I thought I'd provide some info on the FM issues being discussed and what we are doing about them. First off there are a handful of issues here and there that are being worked. However there are other issues being noted where the data and behavior is correct. Since I am not a fan of opinion or feeling and prefer quantitative results lets discuss military power behavior:
As you can see there is clearly work to be done across the flight envelope in regards to military power level flight acceleration. I have been working on a fully automated test case tool that stimulates the FM code to generate the above outputs along with FM adjustments to much more rapidly make corrections and adjustments. The nice thing is, it utilizes the same EFM API function calls that all EFM aircraft use, so it will become an extremely valuable tool across all airframes.
We are doing this for every data plot that we have and are capable of digitizing and automating to make further improvements.
In summary, here are the issues we are currently looking to tackle:
- Improving level flight military power behavior (in general too high for most of envelope)
- Improving adverse yaw behavior (not sufficiently modeled)
- Improving dynamic stall behavior (not sufficiently modeled) - along with spin behavior
- Improving accelerated stall behavior (not sufficiently modeled)
- Investigating transonic control modeling
- Increasing drag at high angles of attack
- Improving turn rate (slightly too low currently across flight envelope)
Finally, an issue I'm seeing being brought up multiple times is ground behavior when moving from 9000 RPM through 10000 RM with too quick thrust build up when advancing throttle. Per our steady state engine charts each engine outputs around 800 kg of thrust at 9000 RPM this exponentially spikes to around 1400 kg of thrust at 10000 RPM. I am still investigating this and will create an automated engine test like the mil power test above to ensure the code is generating thrust for a given RPM as expected.
The last issue I'm seeing is roll rate and inertia behavior. From all indications the inertia data is correctly applied but will generate an automated test that analyzes effective roll moment for a given aileron deflection to ensure it is behaving correctly.
My objective is to provide a summary of these test results once their are generated and any corrections applied so you can all take a look at it.