Osprey - very noticeable trim "inconsistency" between flight models
This aspect is most likely very old but I just started "investigating" it in a more methodical way.
When I fly via the Asobo flight model, 0 deg nacelles, the autopilot will pick some trim setting for a stable (level) flight. Say … 270 ktas … and -15% trim … at VS -65 (see image below).
If I then slightly tilt the nacelles to 1deg … and thereby disable the Asobo flight model … I usually get a very "violent" change in flight trajectory. Sometimes it is almost impossible to recover from it by just pulling the flight stick, because even at 100% "pull up" it can not overcompensate the nose dive which gets induced by a large trim value.
To me it feels like the impact of trim on the aircraft is too large in the custom flight model … or at least inconsistent with the Asobo way of calculating the control surface impact of trim.
Here is a screenshot of a regular autopilot flight situation:
Miltech Simulations @MiltechSimulationsWe are aware of this. We could dampen the effect of trimming to avoid those violent changes. However, your testing scenario is quite unrealistic/out of bounds:
- 270 is quite a fast speed (close to max speed on the Osprey), and therefore requires significant trimming to keep the aircraft stable. It is also way beyond the conversion range for the Osprey.
- At this speed, not only the trim is having a significant effect on the custom FM, but also the calculations are affected by the speed, resulting in the pitch down.
- The custom FM is designed in a way that as long as conversion occurs within the bounds (eg. airspeed around 180kts), the aircraft behavior should translate quite well between the stock FM to the custom FM and viceversa. At these lower speeds, you probably won't need as much trim, and therefore the aircraft won't pitch down as violently. That being said, some trim correction may be required as the custom FM calculates things differently.
Miltech Simulations @MiltechSimulationsWould also like to add that the effect of trimming on the custom FM is also designed so it is effective at lower speeds, therefore its sensitivity is "amplified". The effectiveness of trim on the custom FM is currently not a function of speed as that would add another factor of complexity to the equations of motion.
In reply tonenenui⬆:@nenenuiSo I did run two "extreme" test cases, with extreme trim levels and wrote down the numbers in the following table (image, see below).
All test cases:
- took place in "Clear Skies" (no wind)
- started at stable level flight via Asobo AP
- at 4000 ft … VS = 0
- with nacelles at 0 deg
- took place below the conversion speed limit of 200 ktas
- I watched the flight from the external cam and wrote down the engine RPM as displayed by the sim
- I did not touch the throttle once the stable AP flight and target RPM was reached
There are basically two tests (green starting condition) for which I tried the following two steps:
The first was A … where I kept the AP ON:
- A1 … stable AP level flight
- A2 … increase nacelles to 1 deg
- A3 … watch how Ap will adjust the trim to maintain level flight
- A4 … watch the airspeed until in is stable again
The second was B … where I disabled the AP and kept the trim:
- B1 … stable AP level flight
- B2 … disable AP
- B3 … watch how the stable level flight continues even without the AP
- B4 … increase nacelles to 1 deg
- B5 … wait 5 seconds and see which flight condition we get if the trim is kept as during the AP phase
Both tests show that there is an extreme reaction of the virtual Osprey to a very minor change in the nacelles angle.
I think the real Osprey does not react so sensitively to such minor aerodynamic changes.
So I would conclude that there are still considerable misalignments of the two flight models at the edge where the virtual Osprey translates between them … and that could also explain the "hit" which the Osprey often shows, then the switch happens … especially if the trim is not at 0%.
