Page 1 of 1

What sort of data would you want to look at to make more realistic DAT Files?

Posted: Fri Nov 20, 2015 12:39 am
by decaff_42
Title pretty much says it all. If you could look at data from a flight, what kind of parameters would you like to see values for in order to evaluate performance?

I am working on a Flight Analyzer to extract such information, but I'd like to speed things up by organizing analyses into different modules where certain groups of variables would be recorded and/or calculated while others would be ignored. For example if you only were interested in cruise performance, how about only taking throttle, aircraft lift and drag, etc. and ignore roll-rates, g force, etc.

I'm looking for quantifiable parameters, i.e. things that can be calculated.

Re: What sort of data would you want to look at to make more realistic DAT Files?

Posted: Fri Nov 20, 2015 2:25 am
by Gunny
Some quantifiable relationship for handling aircraft for flight characteristics for individual aircraft or even a class structure.So that a 747 will not handle the same as a C-130.Something a builder could look at and find the handling characteristics of that aircraft or even just that class of aircraft realizing of course that there are so many variables that can effect the handling of an individual aircraft even in the same classification and even the same model.

CPITMANE 5.5 #PITCH MANEUVERBILITY CONSTANT
CPITSTAB 3.2 #PITCH STABILITY CONSTANT
CYAWMANE 3.6 #YAW MANEUVABILITY CONSTANT
CYAWSTAB 3.0 #YAW STABILITY CONSTANT
CROLLMAN 3.0 #ROLL MANEUVABILITY CONSTANT

Re: What sort of data would you want to look at to make more realistic DAT Files?

Posted: Fri Nov 20, 2015 2:51 am
by decaff_42
This is a great point. The problem will come with gather flight data with a lot of changes being made to variables. You will have to make a DAT change and then fly the same maneuvers again. This will take time.

YSFlight also has a in-flight alterable input sensitivity which might change things. I will have to look into that more and see if that changes the recorded control surface input value, or the assigned amplitude of the maximum input.

Re: What sort of data would you want to look at to make more realistic DAT Files?

Posted: Fri Nov 20, 2015 8:52 am
by NightRaven
Engine positions + thrust from each engines to allow maneuvers with engines alone.

Surface area + deflection of control surfaces to dictate flight performance rather than arbitrary constants.

Movable and rotatable hardpoints.

Customisable weapons.

Differing GUNPOWER per gun.

Pretty much all I can think for now.

Re: What sort of data would you want to look at to make more realistic DAT Files?

Posted: Fri Nov 20, 2015 11:27 am
by u2fly
NightRaven wrote: Movable and rotatable hardpoints.
Agree, also it must be moveable that will be very usefull for carrier decks and for wheels.

Re: What sort of data would you want to look at to make more realistic DAT Files?

Posted: Fri Nov 20, 2015 12:28 pm
by decaff_42
I think you guys misinterpreted my question. This is NOT a wishlist topic. I am looking to extract performance data from a .yfs file and then use that to see how the airplane behaved.

What aircraft performance parameters would you want to look at to improve DAT FILE numbers?

Re: What sort of data would you want to look at to make more realistic DAT Files?

Posted: Fri Nov 20, 2015 9:21 pm
by waspe414
REFVLAND 110kt #LANDING SPEED
REFAOALD 15deg #LANDING AOA
REFTHRLD 0.3 #LANDING THROTTLE SETTING

After maneuverability constants and thrust, those are the most significant variables for dogfighters, affecting how closely the velocity vector sticks to boresight, which affects turn rate and high-AOA drag.

I recall doing some analysis of those years ago, when I figured out what they did, but I doubt I have it anymore.

Re: What sort of data would you want to look at to make more realistic DAT Files?

Posted: Sat Nov 21, 2015 2:02 am
by Barr
This for developing a flight analyzer right?

This may or may not be related,either way feel free to delete.

Does ys even realistically calculate fuel consumption? For example a plane on a climb would be using more fuel even if the throttle were equal when cruising. If it does a total fuel used, and seperate that into fuel used during climb, cruise and decent.

Also I always wanted to know how smooth my landings were. Ys will let you land as long as your rate of descent is less than, I don't know but for an example, 2000 ft per second. So a calculation based on speed, rate of decent, time etc to figure out the amount of g force on touchdown. Maybe the same can be done with take off?

Again I dont know much.

Re: What sort of data would you want to look at to make more realistic DAT Files?

Posted: Sat Nov 21, 2015 2:53 am
by decaff_42
I spent some time working on this today so I think I can give you a better understanding of what this will do.

(1) Fly a flight with the aircraft you wish to test.
(2) Save the flight as a .yfs file
(3) Using the setup page, import the aircraft DAT file and the .yfs file generated from the flight.
(4) Select the time-frame you want to analyze
(5) Select what frame-rate you want the data. (long flights choose a lower frame rate)
(6) Select what type of parameters you want to calculate. (Manueverability for angles and rates, Cruise performance for thrust, engine, drag) I guess I could include a landing and take off portion.
(7) run the analyzer

I am still working on the analysis portion of the code. So far I have been able to impliment my AoA code which seems to work well. I will have to do some further checking of this to make sure it works properly. I have so many things that could be something a modder would want to look at. Do you want bank angle rate? Pitch vs elevator input, etc.

(8) Select parameters to plot
(9) generate plots

These last two, I'm not sure how it would work, but if not feasible else, you could make your own plots of the raw data and see what data you want.

Re: What sort of data would you want to look at to make more realistic DAT Files?

Posted: Sat Nov 21, 2015 4:29 am
by Flake
What say you do some numerical statistics on common CLA functions and derived variables like velocity and acceleration?

For example: Mean, Median, Mode, Standard Deviation/Std. Err/Variance, Mean Average Deviation/Sum Squares Error of Mean...

Applyable to things like throttle position, VGW position, angle of attack/slip, airspeed...

That's a lot of possible options but really you could apply it to so many dat variables - eg: testing the variance of airspeed while performing certain manuevers, comparing bleed of airspeed to loading of G, etc. - could be used to compare .dats and balance performance.

Re: What sort of data would you want to look at to make more realistic DAT Files?

Posted: Sat Nov 21, 2015 10:22 am
by u2fly
Here are some free excel sheets for design and analyze aircrafts, that maybe have some ideas for this.
1) http://www.aircraft-certification.de/in ... nload.html
2) http://www.aerodyndesign.com/fwd_tool/fwd_tool.htm