Over-torque engine failure

[whamil77 183]

The developer coded an engine failure if maximum torque is exceeded for some period of time (it appears to be 2 minutes for the left engine or 3 minutes for the right engine).  If you reduce torque below the limit it restarts the clock.  In other words, it's not cumulative.  And all of that is fine. 

But....the number they used as maximum torque is 1484 ft-lbs which is not the torque limit of the gearboxes.  The RW performance charts include power settings up to 1628 ft-lbs and include a note stating the gearbox limit is 1628 ft-lbs.  The 1484 ft-lb limit is only associated with Vmc testing at 2200 propeller RPM.  It is not a mechanical limit.  If you want to roll the dice and blow off Vmc in this airplane and use 1628 ft-lbs and 2200 RPM to takeoff and climb like a rocket, that's fine.  The hardware is perfectly capable of doing that.

The fix is relatively easy.  In the "Gauge_UPDATE_DIG.xml" file simply replace the 1484 torque limit with 1628.

[FDEdev 955]

Changing this number to 1628 wouldn't make much sense because if you fly with the throttle fully open and with the props full forward you only get 1550ft-lb, hence no engine failure at all, never. 

Point of this failure simulation is to get the sim pilots to respect the engine limits. It's a nice feature but IRL a gearbox limit can be disregarded for many many hours before it fails.

Furthermore if you reduce the prop RPM to 2000, the engine failure limit due to overtorque climbs to 1628ft-lb.

 

[whamil77 183]

1 hour ago, J35OE said:

Changing this number to 1628 wouldn't make much sense because if you fly with the throttle fully open and with the props full forward you only get 1550ft-lb, hence no engine failure at all, never. 

Point of this failure simulation is to get the sim pilots to respect the engine limits. It's a nice feature but IRL a gearbox limit can be disregarded for many many hours before it fails.

Furthermore if you reduce the prop RPM to 2000, the engine failure limit due to overtorque climbs to 1628ft-lb.

 

To your first point....that is correct.  At 1550 ft-lb and 2200 RPM no mechanical limit has been reached therefore, no failure.

To the second point, it's fine to use a failure simulation to get pilots to respect engine limits.  But in this case it is arbitrary and inaccurate.  No limit has been exceeded, but the engine fails.  May as well roll some dice and do a Monte Carlo or better yet, set up a random failure in P3D settings.

Your third point....in the simulation if you select 2000 RPM and any torque setting above 1484 ft-lb torque (up to 1628 ft-lbs is perfectly legitimate in the POH), the left engine will fail in 2 minutes and the right will fail one minute later.  Again, arbitrary and inaccurate.  The simulation does not reset the torque limit based on RPM, nor should it.  It should have done what I did and set the torque limit to the actual mechanical limit.  The mechanical torque limit is 1628 fl-lbs.  That's why a 2000 RPM and 1628 ft-lb power setting is authorized and documented in the POH. 

There are three limits in play.

1.  Horsepower limit - An aerodynamic limit based on the airplanes ability to remain controllable with a critical engine failure at Vmc.  In this airplane the limit is 620 SHP.  That equates to any combination of torque and RPM that makes 620 SHP as long as the mechanical torque limit (1628) is not exceeded and the prop RPM limit (2200) is not exceeded (1484 ft-lb @ 2200 RPM up to 1628 ft-lb @ 2000 RPM).  These same engines/gearboxes were limited to 550 SHP in the King Air E90 due only to aerodynamic considerations (1315 ft-lb @ 2200 RPM up to 1450 @ 2000 RPM).  Any torque limit that is below the gearbox mechanical torque limit is a aerodynamic Vmc limit, not a mechanical limit.  Cheating on Vmc does not cause engine failures.  Cheating on Vmc causes control problems if an engine fails above the power limit, but it does not cause an engine to fail.  

2.  Prop RPM limit - Mechanical, metallurgical, and tip-speed limit.  2200 RPM for this airplane.

3.  Gearbox torque limit - Heat and metallurgical limits.  1628 ft-lbs for the PT6A-28. 

Setting the torque limit to 1628 ft-lbs in the gauge update makes the simulation accurate.  Users will experience fewer engine failures but the simulation will be accurate.  Users wanting to experience random engine failures not associated with exceeding the aircraft's limitations can select that in the settings tab of the simulator I believe.   

 

[FDEdev 955]

2 hours ago, whamil77 said:

To the second point, it's fine to use a failure simulation to get pilots to respect engine limits.  But in this case it is arbitrary and inaccurate.  No limit has been exceeded, but the engine fails.  

Hm, the AFM says in the limitation section, engine operating limits:

Takeoff/MCT/single engine emergency: 620shp = 1484ft-lb at 2200RPM

Max climb/max cruise: 620shp = 1628ft-lb at 2000RPM 

Note 1 says: max permissable torque is 1628ft-lb. N_P must be set so as not to exceed shp limitations.

 

So by exceeding 1484ft-lb at 2200RPM you are definitely exceeding operating limits, in this case the maximum permissable shp limit.

 

[whamil77 183]

I had this discussion with Beechcraft instructors and engineers when I went to King Air school so many years ago.  In the "flat-rated" world of turbine aviation, manufacturers and the FAA must come to grips with how and when all that extra power can be used.  It's real easy for the FAA.  They simply define the Vmc performance requirements.  And that definition has to be met at a SINGLE airspeed.  The manufacturer can opt to use more power, hence better takeoff and climb performance, but it comes at the price of a higher Vmc which will impact approach speeds and make the landing numbers worse.  It becomes a balance between takeoff/climb performance and field requirements.  The rules provide for only one certified Vmc and is defined by power and airspeed chosen by the manufacturer.  It's a shame because the ability to maintain control after an engine failure at 180 KIAS is much better than it is at 90 KIAS.  You could really have a "sliding-scale" power limit based on airspeed.  The Blackhawk and Raisbeck airplanes would be absolute rocket-ships.

It would work like this in my world.  Vmc would remain as it is, a relatively low airspeed point well below blue-line that could safely be maintained following an engine failure on takeoff or approach.  I would also define a Vmp, a speed above which the pilot could maintain control after an engine fails while producing its maximum available and allowable power.  On takeoff or a go-around, Vmc would be in play and the flat-rated power limit would apply.  However, once you accelerated to Vmp, all the available and allowable power could be used.  Turboprops would be climbing like angels.  It could easily be done but the Government (FAA) thinks we'd be too stupid to deal with it.   

Back to your point.....

Yes, they are listed as engine limitations so we don't have to do the math and try to figure out power in our head.  But you correctly state that its a power limitation which means it is a Vmc issue not a mechanical issue.  Look at the King Air A100.  Same engines, same gearboxes.  With its extended fuselage the airframe can aerodynamically handle all the power a PT6A-28 can dish out at Vmc.  The torque and power limits on the A100 are 1628 ft-lb and 2200 RPM respectively, which adds up to the thermodynamic limit of the combo which is 680 SHP.  Those same engines in a Cheyenne II can produce 680 SHP without reaching any mechanical limit.  But they are flat-rated to 620 SHP because of aerodynamic limits.  Because those limits are read on engine gauges, they are included in the engine limitations section of the POH.   

In this simulation, setting 1485 ft-lbs of torque at any RPM will fail the left engine in 2 minutes and the right engine in 3 minutes.  1485 ft-lbs torque and 2000 RPM adds up to 565 SHP, well below the 620 SHP limit. 

[FDEdev 955]

9 hours ago, whamil77 said:

In this simulation, setting 1485 ft-lbs of torque at any RPM will fail the left engine in 2 minutes and the right engine in 3 minutes.  1485 ft-lbs torque and 2000 RPM adds up to 565 SHP, well below the 620 SHP limit. 

Guess you learn something new every day. I was sure that the engines don't fail at the 1628/2000 setting. To confirm this I placed the Cheyenne onto the runway at this power setting, selected  sim rate x64 and waited 10sec...no engine failure...what I didn't know until now was that the timer apparently doesn't use the sim rate! So even at a simulation rate of 64 the engine fails after 2 'real' minutes.

This 1485ft-lb limit is definitely a serious bug in the Carenado model as it prevents the pilot from achieving the flight manual performance!

What's puzzling is the fact that apparently nobody has encountered this problem.

Apart from this issue I don't agree that if an engine failure occurs when you are exceeding an engine limit, even for the 'wrong' reason, is identical to rolling a dice. 

Pilots usually tend to observe limits and as there are many limits, only few question them. I don't know how many RW turboprop pilots have the detailed knowledge you have!

If you can run the engine at full power indefinitely (like many users do if you check the various youtube 'reviews') there will be complaints that there's no failure modelling. 

There are many items add-on designers are trying to make more 'realistic' and this item is one of them.

E.g. I really dislike the rather new introduction of the constant cockpit shaking during the take off run as one normally doesn't experience such vibrations IRL, except a few military eastern block runways. Feels more like one or more tire failures. 

I'm not sure about the gearbox limit as I don't know if Piper and Beechcraft are using identical gearboxes and as we have seen in the discussion about the igniters where you accused Carenado of simulating the wrong system, there are significant differences between these two.

 

However, I highly appreciate your detailed insights (which would be even more appreciated at e.g. PPRUNE!) :)

 

 

[sinonquoi 1]

Guys,

Thank you for this very insightful conversation, it really helps understand the realities of this aircraft.

In reality isn't NG limit also a factor (101.5% if I'm right?). In the simulation if I want to keep NG at say 99% max during the climb there is no way torque will be at limit, except maybe at a pretty low altitude. I usually set rpm at 2100 for the climb. Is that correct ?

Since you know so much about the real thing, can I throw again a basic question I mentioned in the v1.2 thread : is there a propeller deicing switch, I could not find it ?

Cheers!

[FDEdev 955]

Although N_G  is a limit there's no need to constantly monitor this gauge, but of course it's good airmanship to include it in your scan of the less important items.

Climb/approach/landing RPM is 2000, cruise between 1900-2000.

Unfortunately you can't use the correct 1628ft-lb torque for climb because the engine fails above 1485, except if you manually change the torque limit :(

There's no separate prop deice switch because the single switch on the overhead panel controls the intake deflector, air intake deice and prop deice.

[sinonquoi 1]

Thanks so much ! Yes I have modified the torque limit in the gauge so I can exceed 1485.

For the prop deice does that mean it's not a good idea to switch engine anti ice on the ground as FOD prevention since it also heats up the props ?

Thanks !

[FDEdev 955]

Using the intake deflectors as FOD protection is definitely not recommended as the manual states that the intake lip heaters will be damaged if they are operated without propeller slipstream, which means not below 1850RPM.

Btw, this applies to many turboprops, not only the Cheyenne II.

[19dcavscout 39]

For other PT6 aircraft N1, or Ng is a gauge you look at with higher altitudes. The -67 for example will top out in regards to N1 at about 18k feet (temp dependent), and its similar to that with -42s.  When climbing, I usually continue to push my torques until I either hit the TGT or N1 limit then just leave the power levers alone. Just for reference, this is in relation to flying the PT6A-67 and -42 engined King Air 200's.

[Paul Eirik 0]

On ‎5‎/‎7‎/‎2017 at 7:23 PM, whamil77 said:

To your first point....that is correct.  At 1550 ft-lb and 2200 RPM no mechanical limit has been reached therefore, no failure.

To the second point, it's fine to use a failure simulation to get pilots to respect engine limits.  But in this case it is arbitrary and inaccurate.  No limit has been exceeded, but the engine fails.  May as well roll some dice and do a Monte Carlo or better yet, set up a random failure in P3D settings.

Your third point....in the simulation if you select 2000 RPM and any torque setting above 1484 ft-lb torque (up to 1628 ft-lbs is perfectly legitimate in the POH), the left engine will fail in 2 minutes and the right will fail one minute later.  Again, arbitrary and inaccurate.  The simulation does not reset the torque limit based on RPM, nor should it.  It should have done what I did and set the torque limit to the actual mechanical limit.  The mechanical torque limit is 1628 fl-lbs.  That's why a 2000 RPM and 1628 ft-lb power setting is authorized and documented in the POH. 

There are three limits in play.

1.  Horsepower limit - An aerodynamic limit based on the airplanes ability to remain controllable with a critical engine failure at Vmc.  In this airplane the limit is 620 SHP.  That equates to any combination of torque and RPM that makes 620 SHP as long as the mechanical torque limit (1628) is not exceeded and the prop RPM limit (2200) is not exceeded (1484 ft-lb @ 2200 RPM up to 1628 ft-lb @ 2000 RPM).  These same engines/gearboxes were limited to 550 SHP in the King Air E90 due only to aerodynamic considerations (1315 ft-lb @ 2200 RPM up to 1450 @ 2000 RPM).  Any torque limit that is below the gearbox mechanical torque limit is a aerodynamic Vmc limit, not a mechanical limit.  Cheating on Vmc does not cause engine failures.  Cheating on Vmc causes control problems if an engine fails above the power limit, but it does not cause an engine to fail.  

2.  Prop RPM limit - Mechanical, metallurgical, and tip-speed limit.  2200 RPM for this airplane.

3.  Gearbox torque limit - Heat and metallurgical limits.  1628 ft-lbs for the PT6A-28. 

Setting the torque limit to 1628 ft-lbs in the gauge update makes the simulation accurate.  Users will experience fewer engine failures but the simulation will be accurate.  Users wanting to experience random engine failures not associated with exceeding the aircraft's limitations can select that in the settings tab of the simulator I believe.   

 

Hi, I am experiencing the engine failure without exceeding any limits. All gauges is in the green zone. I usually cruise with 1900-2000 RPM and 90-100% throttle, which result a Torque about 1050 ft-lbs at FL230. I have tried to increase the limit in the GAUGE_UPDATE_DIG.xml file, but I don't know how to make a cabinet file or overwrite the file. 

Do you have any suggestions to avoid the engines failing at each flight?

 

I'm flying this aircraft in a world trip, so it would be helpful to have a aircraft without any risk of engine failure.

I appreciate all answers I can get :-)