Flying Qualities

[Ivan]

Sometimes we can learn from the kids....

My son Michael recently has become more interested in CFS and mission flying.
Most of the time he flies my FW 190A, but sometimes he flies the stock aircraft as well.

One of the stock missions involves shooting up a couple gunboats and a German staff car under the Eiffel Tower.
I believe that to properly execute this mission requires being able to fly through the base of the Eiffel Tower. He had some trouble doing that initially, so he asked Dad to demonstrate.
In doing this, I noticed that some aircraft could do it easily while others would take much more effort.
We all know this to be the case, but the Eiffel Tower appears to be a very good test of general handling.
Some aircraft such as my FW 190A and P-40s flew through easily while others such as the Macchis and Ki-61 had a much tougher time. I normally do a lot of testing for straight line performance but not too much rigorous testing for handling. Seems like we will be visiting Paris more often.

What do you all look for in handling? Which particular downloads do you believe have the nicest handling and what do you like about them?

- Ivan.

 

[arfyhun]

My favourite plane is the Avro Lancaster BI PD821, VN-E Version 2.5 released 08-20-2001. Aircraft, Main Panel and textures by Roger Lowery

It trundles along at a pace suited to my abilities.
I can get it in the air, swan around AND land it without putting the oleo legs up through the wings or me in a virtual coffin.

Fighters are not my forte. However hard I try I cannot get out of flying too shallow a glide slope.

So with the Lanc, once airborne, I can close the cloud base right down and fly on instruments chasing NDBs.

Extacy!

Graham.

 

[smilo]

reminds me of the good old days flying with the aac.
of course, we flew stock ac only.
my fave was the fw190.

we would fly figure 8's through the tower
just to see how many we could do in succession.
quite the challenge and loads of fun.

 

[Ivan]

.....Five Years Later

Hello Gentlemen,

Those were interesting times five years ago.
My Son who was a little interested in flight simulators back then is much more into console gaming today.
His questions started me on a series of experiments and development that still gets refined from time to time.

Before writing this post, I found Arfyhun's Lancaster B.I VN-E and took it for a spin.
It does seem like a pretty nice package and behaves well if trimmed properly.
With a CFS instead of FS 98 Flight Model and a reasonable Virtual Cockpit, it would be very close to current standards.
The Instrument Panel(s) layout is quite interesting and probably the best I have seen for a Lancaster.

The original reason for this thread was to find out what others thought were "Good" handling characteristics.
The responses were not quite what I had expected. I was thinking more along the lines of good control response and precise directional control such as might be required to align with a stationary object such as the Eiffel Tower (or perhaps a Carrier Flight Deck or narrow Runway).

Some aircraft are obviously better at this than others and my goal was to be able to recognize the specific qualities that made for good handling and then be able to reproduce those qualities at will.
Before heading along this path, my attempts at flight models were mostly directed at straight line performance, such as speed, and climb and other measurable factors such as turn and roll rates.

What I find most memorable about "Testing" a flight model is the informal "Fun Flying" part where one gets a general impression of ease of handling and response. How easily can I buzz the Airfield? Can I hold my direction in a Slow Roll? Does the Aeroplane want to tighten up a Turn? How easily can I make a level Turn? How easily can I control direction on the Take-Off or Landing?

These handling characteristics tend to be much harder to reproduce than straight line performance. One is also never quite sure that one's own subjective opinion is really in line with the pilots' evaluations.

Two recent / current projects started me thinking about these factors and looking for this thread that I knew I had started a while back. (I had no idea it was 5 years ago though.)
The Bell P-39 Airacobra is an aircraft with a great number of contradicting opinions.
In the air, it apparently handled very well at low altitude but was underpowered.
Its Take-Off and Landing behaviour was excellent as per the design priorities.
It also had a few quirks which I tried to interpret and build into the model.

By itself, the Airacobra would not have prompted a continued discussion on "Flying Qualities".
I also recently came across the texture files for "Smilo's Hammer" which was my old P-47D-23 model painted (I believe) by PJ Dunbar. I had not visited the Thunderbolt Projects recently except to put together a package to be able to view my new P-47D-25 model. I had not done any testing for quite some time except to confirm some performance figures by using the Autopilot.

When I tried flying "Smilo's Hammer" to see if things looked and felt right, I found that there were quite a few minor issues that I had not done back then that are now just part of the standard process. Although the Autopilot had no problems, I found that there were a few issues with Trim Settings and other handling issues.

Seems like it might be a good time to revisit some of the older projects to see how they hold up against current standards.

- Ivan.

 

[Ivan]

Trim Settings

The typical Trim Parameters in CFS stock AIR files and many add-ons are much too coarse.
It may not be possible to trim an aircraft for straight and level flight
My Thunderbolts are examples of this.
They fly well enough under Autopilot and can be thrown around reasonably well, but for a long cruise, they cannot be trimmed to fly "Hands Off". (At least *I* have not been able to do it.)
Finding a combination of trim settings that results in straight and level flight is usually quite difficult but usually one can get pretty close.
I also prefer that the aircraft be set for a slight climb at full power because it is the most common configuration unless the actual aircraft is known to behave differently.

Opinions will differ here but the general idea is to have the aircraft trimmed so that it has no tendency to pull in any direction and all direction changes are a direct result of pilot input so that aerobatics are more predictable.

Arfyhun's Lancaster actually is trimmed slightly nose down by default but the Aircraft Checklist addresses the needed trim changes to fly level. Neutral handling is probably less important with an aircraft that is not intended for aerobatics.

- Ivan.

 

[Mike71]

The typical Trim Parameters in CFS stock AIR files and many add-ons are much too coarse.
It may not be possible to trim an aircraft for straight and level flight
My Thunderbolts are examples of this.
They fly well enough under Autopilot and can be thrown around reasonably well, but for a long cruise, they cannot be trimmed to fly "Hands Off". (At least *I* have not been able to do it.)
Finding a combination of trim settings that results in straight and level flight is usually quite difficult but usually one can get pretty close.
I also prefer that the aircraft be set for a slight climb at full power because it is the most common configuration unless the actual aircraft is known to behave differently.

Opinions will differ here but the general idea is to have the aircraft trimmed so that it has no tendency to pull in any direction and all direction changes are a direct result of pilot input so that aerobatics are more predictable.

Arfyhun's Lancaster actually is trimmed slightly nose down by default but the Aircraft Checklist addresses the needed trim changes to fly level. Neutral handling is probably less important with an aircraft that is not intended for aerobatics.

- Ivan.

Flying Qualities, flight stability and control effectiveness are all an inter related ball of string, depending on what the airplane was designed to do, and has - and always will be - a set of compromises.

I was lucky enough to be selected to attend the US Navy Test Pilot School in the mid 1970's. This was a one year course, 1/2 day classroom, 1/2 day flying lab work, and it was all intense and meaningful. Studying and learning to observe/test various types of aircraft gave us a great grasp of what these things were all about.

Regarding ANY aircraft (fixed wing) in any FS program, I would make TWO important observations for readers to consider:

1. Regarding pitch trim - longitudinal trim sets a stability point for the aircraft to hold an INDICATED AIRSPEED - NOT a level altitude (a common misconception). Given a fixed weight, cg, configuration and power setting the airplane is trimmed "hands off" it will try to hold that IAS whether level, descending or climbing. Change any of the fixed parameters, the trim point will change and the airplane must be retrimmed to hold the original airspeed. Any disturbance - pilot input, turbulence, etc will knock the plane off the trim speed but stability will make it attempt to re-establish that speed through a series of decreasing oscillations over time. I have flown a lot of airplanes - 65 hp Aeronca Champs, P-3s, KC-135, C-47, tactical jets, turboprops, recip twins, etc. I have never seen ONE that holds perfect altitude for any significant length of time, even when well trimmed - without some minor occasional control inputs. Generally - and obviously - the faster the airplane, the larger and faster altitude excursions will be when the airplane tends to vary from level cruise at a trimmed speed.

2. I think there are many FS fans who may set their controllers - joysticks, etc. improperly, perhaps adjusting them for various planes they are trying to fine tune. Over many years, I have decided that it is much better to set up a controller with a 50% sensitivity and null zones of 25%. These are for my Logitech Extreme 3D Pro. Obviously other controllers - and users - will have different opinions.

A good null zone is important; it allows the pilot to rest his / her hand on the control column without accidental disturbance, just like having a little freeplay in the steering wheel of your car. It also helps prevent "cross- coupling" of inputs, such as accidently putting in a little aileron in a high-G pullup because you didn't pull the control column perfectly straight back. Too much null zone or "free-play" results in decreasing control predictability and "rachety" control inputs.

I then use the [Flight Tuning] section in the aircraft.cfg file to tailor trim effectiveness and control sensitivity in all three axes for any individual airplane type. This has satisfied me over many years in FS5 through FSX-SE and P3D. If a certain plane does not have a [Flight Tuning] section you can add one and it will work 99% of the time. A typical one looks like this:

[flight_tuning]
cruise_lift_scalar = 1.0
parasite_drag_scalar = 1.0
induced_drag_scalar = 1.0
elevator_effectiveness = 1.0
aileron_effectiveness = 1.0
rudder_effectiveness = 1.0
pitch_stability = 1.0
roll_stability = 1.0
yaw_stability = 1.0
elevator_trim_effectiveness = 1.0
aileron_trim_effectiveness = 1.0
rudder_trim_effectiveness = 1.0

The value 1.0 is a default; less and more than 1.0, or any given value creates a corresponding change in the parameter.

I will now rest my ageing brain and let others wax eloquent on addition ideas.

 

[Ivan]

Hello Mike71,

Thanks for your comments, obviously based on real-world experience as well as experience in the virtual world.
Hopefully this discussion will get more interesting than it would have as a monologue.

I do understand that different types of aircraft would have different stability and response characteristics that are considered "good behaviour". I was thinking along the lines of the typical air superiority or escort fighter of the WW2 era. Most aircraft of that type would be expected to be able to be trimmed to fly formation with pretty minimal control inputs.
As for the condition of the aircraft, most of the factors are pretty well defined: Military Power (non-WEP), Full Ammunition, Full Fuel (most of the time). As for exact Altitude, I am not that particular but usually it is around 2000 feet. This is not a realistic cruise condition but does give "reasonable" trim settings to use when maneuvering at low level.

One of the things that makes this difficult at times is that in most of my flight models, I tend to push the CoG back pretty close to the Neutral Point (or where I estimate it to be) and longitudinal stability is pretty marginal. In real life, this is supposed to be optimal for performance at the cost of easy handling. I am actually not so sure that in Combat Flight Simulator it is reflected in the modelling.
This kind of configuration in the AIR file tends to create a tendency toward a long period oscillation and not a lot of longitudinal stability.
Thanks for the comment about "Trimming for Indicated Air Speed". This is the kind of thing I know I have seen in books but just had not connected what I was reading with what I was seeing in the simulator.

I believe the Trim Settings that you are describing in the Aircraft.CFG are modifiers for the values that are in the AIR file and work with the more modern simulators. With the older simulators such as CFS, we just do the editing directly in the AIR file (Recprd 516 and 1101). It is also the kind of thing that the designer should get as close to "Correct" as possible.
The kind of issue that I was originally getting at was that if the notches of trim are too coarse, one could from a constant Left bank to a constant Right bank without being able to find something in between.

Regarding the Null Zones that you described in configuring the joystick for your computer:
My own setup has minimal Null Zones. Sometimes that can be a problem as a joystick wears a little.
I had one joystick (that I bought new) that would give and intermittent Rudder input.
After I got a second joystick, I finally realised that the first one was simply out of adjustment and had been that way since it was new.
I have also had another joystick that behaved pretty well for most of a decade and then gradually develop the same problem.
The only other aspect I use for setting up a joystick is to make sure that full deflection of the joystick registers as full deflection as detected by the computer's calibration utility.
My own believe regarding Null Zones and Response is that it is the job of the Designer to manage them via the control response graphs (Records 341, 342, and 343) to give a different "feel" for different aircraft.
There is a lot more to discuss regarding tuning of these control response graphs; They were originally intended to be the main subject of this thread.

Thanks for taking the time to write a very informative response.

- Ivan.


Regarding Control Effectiveness:
That obviously has its limitations when the

 

[Mike71]

Thanks Ivan -

I take you seriously regarding CFS air files, since I have only dealt with MS and P3D sims. However my observations may still help many who do have trouble, and common misconceptions, regarding longitudinal trim and stability in general.

Combat airplanes in formation certainly are a different task than flying single ship, and do take a lot of work even when well trimmed. That is why the two basic types of today (parade and cruise) are used. "Loose cruise" allows each pilot to look around a little, but obviously not much good IFR.

It seems like your observations on tuning the CFS air file trim curves is the issue, but at that, I obviously know less than you do since I have never tinkered with the program.

Press on, and good luck.

 

[Ivan]

Hello Mike71,

I remember now where the issue of IAS and Pitch Trim came up:
"The Throttle controls Climb Rate and the Elevator controls Air Speed."
If the Elevator controls Air Speed, then Elevator Trim must also control Air Speed..... (?)
I had always thought this was a non-intuitive but obviously correct way of describing things.
I guess the technically correct way of describing things would be that the Pitch Trim is adjusted until Thrust and Drag (Air Speed) are balanced and no altitude is being gained or lost?

In this field, I am just a hobbyist who has done a fair amount of reading, so it is good to have someone with formal training and possibly a better insight into how things should work. I find that every aeroplane I try to model has taught me something new and I have been playing with flight models for quite some time.

If you are already working on Flight Models for Microsoft Flight Simulators, then you are actually working with the same parameters that I am. I believe the newer simulators just have a superset of the older variables from CFS and F 98. I have actually been using SDK's for the newer simulators to work on CFS "stuff".

- Ivan.

 

[Mike71]

Hello Mike71,

I remember now where the issue of IAS and Pitch Trim came up:
"The Throttle controls Climb Rate and the Elevator controls Air Speed."
If the Elevator controls Air Speed, then Elevator Trim must also control Air Speed..... (?)
I had always thought this was a non-intuitive but obviously correct way of describing things.
I guess the technically correct way of describing things would be that the Pitch Trim is adjusted until Thrust and Drag (Air Speed) are balanced and no altitude is being gained or lost?

- Ivan.

Ivan -

ABSOLUTELY - pitch controls IAS. Excess power at any held IAS results in a climb with a rate proportional to the excess power required for level flight at that IAS, and vice versa for rate of descent.

I would not think of pitch trim as a quasi- independent variable; it is only a means of reducing pilot control force when he holds the control column (and therefore the elevator/stab) at a position required to hold the desired pitch angle that gives the target IAS. This is the case no matter in a climb, level or a descent. Reducing the force to near zero helps in making small, precise changes in pitch necessary to hold the target speed; unless you are far more perfect than average humans, no one can control pitch within say plus or minus half a degree for very long, while also executing other tasks.

In level flight at constant IAS, the impression typically is to hold altitude with elevator trim; however the reality is, if you need nose up trim, the airplane is actually trimmed for a higher IAS than cruise and it is trying to seek it by descending.

If the thrust is changed in a climb, unless the thrust centerline and several other minor factors all act through the center of gravity, you will have to make some trim change as well as pitch change to hold the same target airspeed.

The military stresses to no end, in flight training, to control airspeed with pitch, ROC or ROD with power and CONSTANTLY TRIM, TRIM, TRIM!!

This whole ball of wax does get confusing, but in the end, professional civil and military pilots always stay out of trouble by practicing the religion of pitch = IAS and throttle = ROC/ROD. it gets a little muddy at times because you have to work both of them together rather quickly and smoothly in regimes like an instrument approach or slow flight. You see your IAS decreasing and add power - a normal reaction. BUT, at the same time you decrease pitch a little to maintain a given ROD/ROC; you are really thinking backwards, but that is normally "sorta OK".

They both work together and after time and experience it becomes pretty much automatic. Unfortunately, many landing accidents result in an airplane getting very slow and full power does not decrease ROD quickly enough and the pilot keeps trying to ease the nose up to help - but induced drag goes up rapidly as well -- bang into the ground. Most jet planes (airliners, bizjets / military) can easily develop rates of descent of 2000 fpm and more at high pitch angles/slow speed at FULL power. Decreasing pitch at that point, at low altitude is a loser because the ROD will increase, too close to the ground. The secret - don't get into that situation.

 

[Ivan]

I would not think of pitch trim as a quasi- independent variable; it is only a means of reducing pilot control force when he holds the control column (and therefore the elevator/stab) at a position required to hold the desired pitch angle that gives the target IAS. This is the case no matter in a climb, level or a descent. Reducing the force to near zero helps in making small, precise changes in pitch necessary to hold the target speed; unless you are far more perfect than average humans, no one can control pitch within say plus or minus half a degree for very long, while also executing other tasks.

Hello Mike71,

I do understand that in reality Elevator Trim really just affects where the Elevator center position is, but in the simulator it doesn't quite work that way. It acts as if it were a second pitch control.
There are a couple ways to confirm this.
The easiest is to note that if full control deflection does not allow something to happen, then trimming the control in that direction will not have the effect of adding addition control deflection.
Thus, unless aerodynamic forces limit control deflection, trim should make no difference, but in the simulator it does make a difference.
If you set your joystick to have a fair sized Null Zone at the center and try to get the joystick centered, doesn't that defeat the purpose of being able to make small, precise changes?
Along the same lines, the Blue Angels, when flying the F8F Bearcat, were described as trimming their aircraft so that a constant back pressure on the stick was required for level flight. The idea was that a change in pressure was more precise than a change in direction of control force.

At the moment I am having some issues in trying to set up what should be a simple AIR file.
The usual methods do not appear to be working and I need to figure out why not.

- Ivan.

 

[Mike71]

Hello Mike71,

------If you set your joystick to have a fair sized Null Zone at the center and try to get the joystick centered, doesn't that defeat the purpose of being able to make small, precise changes?
Along the same lines, the Blue Angels, when flying the F8F Bearcat, were described as trimming their aircraft so that a constant back pressure on the stick was required for level flight. The idea was that a change in pressure was more precise than a change in direction of control force.-----------

At the moment I am having some issues in trying to set up what should be a simple AIR file.
The usual methods do not appear to be working and I need to figure out why not.

- Ivan.

When I referred to my null zone settings in FS, that was for my preference and my joystick. You are right, there is no easy answer for too large or too small. I do know that the setting scales in FS (the visual sliders) do not seem to represent the actual result very well. In real life, military airplanes have design specifications regarding "free play" and associated control response. I would have to see what FAA Fed Regs say, I have never seen HQ's discussed to any great degree by the FAA, it is mostly performance requirements it seems to me. However, I have never really researched it.

WRT the Blues and F8F with nose down trim - were they talking about flying inverted? Also, the F8F did not have electric trim as far as I know. Therefore the pilot would be pretty over taxed between throttle and trim manipulation going through there various maneuvers in a short period of time. They may have found a happy medium in doing it that way. They also were not going through massive speed changes compared to a jet like the F-18 or F-16 that require considerable retrimming.

When you say "The easiest is to note that if full control deflection does not allow something to happen, then trimming the control in that direction will not have the effect of adding addition control deflection.
Thus, unless aerodynamic forces limit control deflection, trim should make no difference, but in the simulator it does make a difference." -- not sure exactly what you mean by all that; an example?? What type of stab are you referring to - a simple anti-servo system, a flying tail or a trimmable tail plane?

As I said, I know little to nothing regarding mods to an air file. Flt Sims are certainly not perfect representations of real life, but they are still very enjoyable.

 

[Ivan]

WRT the Blues and F8F with nose down trim - were they talking about flying inverted? Also, the F8F did not have electric trim as far as I know. Therefore the pilot would be pretty over taxed between throttle and trim manipulation going through there various maneuvers in a short period of time. They may have found a happy medium in doing it that way. They also were not going through massive speed changes compared to a jet like the F-18 or F-16 that require considerable retrimming.

I believe they were not describing inverted flight. Otherwise back pressure would not be required to maintain level flight.
If you think about the routines that the Blue Angels fly, they actually go through a pretty wide range of speeds because they tend to do quite a few vertical maneuvers. I believe this description was about the F8F Bearcat and I don't know how it behaved through the speed range, but at one point, they also went back to the F6F Hellcat and that aircraft was known for pretty large longitudinal trim changes with varying airspeed.
Some aircraft require a LOT of trim adjustment with increasing airspeed. Others do not.
The hard part is trying to figure out what "A Lot" means when trying to create a flight model.

When you say "The easiest is to note that if full control deflection does not allow something to happen, then trimming the control in that direction will not have the effect of adding addition control deflection.
Thus, unless aerodynamic forces limit control deflection, trim should make no difference, but in the simulator it does make a difference." -- not sure exactly what you mean by all that; an example?? What type of stab are you referring to - a simple anti-servo system, a flying tail or a trimmable tail plane?

Your question about the actual structure and mechanism of the Tail Plane is beyond the scope of what I believe the AIR file can represent.
The representation in the AIR file is of performance and not the actual mechanism,
In representing the actual physics, there are some problems.

Here is an example to illustrate the situation I was describing:
In my flight model for the A6M2 Model 21 Type Zero fighter, the aircraft gets nose heavy as speed is reduced.
(I am going to make up some numbers here because I can't remember what the real ones were.)
If one tries to maintain level flight with the gear and flaps down, at about 80 knots, the nose will begin to drop even with the stick all the way back.
....BUT, by dialing in some Elevator Trim, there is enough control authority to keep the nose up until the actual aerodynamic stall at about 55 knots.
At the time I worked on that flight model, I didn't know any better. Actually I probably did, but just never connected the dots.
I haven't gone back to fix the issue because it is one of the least broken of the flight models I have worked on and there are other issues I need to prepare for before revisiting it.
Recognition that there were a few broken things is why I never released the follow-on A6M3 Model 32 version.

Another interesting thing about Flight Simulator physics is what happens when we taxi at medium / high speed in a tail dragger with the tail wheel down and pull back on the stick. What should happen? What actually DOES happen? Doesn't make sense, does it?

- Ivan.

 

[Mike71]

I believe they were not describing inverted flight. Otherwise back pressure would not be required to maintain level flight.
If you think about the routines that the Blue Angels fly, they actually go through a pretty wide range of speeds because they tend to do quite a few vertical maneuvers. I believe this description was about the F8F Bearcat and I don't know how it behaved through the speed range, but at one point, they also went back to the F6F Hellcat and that aircraft was known for pretty large longitudinal trim changes with varying airspeed.
Some aircraft require a LOT of trim adjustment with increasing airspeed. Others do not.
The hard part is trying to figure out what "A Lot" means when trying to create a flight model.



Your question about the actual structure and mechanism of the Tail Plane is beyond the scope of what I believe the AIR file can represent.
The representation in the AIR file is of performance and not the actual mechanism,
In representing the actual physics, there are some problems.

Here is an example to illustrate the situation I was describing:
In my flight model for the A6M2 Model 21 Type Zero fighter, the aircraft gets nose heavy as speed is reduced.
(I am going to make up some numbers here because I can't remember what the real ones were.)
If one tries to maintain level flight with the gear and flaps down, at about 80 knots, the nose will begin to drop even with the stick all the way back.
....BUT, by dialing in some Elevator Trim, there is enough control authority to keep the nose up until the actual aerodynamic stall at about 55 knots.
At the time I worked on that flight model, I didn't know any better. Actually I probably did, but just never connected the dots.
I haven't gone back to fix the issue because it is one of the least broken of the flight models I have worked on and there are other issues I need to prepare for before revisiting it.
Recognition that there were a few broken things is why I never released the follow-on A6M3 Model 32 version.

Another interesting thing about Flight Simulator physics is what happens when we taxi at medium / high speed in a tail dragger with the tail wheel down and pull back on the stick. What should happen? What actually DOES happen? Doesn't make sense, does it?

- Ivan.

The reason I ask about the type of stab arrangement is that a trimmable stab and separate elevator could cause the effect you describe.

 

[Ivan]

The reason I ask about the type of stab arrangement is that a trimmable stab and separate elevator could cause the effect you describe.

Thanks Mike71,
That is an excellent point. I hadn't thought about that.

In your prior post, you mentioned FAA guides for handling qualities and performance.
I don't believe those would be particularly useful for what we are doing here because we really are not designing "new" aircraft but instead are trying to mimic existing aircraft and their known handling characteristics. Where those characteristics are not known is about the only place we get to fill in the blanks.
In addition, as I see it, the FAA is concerned with safety while we are more interested in fighting qualities or how easily the weapons can be pointed at an enemy.

The original topic that I had in mind was trying to figure out why some CFS aircraft were able to fly through the base of the Eiffel Tower without any difficulty while others were quite difficult to line up.
In the original post, I had mentioned the Kawasaki Ki 61 and how it was quite difficult to guide through the Eiffel Tower.
Its descendant was the Ki 100 Army Type 5 Fighter.
The Ki 100 wasn't particularly fast nor did it have a great deal of engine power but seemed to do relatively well in combat.
The Japanese flew it against the Ki 84 Hayate which was one of their best fighters for straight line performance and found the Ki 100 to be the superior aircraft even though the Ki 84 had significant advantages in speed and climb and engine power and the wing loadings were not that different.
If the straight line performance was inferior, then the difference had to be how the aircraft handled.
This started me looking at the places that I could tune a flight model for a different "feel".
There actually are quite a few places, but the easiest to tune and easiest to explain is what I am calling "Control Modulation".
(Perhaps there is a more proper term for this.)
This can be adjusted using the Control Response Graphs - Records 341, 342, and 343.

.......

 

[rpetty]

For Mike71 I was curious with your background do you feel that the flight test techniques used in actual flight tests for handling qualities can be replicated with some degree of fidelity in MS FSX?

I used to be a member of SimNASA and was in the process of developing FTTs in performance tests for FSX. Do you think the Harper Cooper scale (modified) could be of use?

 

[Ivan]

The Focke Wulf 190A

Many many years ago, I worked on a FW 190A for Combat Flight Simulator.
One of the characteristics of the FW 190 series is that they have a relatively high Roll Rate (about 150 degrees per second).
The FW 190 is also a fairly heavy aircraft, so its moments of inertia are not extremely low.

To be able to get the high roll rate and have it controllable, the Aileron Control Force must be fairly high but the Roll Damping also needs to be balanced so that it is controllable when the stick is centered. (My opinion.)
Using the typical Aileron Control Graph (Record 342) would give us the high roll rate at full control deflection but makes precise control of the bank angle very difficult.
This is the same situation as having the mouse adjusted too precisely and not being able to select a particular character on the screen.
The solution that used at the time was to adjust the graph so that toward the center, a small movement of the joystick resulted in only a small control effect but the curve gradually changed so that at full joystick deflection, one would get the full control effect.
This actually worked out pretty well.

This idea obviously would work (where appropriate) with the other control axes though the tuning tends to be a bit subjective.

- Ivan.

 

[Mike71]

For Mike71 I was curious with your background do you feel that the flight test techniques used in actual flight tests for handling qualities can be replicated with some degree of fidelity in MS FSX?

I used to be a member of SimNASA and was in the process of developing FTTs in performance tests for FSX. Do you think the Harper Cooper scale (modified) could be of use?

I don't think real life test techniques can normally be used, because the repeatability of results across the entire spectrum of FS users is dependent on such a wide range of controllers and their very different qualities/adjustments. That is why I personally use the [flight_tuning] section of the aircraft.cfg file to establish what I consider quasi-realistic FQ's for individual aircraft. I now know via Ivan's comments that this is apparently not a successful way to go in CFS programs apparently.

Real life performance testing on the other hand is useful in establishing quasi-realistic fidelity.

 

[Mike71]

Thanks Mike71,
That is an excellent point. I hadn't thought about that.

--------In your prior post, you mentioned FAA guides for handling qualities and performance.
I don't believe those would be particularly useful for what we are doing here because we really are not designing "new" aircraft but instead are trying to mimic existing aircraft and their known handling characteristics. ---------
In addition, as I see it, the FAA is concerned with safety while we are more interested in fighting qualities or how easily the weapons can be pointed at an enemy.

The original topic that I had in mind was trying to figure out why some CFS aircraft were able to fly through the base of the Eiffel Tower without any difficulty while others were quite difficult to line up.
In the original post, I had mentioned the Kawasaki Ki 61 and how it was quite difficult to guide through the Eiffel Tower.---------------------

.......

I only mention FAA matters because they are apparently lacking - i.e., they certainly address safety across a wide spectrum, specifically as a performance requirement. HQs on the other hand seem to be limited to stall characteristics as best I have seen.

With respect to flying under the Eiffel Tower, I would note that, in my opinion, "tight in the loop" controllability in FS programs is limited by the 2D virtual world. peripheral vision and depth perception are critical in such a maneuver, and the inability of FS programs to realistically simulate this makes these types of tasks very difficult and lead to overcontrol in many instances. I firmly believe this is one reason one why so many pilots find making reasonable landings (especially flaring/lineup) in FS difficult and discouraging.

Equally important are kinesthetic senses - the subconscious awareness of g-forces such as an increase in sink or climb rate, yawing, etc. These sensory values may be subtle, bad add to a pilots ability to monitor his entire situation. The "momentum efect' in VC mode in FSX, Steam etc. is a noble attempt to simulate these, but in my opinion, are severely limited and somewhat inaccurate in many cases.

 

[Ivan]

I only mention FAA matters because they are apparently lacking - i.e., they certainly address safety across a wide spectrum, specifically as a performance requirement. HQs on the other hand seem to be limited to stall characteristics as best I have seen.

Actually if this is the kind of thing you are interested in, there are quite a few NACA Wartime Reports on specific fighters that seem to be evaluating them against a very detailed standard. I used one of those reports when working on the P-39D/F Airacobra flight model. The specific report that I used is NACA Report L-602. There is much more detail in it than I know how to model and there are some characteristics that I found I had to under-model in order to make the CFS model more flyable.

With respect to flying under the Eiffel Tower, I would note that, in my opinion, "tight in the loop" controllability in FS programs is limited by the 2D virtual world. peripheral vision and depth perception are critical in such a maneuver, and the inability of FS programs to realistically simulate this makes these types of tasks very difficult and lead to overcontrol in many instances. I firmly believe this is one reason one why so many pilots find making reasonable landings (especially flaring/lineup) in FS difficult and discouraging.

While I agree with you that there are some serious vision issues with any desktop flight simulator on a single screen, I don't believe those are really the dominant factor in flying through the Eiffel Tower. The reason I do not believe this is because with certain aircraft, it is easy to do this and with others, it is quite difficult and it is the same pilot, controller, simulator program and computer being used.
Doing a proper flare in landing is fairly difficult in CFS at least for me because the "Ground" is not very detailed and all the little cues that we would see if flying in the real world such as bushes, rocks, patterns on the runway, blades of grass, etc. are just not present. The distance to the ground is hard to confirm.
You are right that peripheral vision is a serious limiting factor. Most people have around 180 degrees of binocular vision and in the game we have about 90 degrees of MONOcular vision.

Equally important are kinesthetic senses - the subconscious awareness of g-forces such as an increase in sink or climb rate, yawing, etc. These sensory values may be subtle, bad add to a pilots ability to monitor his entire situation. The "momentum efect' in VC mode in FSX, Steam etc. is a noble attempt to simulate these, but in my opinion, are severely limited and somewhat inaccurate in many cases.

I can't address FSX because I have no experience with it. The rest of the missing sensations is just what we give up with a desktop simulator and even the full motion simulators don't give that realistic a feel.
I figure this is one of those situations where we just have to go with what we have.

- Ivan.