questioned the 1 in 500 per minute figure.DJ

Fair point. We might look and see if we can adjust the random settings for a future update.Rob

[briansommers 74]

But as I said, it's early days yet and we are relying on the mathematical probabilities set; it's quite rare for FSX to always reflect mathematical probabilities accurately!Rob - RealairI kind of chuckled at that comment, but when you really think about it, I think it is well, sad really.