What Key info is need to create an FDE?

[OleBoy]

This information has been mentioned numerous times in many threads, the data numbers for weight, engine HP, airfoil, and etc.
What (if when creating flight dynamics) is the optimal info to have on hand if available?

[fliger747]

Basics are stall speeds, HP, weights, CG, all basic stuff. For fighter aircraft max speeds at sl and altitude and SL are helpful. However, if you can find the actual lift curves for the airfoil in question it is helpful. Pilots notes about stall and handling are also helpful. I am doing a FDE for a plane I have thousands of hours in as captain, and still need more info.... Depends on the level of veracity you are trying to recreate.

I recommend The utility Acraft Airfile Manager and the AFSD utility by Herve Sors. The later allows accurate testing.

Good luck! T

[Dev One]

I would recommend that you find a donor aircraft .cfg & .air file that is simple, most aircraft models seem to have a very long list of items in the .air file that seem to do nothing, but I found an ALPHA Westland Lysander - freeware - & that has quite a simple listing & it works very well modified for my current WIP. I know its for FS9, but I should think it will be OK for a FSX native. As for the camera definitions in the FSX .cfg file you can copy & paste those from A.N.Other & then adjust the position dimensions to suit.
The biggest problem I have is genarally related to engine & propellor data - there are 4 items in the .air file that can be very difficult to adjust, (as well as spin parameters), although I use Jerry Beckwiths FDWB - freeware - not sure if its still available to download as its been replaced by his payware Airwrench.
FDWB has built in data for common aerofoils.
Good luck
Keith

[fliger747]

A little more to technique: I once having weights, dimensions correct, all the airframe stuff, then get the HP right. It may be easier to steal a prop table from a similar plane. Then adjust the drag to get speeds right.

Good luck!

T

[mal998]

yep...all that and about 6 months worth of fiddling and testing.

[JohnC]

This information has been mentioned numerous times in many threads, the data numbers for weight, engine HP, airfoil, and etc.
What (if when creating flight dynamics) is the optimal info to have on hand if available?

The most time saving and quality ensuring process I can recommend is:

(1) Set the lift (i.e. Lift curve, incidence, and stall speed)
(2) Set the maximum thrust by combining discrete data real world data points with established performance trends across the range of the flight envelope (unless drag polar data is available)
(3) Tune the drag to match performance tables (typically climb/cruise)
=> Light weight high speed work well for zero lift drag, while slower speeds and higher weight work well for induced drag via Oswald Efficiency
(4) Tune the minimum/idle thrust to match descent charts.

Drag should come after thrust because the Drag force is dependent on the cross-sectional area of an aircraft, and FSX defines this automatically. So you can never set a drag coefficient in the airfile to a real world measured value and expect a realistic result,


...oh yes, and find an experienced pilot or two to tell you how the aircraft should feel in the air.

[OleBoy]

Uhh, I think I should have been more clear in my first post.

First, doing an FDE is quite a ways away. Although what I'm wanting to do is to collect the needed data from things. Flight manual? POH?
I've no idea. This Aeronca K has limited data available. The "Chief" on the other hand has a little more. That I can find.

My curiosity has me searching, but I'm not sure what I need for papers that may have pertinent info. If any.
All the info shared so far is great. It's just overwhelming!

[Sundog]

Uhh, I think I should have been more clear in my first post.

First, doing an FDE is quite a ways away. Although what I'm wanting to do is to collect the needed data from things. Flight manual? POH?
I've no idea. This Aeronca K has limited data available. The "Chief" on the other hand has a little more. That I can find.

My curiosity has me searching, but I'm not sure what I need for papers that may have pertinent info. If any.
All the info shared so far is great. It's just overwhelming!

In that regard, basically gather all of the best referenced info you can. A POH/ Flight Manual always has good references in it, such as stall speeds, how to properly operate the systems, weight limits, etc. Also books with pilot reports, etc. If you're doing a GA aircraft, search for articles from the GA mags on pilot reports. If you know someone who has flown the type and or flies it regularly, they could prove invaluable, especially if that person is actually a trained test pilot; not likely but we've had that pleasure and it does make a difference.

Now with regard to writing the FDE, back when we were making FS add-ons, we found modeling the propulsion system first and getting it right made the rest much easier. The key isn't so much what you put into the tables as values as much as what you get out. I know one of the programs listed by Fliger has a test panel, IIRC, where you could see the values being output as you were testing. I don't recall how we did it, but we first made sure the engine was putting out the correct power, etc, for a given density altitude. Once we had the engine modeled correctly, then we would work on the aerodynamics/handling/stability. But one of the problems you will run into is all of the different joysticks and the personal settings for them. I forgot what JS Ed had, but there were times he would be testing the FDE and he thought it was fine and I would try it with my Saitek and I would end up with PIO's (Pilot induced oscillations). So when you make your first plane and hand it to others to test, you'll have to take that into consideration with regard to their comments.

I've been thinking of what parameters may be needed to make a program that would calibrate a plane to a users JS, but I still need to think about how to clarify it. It would probably be a great project for Rob B. ;)

[Ivan]

Fliger747 touched on Horsepower, but I believe a bit more is needed:

I work with CFS1 but I believe the general principles apply pretty much everywhere.
Horsepower and Thrust influence a LOT of the performance of your aircraft.

For a Propeller Driven fighter, I would gather at LEAST the following information about the Engine:
Horsepower @ RPM for Sea Level and Critical Altitude.
Manifold Pressure to achieve this power.
Get several Power @ altitude data points if possible.
Decide whether you will be simulating Take-Off and War Emergency Power - With my recent A6M2 Type 0, I chose NOT to represent Take-Off or WEP.
Propeller Diameter
Propeller Pitch Range - Minimum and Maximum Pitch.
Engine Reduction Gearing
Engine Displacement & Cylinder Count

For the Airframe, a lot of it is a "Fill in the Blanks" kind of thing.
You should also get a pilot report or something describing the handling of the aircraft:
How do the controls effects change with varying airspeed?
How does the trim vary with airspeed?
How does extending Flaps and Landing Gear change the Pitch Trim?
If you don't know, extending either usually causes a nose down pitch.
Extending flaps may give enough extra lift that the nose down pitch isn't obvious because the aeroplane climbs.
The pilot report hopefully will also describe any "weird" handling, stall - spin characteristics, stability on each axis.
(An example of "Weird" would be the directional instability and yaw caused by ailerons on F-100 Super Sabre.
Another example would be for the F7F, the Rudder will roll the aeroplane faster than the Ailerons.)

You will also need the "Zero Fuel Weight" for the aircraft.
US aircraft sometimes have a "Basic Weight" specified.
If you add the crew weight (200 pounds per person), you should come close to the ZF Weight.
Another way is to find a known "Loaded" condition such as "Normal" Loaded weight and deduct the disposables such as fuel and ammunition.
Keep in mind that sometimes this "Normal" loaded condition isn't really "Normal".
(As an example, for several of the USN fighters of WW2, only a partial ammunition and fuel load is carried for "Normal" loaded; "Fighter-Overload" fills everything.
Another example is for the J2M Raiden and a couple other Japanese fighters, the Auxilliary fuel tanks are not filled for "Normal" loaded weight.)
For this information, I spent an awful lot of time trying to find the ammunition weight for various WW2 Aircraft Guns.

Hope this helps.
- Ivan.

P.S. Even with all the numbers, there is generally a lot more tweaking to get your performance and handling right and I am pretty sure there are people here who know how to do it better than I do.

[Milton Shupe]

For your single engine aircraft, I recommend downloading and trying Jerry Beckwith's AirWrench.

It is usable for a single engine craft like yours and will show you what you need to generate a proper FM.

It is quite inexpensive and for a novice FModeler getting started, it's the "Cat's Meow".

http://www.mudpond.org

[Sundog]

For your single engine aircraft, I recommend downloading and trying Jerry Beckwith's AirWrench.

It is usable for a single engine craft like yours and will show you what you need to generate a proper FM.

It is quite inexpensive and for a novice FModeler getting started, it's the "Cat's Meow".

http://www.mudpond.org

Seriously, this.

I think the only restriction is you are supposed to only use this for freeware.

[fliger747]

Yes Airwrench is a good utility and I don't think Jerry is getting into the 1% stratosphere with it. The only real beef with it is it recalculates everything each run, so any off field changes will be erased. Additionally there are some great tables it generates.

T

[Lionheart]

FDE's are incredibly difficult at times. Its like a stack of sticks, the game of 'pick up sticks'. You need to move one and you end up moving 3 or 4 more.

If you change the HP rating, it does nothing except change peak sound. Figure that one out....?! To increase HP in FS FDE's, increase/decrease the cubic inches.

You want to match as many of the dimensions as possible in the 'airplane_geometry' section. Fuel and weights balance is also key.


Getting your landing gear to behave properly is also VERY DIFFICULT! Have fun. FDE's give me nightmares. and gray eyebrows... grim!

[bstolle]

1.If you change the HP rating, it does nothing except change peak sound. Figure that one out....?! To increase HP in FS FDE's, increase/decrease the cubic inches.
2. You want to match as many of the dimensions as possible in the 'airplane_geometry' section
1. Definitely can't confirm that (such problems only occur if the air file doesn't match the cfg file and vice versa)
2. These data are mostly eye candy. Lift, flight control effectiveness etc. should be defined within the air file.

[OleBoy]

Thank you everyone, for your response and direction.

I've read bits and pieces through out the forums over time, about different situations and areas pertaining to flight dynamics. Generally, I just moved on. I was over whelmed with it. As for the most part right now. For the most part, I've come know by name those members who are knowledgeable in the areas of working with flight dynamics that come to SOH. And again (for the most part. LOL!), those members have responded. Very in depth!

At this point based on all that needs to be learned, added to, and on top of creating a model, it's more than I honestly want to have to deal with. Or want to deal with really.

[Ivan]

FDE's are incredibly difficult at times. Its like a stack of sticks, the game of 'pick up sticks'. You need to move one and you end up moving 3 or 4 more.

If you change the HP rating, it does nothing except change peak sound. Figure that one out....?! To increase HP in FS FDE's, increase/decrease the cubic inches.

You want to match as many of the dimensions as possible in the 'airplane_geometry' section. Fuel and weights balance is also key.

Getting your landing gear to behave properly is also VERY DIFFICULT! Have fun. FDE's give me nightmares. and gray eyebrows... grim!

I don't believe that changing cubic inches is the correct way to go. CID should match the actual aircraft's engine as should Manifold pressure, RPM, gearing, prop specs and all the other stuff I mentioned in my earlier post. To tune, you should be changing Torque and Friction loss modifiers which also let you tune the power curve of the engine to some extent. Once you get the Sea Level setting, you should tune the Supercharger boost to get your correct power at altitude. It isn't really difficult, just time consuming to test.

Landing gear tuning also isn't all that difficult if you want the critter to behave well. I don't have my notes in front of me so I can't quote the multiplier I use, but the basic idea is to tune the spring rates to be related to the weight of the aircraft. The damping is about (from memory) 1/3 the value of the spring rate. The contact points should be around 4-6 inches below the actual wheel's location (at least for CFS1). Don't worry about exact matches at this point. Close is good enough for a first pass.
I wrote up a spreadsheet to calculate the load on each wheel based on CoG, Ground Angle, and Contact Points location. The idea is to figure out the weights on each wheel and to adjust the spring rate to match. Once you have set the final spring rates you will be using, go back and adjust the contact points, fuselage angle, and release height in record 300. Two passes is generally all it takes to get decent behaviour.

Having spring rates very mismatched to the loads is bad because sometimes you will get a shake, wobble or strange bounce on only the mains or tail gear. To test, I slew the aircraft up about 3 feet and drop it. If it is well behaved, it will bounce once maybe. Once you have the "well behaved" numbers, you can mess a bit with the damping and perhaps even the spring rates to cause bad hehaviour that the original aircraft (Example: Corsair bounce on landing) may have had.

Hope this works for you also.
- Ivan.

[fliger747]

Engine HP:

Use the correct displacement and shaft RPM. Try to use the correct compression ratio and prop gearing ratio if it has reduction. For a constant speed prop unit set the max available manifold pressure and critical altitude if it is turbo or supercharged. Then test for proper HP. Delivered HP can be tweaked by adjusting the compression ratio somewhat or better the internal friction, which I think is the 509 table in the .Air file. More high RPM friction means less net HP etc. If you have good data of various RPM/MP vrs HP you can go a bit further by tweaking both ends of the friction line.

Good Luck! T

[Ivan]

Hello Fliger747,

Perhaps the differences we are encountering are because of different versions of simulators.
My experience has been that the Critical Altitude number has no observable effect on the engine output.
The Supercharger Boost Gain seems to have a large effect with very small changes in the number.

I believe that the following can be set according to real life specifications:
Displacement per cylinder
Cylinder Count
Maximum RPM
Compression Ratio
Maximum Manifold Pressure
Reduction Gear Ratio
Propeller Diameter and Pitch Range

The process I use is the following and it seems to work pretty reliably at least for CFS1:
First, find a stock aircraft that has a similar Propeller Power Coefficient.
Copy that record to replace your record 512. Replace Record 511 if you wish.
Beware, the Stock P-51D has some strange curves.
The Power Coefficient is basically a measure of how hard your prop is to spin.

Next, set all the parameters listed above. Don't worry about the supercharger yet.
Take it for a test flight using Mr. Beckwith's very excellent set of test gauges.
See what the power output is at sea level. I test at 500 feet. The correction to sea level is negligible.
If the power is too low, adjust the Torque value higher or Friction Loss lower.
If the power is too high, adjust the Torque value lower or Friction Loss higher.
Within 5 HP is good enough at this point. Don't worry, you'll be back here later.

Adjust the Supercharger factor to get the correct power at critical altitude:
Adjust the Critical Altitude number if you want. I haven't seen that it makes a difference in actual performance.
Run a power test at each altitude. I test at 2500 foot intervals up to about 40,000 feet for WW2 Fighters.
You may quit a lot earlier if your aircraft has a much lower ceiling.
This is quick because you only need to slew the aircraft to each level and let it stabilise for about 5-10 seconds.
Record Manifold Pressure and Horsepower at each altitude.
Even partial tests are good because you will be changing the Supercharger factor and testing again until your curve looks right.

When you are satisfied here, Fine tune the Coefficient of Drag to get reasonable max speeds at SL and Critical Altitude.
I usually do this with autopilot and record the number when the mph reading doesn't change for 15 seconds.
You might be asking: What does drag have to do with engine performance?
It probably doesn't relate much, but is necessary for the next step: A Service Ceiling test.
Using Mr. Beckwith's excellent test panel, set it to climb and slew to a few hundred feet under your expected ceiling.
See where the aeroplane's climb rate falls below 100 feet per minute. You DID set the proper load condition, right? Hope so.
Sometimes the service ceiling spec calls for 500 feet per minute. Check your references.

Here is where life gets interesting.
If the ceiling is too low, adjust your Torque LOWER and Friction LOWER to get the same Sea Level Output.
You will find it difficult to match the exact output which is why you were only trying for within about 5 HP earlier.
If the ceiling is too high, adjust Torque Higher and Friction Higher as well.

The net result is that with this method, you can adjust extreme altitude power output without affecting Sea Level or Critical Altitude much if at all.

By the way, how did this thread get moved to the FSX area? I know it started in Flight Dynamics.
- Ivan.

[fliger747]

The critical altitude in FS9 and FSX is quite important, as above that the engine MP will fall off with the pressure lapse rate. Example, a R2800 with a critical altitude of say 21000 ft in high blower will maintain the rated MP all the way to that altitude, as well as rated HP. This is not quite correct as high blower at critical alt in the real plane will use about 400HP, so the engine actually delivers only 1600 to the prop instead of 2000. However thrust gained from the stacks compensates a bit for this. At any rate things tend to work out about right.

use the AFSD utility and watch the MP decline with every 100 ft climb above critical altitude, and watch the HP decline!

Cheers: T

[Ivan]

Its interesting that MS is doing things differently now. I think I was given a copy as a present but just never installed it.
Perhaps I should look at how things work with FSX.
- Ivan.

[warchild]

There was a time when microsoft was on the right track and moving towards making a solid simulation foundation. Thats why Lockheed became so interested in MS's ESP platform, but But thats all gone now. Microsoft is not interested in making a simulator. it's barely interested in making toys and games and really prefers someone else make those too while they try and take some of the credit.. FSX and P3D give us a solid simulation foundation, and theres plenty of the usual undocumented features to pull from to rival a semi mid level simulator ( if you can afford the hardware )..

[OleBoy]

Thanks all for the input. I just bought AirWrench:salute:

[warchild]

Just remember with AirWrench that it will only output correct figures, if you input correct figures.. Sparks was real good about stating that point. but its a great lil program to learn from..

[Ivan]

I believe the Air Wrench manual is a very good read.

My goal isn't the AIR file that is the end result of using Air Wrench.
I am much more interested in learning how things work in the AIR file.
That is why the old discussions at the 714th were so interesting to me.

- Ivan.

[warchild]

Welll, theres the way things work in the airfile, and then theres the way things work in the airfile.. I know that sounds repiticious, but each entry has what it does, and then, each entry has a relationship with other parts of the model ( both files ) that causes it to have an expanding effect. Its the relationships and understanding how they function that really define how a flight model works, and to be honest, theres no book that could possibly define all of those without driving the author stark raving mad. Frankly, i'd be more inclined to recommend taking a course in fluidics. Same thing as a flight model minus the engines and other variables.