Following some discussion on Spitfire Stability I thought I'd try to measure the hands off stability. I can't be certain that it is correct but we may just have the unstable Spitfire along the lines of the A&AEE (and NACA) findings. I can't vouch that it is a true representation of the condition, I just tried to fly it and see what happened.
Attached is a zip file containing an excel spreadsheet (actually created in Open Office) which is the .csv file output from the test mission but saved in excel format so I could create the chart. A jpg of it is in the attached file and also embedded in this post for easy viewing. The mission file Stability.mis and its Stability.cs script file are also in the zip so if anyone wants to try it, drop those into your My Documents/1C Soft Club/il-2 sturmovik cliffs of dover/missions/ folder.
You will see from the 9000 records of raw data and the fact that I only charted records 4000 onwards that I had a job getting it to settle. I set record 4000 as time zero and charted against time in minutes. I followed the A&AEE method of pilot controlling until steady then releasing the controls. The only adjustment I made was very light and careful aileron adjiustment as it is impossible to fly it hands-off at the A&AEE test rpm of 2580 and full boost (not BCO). Under these conditions it gradually began to oscillate longitudinally.
I charted the RPM, Manifold Pressure (cockpit reading), Altitude and, for us perhaps the most important of all, the Pitch and Speed variations. You can see from about 4.8mins I finally had it settled down and the hands off flying began creating oscillations of pitch and associated speed changes: the descending speed increased until it brought the nose up or it decayed until it brought the nose down. The oscillations are particularly noticeable in the indicated airspeed and Pitch Angles (positive = nose down). The overall progress was a gradual climb but that may be due to my elevator trimming which is quite tricky to get stable with my rotary. Perhaps a minutely lesser trimming would have kept the average altitude generally the same.
Whether or not this truly represents what we are looking for I cannot say. It could just have been bad trimming or indeed instability but I made no elevator inputs at all. Perhaps some of you would like to give it a go.
The mission is included so you can try it yourself. Be sure to edit the file path for the output file in the file Stability.cs You will find it down around line 114, where it reads:
fstring = "E:/CoD_data/BlackBox_Data/" + strfile + "-" + cur_ac.ToString() + ".CSV";
You can see I am using my E; disc. You can edit the disc and path to suit yourself. Use Notepad, Notepad+ or Microsoft Visual C# Express. Dont edit the part from + strfile onwards.
By the way I have now included Elevator trim setting in the output file which was not in my own results. It may be of use, perhaps as a marker for when you finally got it settled. Select a series of records you believe represents the settled test period and chart them.
Edits/Additions:-
EDIT 1: The title of the chart reads 2850rpm, that should read 2580rpm
EDIT 2: Files replaced. Found reason for erratic Pitch curve - wrong data column plotted
One reason why the oscillations are not as frequent as the A&AEE/NACA tests may be that the control column was not truly 'free' as it is held by the joystick springs.
EDIT 3: zip file replaced. Control deflections now included in output file.
EDIT 4: Speed at Altitude chart added. Only a few checks were necessary to tell that it is well below par.