New flight dynamics for 5 WW1 biplanes
Results 1 to 14 of 14

Thread: New flight dynamics for 5 WW1 biplanes

  1. #1

    New flight dynamics for 5 WW1 biplanes

    Hello Folks,

    With help from Ravenna, Kdriver and Martin Klein, I have just uploaded some new flight dynamics for 5 WW1 biplanes:

    - Curtiss_Jenny
    - Hanriot HD-2 Floatplane
    - Airco DH-9
    - Airco DH-4
    - Sopwith Triplane

    They will be available in the library as soon as they are approved by the librarian:

    http://www.sim-outhouse.com/sohforums/local_links.php?action=jump&catid=124&id=26887


    The objective is to correct some inherent in-flight instabilities on the flyable model, but most importantly, the suicidal tendencies of their AI versions during missions.

    Kindly make the necessary changes to file names and the Flightsim Number sections in the Aircraft.cfg for the specific model chosen.

    Changes involve mostly corrections to the engine performance graphs and propeller graph tables, as well as the Side Force due to Rudder entry in the Primary Aerodynamics section, and sometimes a correction in the Aircraft.cfg CoG entry.

    Hopefully they will be of some help.

    Cheers,
    Aleatorylamp
    "Why make it simple if you can also make it complicated?"

  2. #2
    SOH Staff Devildog73's Avatar
    Join Date
    Sep 2005
    Location
    The Melbourne in FL, USA
    Age
    72
    Posts
    2,618
    aleatorylamp,

    Thank you SO much!

    I have been avoiding WW-I aircraft and missions because I just could not master the flight of the biplanes.
    I continually crashed, where in more modern WW-II and newer aircraft I had no problems....

    DDog
    Devildog73

    Semper Fi
    "Earned RESPECT seldom needs to demand respect"

  3. #3
    Hello Devildog,
    Iīm glad indeed that you have found this useful!

    I know very little about missions and the way to set them up, but when Ravenna pointed out the flaws in the behaviour of some AIīs, i.e. suicidal tendencies, I decided to inspect their FD. I found the flyable model had some serious stability problems.

    A comparison of their FD with those of several other similar aircraft, seemed to show some issues that were a bit out of line with what seems to be more conventional, so I made some changes. These made the flyable model easier to handle, with a more realistic performance.

    Then, Ravenna tested the revised FD in the corresponding AI aircraft in missions, and it seems to have fixed the problems.

    Letīs see how it goes!

    Cheers,
    Aleatorylamp
    "Why make it simple if you can also make it complicated?"

  4. #4

    New CFS2 Flight Dynamics for 5 WW1 aircraft

    A new entry has been added to Add-Ons Library, category CFS 2 Aircraft - World War I

    Description: New .air and Aircraft.cfg files for:


    - Curtiss_Jenny
    - Hanriot HD-2 Floatplane
    - Airco DH-9
    - Airco DH-4
    - Sopwith Triplane


    These correct some inherent in-flight instabilities on the flyable model, but most importantly, the suicidal tendencies of their AI versions during missions.


    Kindly make the necessary changes to file names and the Flightsim Number sections in the Aircraft.cfg for the specific model chosen.


    Changes involve mostly corrections to the engine performance graphs and propeller graph tables, as well as the Side Force due to Rudder entry in the Primary Aerodynamics section, and sometimes a correction in the Aircraft.cfg CoG correction.

    These new flight dynamics are supplied with help from Ravenna, Kdriver and Martin Klein.
    Enjoy!
    By Stephan Scholz, July 2020

    To check it out, rate it or add comments, visit New CFS2 Flight Dynamics for 5 WW1 aircraft
    The comments you make there will appear in the posts below.

  5. #5
    SOH-CM-2023 Ravenna's Avatar
    Join Date
    Jun 2005
    Location
    Newcastle, Australia
    Age
    71
    Posts
    2,355
    Err, still testing today, but at least I can say that the Sopwith Tripe takes off without making holes in the airfield. As for the rest the player aircraft are nicely done. Thanks Stephan!

    File the Jenny FD somewhere safe for later.

  6. #6
    Hi Ravenna,
    OK, no hurry. Letīs see how it goes with the other AIīs then.

    It seems the main stability issue causing the AI failures was the entry for Side Force - Side Slip Angle
    in Primary Aerodynamics (not Side Force due to Rudder - my mistake), that caused abrupt side-slipping and made the AIīs crash into the ground at low altitudes.

    This entry appears to have been a maximum value, the intention behind which I presume was to make AI evasive manouevers much quicker. Anyway, I put in more conventioinal values there.

    Hopefully the AIīs will be useful in missions now!
    Cheers,
    Aleatorylamp
    "Why make it simple if you can also make it complicated?"

  7. #7
    Gentlemen,

    Keep in mind that many aircraft from the Great War really were not well behaved and were quite unstable in certain conditions.
    I would expect a floatplane conversion to be one of those because of the increased "keel area" of the floats that often extend ahead of the Aircraft CoG. Aircraft design wasn't really a science back then and structures often deformed under load to create some strange aerodynamics as well. Another observation is that often the Rudders of the time were seriously over balanced and had no fixed fin area for stability. The same can often be said about the full moving Elevators.

    Aleatorylamp,
    Several years ago, I believe the Side Force - Side Slip Angle issue came up in discussion when I was trying to make my brand new P-40N able to fly a knife edge. Although modifying this field worked for this single purpose, it created some seriously strange and not very predictable handling problems and I ended up removing the edits because I could not find a combination that seemed to behave well without breaking something else.

    Also, if CFS2 AI is anything like that found in CFS1, there are certain records (Control Modulation) that make an aircraft much more "pointable" and controllable that are totally ignored by AI. Back around 2002 or 2003, I found this out when trying to tune my FW 190A-8 for the roll response noted in flight testing of the actual aircraft. I was able to do this, but needed to put the control factor for the Ailerons so high that AI simply could not handle it though it was no problem for a human pilot.

    - Ivan.

  8. #8
    Hello Ivan,
    Nice to hear from you. Hope youīre doing well, and thanks for your comments!

    The main reason for updating the .air files for the 5 WW1 biplanes in question, was the suicidal tendency of the AI versions of the aircraft that simply flew themselves into the ground at low altitude.

    Another factor was the difficulty to fly them properly, and I found several issues that I could correct to make the models more flyable. One of them even had a strong CoG offset in the Aircraft.cfg, which proved problematic because I discovered it very late. Very frustrating, as I was only looking into the .air file and was getting no results at all until Kdriver and Martin Klein pointed out that the Aircraft,cfg could overwrite things in the .air file. No wonder... :-(

    As I donīt know very much about programming AI aircraft, I limited my intervention to adjustments which improved the flyable aircraftsī handling, and substituting some incongruent looking entries for more conventional ones.

    Apart from the Side Force Sideslip Angle entry, I found propeller and engine graphs with strange shapes. Some of them simply flat, so I changed them to shapes that I felt were more familiar and more common to this type of aircraft.

    This also involved using my QBasic programs relative "J" Factors and Zero efficiency points, so, applying what I learnt from you, I could adjust performance to more realistic levels, and it appeared to be successful.

    The end result, for the moment at least, appears to have made the AI versions of the 5 aircraft useable in missions.

    Letīs see if there are any more reports to this respect.

    Anyway...
    Cheers,
    Aleatorylamp
    "Why make it simple if you can also make it complicated?"

  9. #9
    Hello Aleatorylamp,

    Thanks for the good wishes.

    As I mentioned before, AI doesn't really fly quite the same way. From what I can tell, controls are "binary".
    They are either full deflection or neutral. The AI also adjusts much faster and doesn't necessarily understand inertia and damping.
    I believe there are also a lot of other things that the AI doesn't really know how to handle that a human figures out by applying some amount of control input and sensing the response. That may be why there are certain flight models that AI can't deal with; they only have one set of responses.

    The CoG settings between different versions of simulator are VERY different. This is one of the reasons it took me so long to release my own versions of the Airacobra. I wanted to create an "Aft CoG" condition but found that the AIR file edits you and I were discussing way back actually had absolutely no effect in CFS (and probably not in CFS2 either) even though they worked in FS 98. The situation is actually handled by pitch moments so if you can directly alter them via the Aircraft.CFG which I cannot, then it is may be the only way to go.
    There was actually a lot of work that went into the flight model, much more detail work than the forum thread would imply.

    By the way, I actually got the Airacobra to "Tumble" at one point. It isn't easy to do though and it isn't asymmetrical in my flight model as it is described in flight reports so there are still some not so obvious discrepancies that I don't know enough to reproduce.

    Be somewhat careful about how you build new propellers. There are some characteristics in the P-39D-2 prop that are not so well behaved. I did not figure this out for a long time because I didn't do any checks of your release until well after I had finished and released my own version.
    Sometimes a nice looking set of curves doesn't actually work well because of interpolation issues. I just ran into this on a very small scale with the recent Me 109E-4 and expect to see the same on the other two flight models I am currently working on.

    Take Care and Apologies for hijacking your thread.

    - Ivan.

  10. #10
    Hello Ivan,

    I donīt know much about AIīs so what you say is interesting.

    With the revised FD upto now, it seems that one AI wingman now seems to behave well, but there still seem to be problems with multiple AIīs - perhaps it has to do with collision bubbles... I donīt know.

    Although what is in the Aircraft.cfg is still quite new to me, and I still donīt understand why it is necessary, but thatīs beside the point. Relative to your comments on the Pitch, Yaw or Roll Moments that can be adjusted in the .air file, it appears that their counterparts in the Aircraft.cfg are the aileron, rudder and elevator effectiveness lines.

    Anyway, we shall see...
    Cheers,
    Aleatorylamp
    "Why make it simple if you can also make it complicated?"

  11. #11
    SOH-CM-2023 Ravenna's Avatar
    Join Date
    Jun 2005
    Location
    Newcastle, Australia
    Age
    71
    Posts
    2,355
    Hi Folks,

    All this makes for interesting reading. I'm not sure I understand all of it! Just what is a Pitch Moment?
    I've spent all day puting the aircraft equipped with Stephan's new FD files through their paces. A few things stand out: Air starting the Sop Tripe, DH.4 and DH.9 present no problem. Aircraft stay in formation.
    The problems start with take offs. Player + one AI presents no great dramas. The occasional crash but generally take offs are incident free. With more than one AI, however my rear view is of multiple crashes- tail up and full forward roll! With 6 AI, when the smoke clears a bit, there is usually one AI that rises to take its place beside the player aircraft.

    Is it possible that the collision bubble is causing the closely grouped aircraft to crash when taking off? I haven't tried to hand edit the mis file to move the aircraft further apart...I just ran out of time today.

  12. #12
    Hello Aleatorylamp,

    You are probably thinking about the Moments of Inertia found in Record 1001. That is not what I am discussing.
    I was only discussing Pitch Moments which are found in Record 538 and Pitch Moment due to AoA Rate found in Record 1101.
    Personally, I don't know that I know how to use the field in Record 1101 though I do understand the concept.
    It is a matter of knowing what it is describing but not knowing what conditions would cause it to happen.

    As I understand it, Record 538 describes what the Aeroplane wants to do when just "existing" in an airstream. If you think about it, the specification of where the Center of Lift is in relationship to the Center of Gravity pretty much describes the forces which would result in the same thing but only on one (Longitudinal) axis.
    Imagine what happens on an aeroplane with a Parasol wing. The stability with the high wing is better because as the AoA increases, there is a greater tendency for the aeroplane to pitch down.

    Most Aeroplanes with a proper CoG loading will pitch down at the stall because the CoG is ahead of the Center of Lift. It makes for predictable handling. The Airacobra with most of its stores expended was still flyable but if stalled, it would not pitch down as quickly and regain flying speed. Things got very ugly if it was in an accelerated stall because sometimes the gyrations would not settle before the 'Cobra hit the ground. That is why there were warnings not to fly aerobatics with aft CoG.

    The Control Effectiveness is a different thing altogether.

    Hello Ravenna,

    I hope my explanation above also addresses what you were asking. I don't know that I am completely correct, but the behaviour of the flight models when I make edits seems to reflect what I just described.

    - Ivan.

  13. #13
    Hello Ivan,

    I didnīt mean Moments of Inertia, but was referring to the control surface moments in the Primary Aerodynamics section: Roll Moment due to Ailerons, Pitch moment due to elevator, and Yaw Moment due to Rudder, and then I mentioned their "effectiveness" counterparts in the Aircraft.cfg, in answer to your comment:

    "The situation is actually handled by pitch moments so if you can directly alter them via the Aircraft.CFG which I cannot, then it is may be the only way to go."

    However, it seems that it wasnīt what you meant. U
    nfortunately it is too complicated for me, and my insight into the matter is not enough, so I canīt offer any more.

    Cheers,
    Aleatorylamp
    "Why make it simple if you can also make it complicated?"

  14. #14

    Making AIīs useable for missions

    Hello Folks,
    More than anything, the objective of the revised FD for the 5 WW1 fighters mentioned on this thread, is to make their AI versions useable in missions.

    Re-working these .air files produced some improvement as regards stability in their flyable versions, but with apparently only limited improvement in the AI aircraft using these revised FD.

    As reported by Ravenna, upto now, only single AI aircraft seem to work on missions, but multiple AIīs seem to interfere with each others, causing crashes.

    Any ideas to remedy this situation will be welcome.
    Cheers,
    Aleatorylamp
    "Why make it simple if you can also make it complicated?"

Members who have read this thread: 0

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •