Advanced Flight Model for other aircrafts!
-
Some inflections, at least for the CL values near critical alpha values, are due to the spline function that i’ve applied to them in Excel, just to not have those sharp looking edges…, although the sim itself might also (i don’t know) interpolate values around those abrupt inflections within the data tables when playing!
Please correct me where you might think there is a problem and if you may…, tell me how should it be, otherwise i might actually make mistakes here or there without having a clear view of what could be so wrong. In terms of CD2AoA, i’ve reworked the data and made some significant reductions at lower AoA, and slightly increased them after the first (with wings swept forward at 16 deg.) and complete airflow separation (stall) that occurs around 11…13 deg AoA; after the first partial separation at around 10 deg AoA and second and full separation (stall) at around 22 deg AoA, when having the wings swept to 45, and finally and finally did the same for the wings swept at 72, with the fully developed stall at around 26…27 deg AoA. These critical angles of attack have been estimated according to the wing’s aspect ratio (which of course vary with the sweep angle). At 90 and -90 deg AoA though…, i’m pretty sure the CD values are very close to real now (depending on Mach number also), having any inaccuracy/error lowering as alpha travels from -45 to -90 and between +45 to +90.
The CL values that i’ve initially obtained were 2% to 8% higher than they should, so at least these weren’t too far from reality and now these are also corrected and much better.
I’d also like to share a very useful (more or less depending on everyone’s acknowledgement on aerodynamics alone) site which can give everyone some good clues about aerodynamics and more:
http://www.aerospaceweb.org/question/aerodynamics/q0194.shtml
Have a good day!
-
Please correct me where you might think there is a problem and if you may…, tell me how should it be
Cd at Null Lift to low.
This is a major error, the rest of the sequence.Check again the lower chart 23_Cl_to_Cd.
Look for Cd at zero Cl. Cd with zero lift is always greater than 0.02, for any Match number and wing sweep.
can build a chart Cd with zero lift point by point directly from the lower chart on 23_Cl_to_Cd (31 page of aerodynamic MiG-23).
or can be made easier.
just take data from the chart 23_Cdo (page 25). Cd0 is the drag coefficient at zero lift force, but in the case of the MiG-23. Cd0 obtained when the angle of attack is less than 0 AOA (asymmetric profile). -
@A.S:
I can provide the original russian EM charts for Su-27, Mig-23ML, Mig-25RB and Mig-29 if that helps.
Text is russian, but the graphs are understandable.I’d be interested in the Su- charts
Thank you.
-
Hello again,
I don’t even know how to start this as I haven’t shared any info in about years because I actually didn’t get the proper time needed anymore, but hopefully I’ve come up with some better aero and thrust data, not just for the MIG-23 ML, but also for 3 other fighters, trusting that these results would be a lot better! ntinue the purpose of gathering/obtaining the aero (also the thrust data, but I’m primarily capable for aero) tables needed to have these birds fly as authentic as they’ve ever been in our simulator.
I want to apologize to OSD and others for not paying too much attention on the lower AoA CD’s which indeed are erratically low…, I’ve done this mistake by confusing those those Cx charts that were written in russian (which I lack) and I mislead myself by interpreting them wrong. I misunderstood those charts, which were probably regarding the CD (Cx) increase due to some added stores or whatever and I can’t believe even now that I’ve let such low CD0 values like 0.002! Not even the aerodynamically shaped water drop can have such a low drag coef. These were the Cx(CD) diagrams/charts that I’ve confused I’m sorry because:
I already felt this was wrong and yet I just manually forced a function between zero lift alpha (AoA) and 15-20 alpha just to make this possible, but I was foul enough not to trust what I already knew:(, also waiting so much until I had this corrected!
The CL is also a bit refined and corrected for subsonic regimes where the function with Mach was inverse and is now corrected. I’ve uploaded 2 versions of the tables, one with CL drop beyond stall AoA (suitable for the AFM) and another version which works for the OFM (the OFM seems to not like CL reduction as the AoA increases) which has an almost constant CL gradient up to about 40…45deg. AOA from where any CL would start to drop. So it won’t simulate a stall for the OFM, but the critical AoA must be somehow limited on the OFM depending on wing sweep (as it’s the case with the variable geometry wings) to the known or calculated critical AoA for each wing position, otherwise we’d witness these aircraft manoeuvre much better than normal which is not the point for a good simulation, that’s why it would be more appropriate to use the correct aero tables on the AFM which accepts CL drops as AoA increases, although for the AFM I don’t know how to make it simulate a wing stall and/or wing rocking at a given AoA.
Now regarding the thrust tables, I have to admit that I’ve used some more empirical functions (that keep increasing thrust as the dynamic pressure and IAS increases) for altitudes >10000 feet which neglect the possible supersonic airflow (mach/shock waves ingestion) that may occur within the air intake of the engine, which in reality would lead to hazardous vibrations which could damage the engine, yet by thinking that the design of these intakes with the dimensions and positions of the splitter plate (boundary layer separator) should not allow such problems occur up to about 2.1 Mach (at least for the F-16), I would neglect the appearance of such situations even for an addition of further 0.4 Mach, thus continuing with the same functions to Mach 2.5 where our tables stop, presuming that between 2.1 and 2.5 there won’t be great changes in the airflow at the engine’s intake! Below 10000 I it would’ve been exaggerated to let the thrust continue to rise with Mach, so I manually tweaked it there (making the thrust slope become negative) to have it more appropriate.
Everywhere I’ve modified some default values/data, I have written “#MODIFIED” in order to tell what has been touched!
With the giving CL’s and critical AoAs (I’ve determined the stall/critical AoAs with some empirical equations with acceptable error) of the MIG-23 ML to say the least, the turn rates achieved are in a nearly perfect match with the real life ML’s manual, so at the moment there isn’t much that there can be done to increase the accuracy of the CL tables for the aircraft that I’ve worked on so far! At least these are the most correct values that I could obtain for the CL in particular! The CD tables might be reworked as needed, as well as the thrust tables, cause these give me a harder time in order to get them better and I can’t be sure of how accurate they are, but even so, from my opinion they should already provide more than 85-95% accuracy, which should be better than the default/actual tables.
You will notice that the CD for AoA +/-90 is 1.15 as arbitrary for Mach=0 and all aircraft start from this value as a general CD for those angles and from those angles of attack, the CD develops accordingly throughout the rest of the AoA range as they are calculated individually for each aircraft. We should know that a flat plate would produce a CD of about 2 at 90 degrees AoA having an infinite wingspan (so that’s only the case of a 2D airflow), yet for a real 3D 1 meter square flat plate (is a rugged example but not too wrong compared to any aircraft in such a situation) the CD would not get any higher than 1.2 at 90 alpha, while the best parachutes only manage to reach 1.55 or so…! According to size and wake flow generated behind the aircraft that would either fall vertically or reach a +/-90 deg. AoA, the most appropriate value would be around 1.15…1.2, which I used. The CD is expected to vary according to AoA/CL and Mach number. The hardest thing is finding the correct slopes/derivatives for each case, which is also a hard time for any CFD anyway, but hopefully I’ll be able to correct them with small but fair steps when I get the time.
I’ve attached the files for all 4 aircraft (F-4E, F-14’s, MIG-23ML and MIG-21) with their respective OFM/AFM model and the aero/thrust charts for MIG-23ML and F-4E which I managed to work on so far:
Here are the tables:
Here are the charts:
OSD, when you have the time, please check these values again! Thank you!
A great day to all the community!:)
-
Can you edit for a little the long post? I recommend to put ech link to a new line for better reading. I have data about MiG-23’s engine and also thrust data about engine of MiG-21bis.
-
Now regarding the thrust tables, I have to admit that I’ve used some more empirical functions (that keep increasing thrust as the dynamic pressure and IAS increases) for altitudes >10000 feet which neglect the possible supersonic airflow (mach/shock waves ingestion) that may occur within the air intake of the engine, which in reality would lead to hazardous vibrations which could damage the engine, yet by thinking that the design of these intakes with the dimensions and positions of the splitter plate (boundary layer separator) should not allow such problems occur up to about 2.1 Mach (at least for the F-16), I would neglect the appearance of such situations even for an addition of further 0.4 Mach, thus continuing with the same functions to Mach 2.5 where our tables stop, presuming that between 2.1 and 2.5 there won’t be great changes in the airflow at the engine’s intake! Below 10000 I it would’ve been exaggerated to let the thrust continue to rise with Mach, so I manually tweaked it there (making the thrust slope become negative) to have it more appropriate.
If you want to have a good feel of what is the relation between thrust and Mach/Altitude, there is a summary of thrust curves for various engines :
http://onlinelibrary.wiley.com/doi/10.1002/9780470117859.app4/pdfThe F404 is quite exhaustive in there. As you can see, depending on the altitude, the thrust decrease appears at different speeds. I believe that with fixed inlets , you basically “overfeed” the engine with air when density * speed² is above a certain threshold. So your assumptions are not too far off.
-
Can you edit for a little the long post? I recommend to put ech link to a new line for better reading. I have data about MiG-23’s engine and also thrust data about engine of MiG-21bis.
Thx “molnibalage” and sorry to everyone about the links! Yes, I could’ve put each one on a different line for easier picking!
-
If you want to have a good feel of what is the relation between thrust and Mach/Altitude, there is a summary of thrust curves for various engines :
http://onlinelibrary.wiley.com/doi/10.1002/9780470117859.app4/pdfThe F404 is quite exhaustive in there. As you can see, depending on the altitude, the thrust decrease appears at different speeds. I believe that with fixed inlets , you basically “overfeed” the engine with air when density * speed² is above a certain threshold. So your assumptions are not too far off.
Hi l3crusader,
Thank you very much for that link! Although I had some general idea of how the output thrust/HP would evolve with Mach and altitude, this one’s a good example which also uses simplified laws for determining fuel flow/consumption and thrust depending on Mach and altitude (although these laws assume that you already have some experimental data to start with and know the “sigma” coef., which might certainly be a variable) for 4 types of aviation engines (props, turboprops, turbofans and turbojets).
From what I see, the F404 fits almost closely in terms of thrust slopes/curves vs height and Mach to those that are being used for the F-16 and other fighter aircraft in BMS, and those are about the same curve shapes I’ve obtained to the 4 aircraft that I’ve uploaded, each of them having it’s unique values with the curve forms varying moderately from one to another. My belief is that the reason why the F404’s thrust curve slopes show an increase with Mach even from sea level in contrast to a turbofan like the TF-30 and other turbofan engines for which the thrust values degrease with Mach number (at low altitudes and at SL especially) up to about 20000’+ of altitude from where they also start to show positive slopes as well, is that the F404’s diagram is based on full afterburner measurements while the rest of turbofan engines show only the 100% MIL thrust.
I’m mostly specialized in aerodynamics and flight mechanics/dynamics and only know about jet engines up to a certain point;), so I must admit that I do not know why a jet engine’s thrust would decrease it’s output thrust if the airspeed or Mach number increases, because logically, as the IAS (indicated airspeed, which depends both on Mach and altitude) increases, so should the 0.5rhoV^2 (dynamic pressure, where “rho” stands for local air density) increase and the higher the pressure felt by the engine’s inlet, then the less need or less work has to be done by the compressor stages to compress air until the combustion chamber achieves a required pressure ratio, so theoretically, the jet (turbofan, turbojet, etc.) engine’s thrust should always increase as airspeed/Mach increases.
I don’t know exactly why does the thrust increase with Mach number (in the case of F404) also has a peak from where the slope becomes negative, but I think this has to do with engine surges that might occur when the shock or mach waves generated by the inlet splitter or anything else might actually get into the engine and reflect throughout the channel and meet the compressor stages which leads to a sudden loss of pressure throughout the intake and first stage, otherwise I don’t know why does a thrust peak occur for jet engines.
I’m looking forward to see how do the modified aero and thrust tables compare now!
Thanks, cheers!;)
-
First there can be only one AFM and that will remain the f16.
You can use the new FM code to be used by other aircraft (like i did for the mirage 2000) but that will never be AFM. That is for semantics.
Secondly, you will never be able to nullify the FLCS from the data files as many many gains are hardcoded and you have a very limited number of parameters extenalized in the data files. As far as FLCS is concerned the NASA model originally built for AFM will certainly be more flexible for you.
Anyway before talking about flcs, we should talk about aeromodeling first.
Before starting to think about using the new FM code and hid Aero moduke it is abolutly mandatory that you work first on CL,CD and thrust tables and ensure that they are producing accurate EM performances.
When you reach that point you cans start thinking to the next step i.e. all other AFM Aero coefficients. I am a little bit surprised to read you take the Cm of the main wing only because the Aero model included in BMS does not fit with that at all, Cm depends on deltah and then therefore depends on elevator action.
Have you read the TP1538?
_–----------
Unortunatly i won’t have time to help you really as i don’t have time anymore for this
Anyway i strongly think you should follow those 3 steps:
- create hffm for your AC (CD, CL, Thrust) no need AFM files here, use OFM
- create Aero AFM data based on the NASA TP1538 modeling ( i think the best is to start from F16 data and tweak them as i did for m2k as it is impossible to calculate them from scratch or find them in the same modeling)
- tweak NASA flcs to unleash you model ( i will give you more details when you will be at that step)
+1
-
Hi A.S,
From what I’m concerned, I believe to have accomplished the hffm (high fidelity flight model) part for OFM giving the CL and CD tables that I could obtain, as well as reading the whole TP1538 which I did 3 years ago and understood that it’s a basis for the F-16’s flight dynamics modelling in BMS, including the aero data which was probably taken from a wind tunnel test. I forgot to talk about the Cm immediately after it was put into question, and yes, in reality the pitching moment coef. can be obtained relatively to wing only, fuselage only, control surfaces and/or empennages only or for the whole aircraft, as a dependency of Cm vs AoA and so I didn’t understand until I was told that in the AFM’s model, the “START OF DELTA CM TABLE” is for Cm vs DeltaH(elevator), but it seems it also takes into account for AoA which is much better. The CM part for aero table is not my concern anymore, but if I were to adjust it I would do it by trial until I’m satisfied with the pitching moment response from the aircraft by knowing it’s overall stability characteristics.
I understand that there are very few around who actually still have time for making BMS’s flight models more realistic, yet as far as I can help (within given limits of accuracy), would be to provide relatively accurate data (between 2…10% error) for CL and CD tables, while the thrust data would rather be obtained from tweak and trial using known data on each engine type, so the accuracy might degrade for even more than 10% for the thrust tables, but the higher the error, the fewer the places where it would occur anyway.
So the first 2 steps I believe to have done quite right according to what I’ve said, while the 3rd step seems a bit too much for me atm as I can’t find the grip on working with the AFM’s stability and input data, although it looks easy.
I’d like to ask one question though: If neighter these final aero and thrust tables (especially aero) aren’t satisfactory, is it still worth it to continue obtaining them for further aircraft in BMS or should I stop here?
Thank you and let’s hope for better!
-
Hi MaverickF22
I’d like to ask one question though: If neighter these final aero and thrust tables (especially aero) aren’t satisfactory, is it still worth it to continue obtaining them for further aircraft in BMS or should I stop here?
This is a question you should ask Mav-JP as he is the creator of the code-environment and hence has more insight regards. Maybe he has a good idea, how to continue from that point.
I myself do understand “flight” as such quite well, but the simulation!-physics (and partially its mathematical language of it) only on an amateur-level, meaning i´m not an aeronautical engineer - Mav-JP is. Nevertheless it´s still good enough to understand or detect flaws and errors.On the other hand, if you need someone, who is still expirienced enough from a piloting point of view and also able to test-fly your work with the ability to communicate back to you in a professional usefull language - so you can actually fine-tune things - then let me know. In case you are aiming for a Mig-29 modelling (provided data is available), then i would be really interested in helping out, because in my own view… it is THE “missing link” to a F-16 Simulation. The Mig-29 was at the time THE! historical counterpart for the F-16 and it would be friggn amazing to have it finally as a “HIFI” (in quotes) fly-able and thereby to “complete” the sim.
S!
Best of luck.
-
Hi A.S and thanks for the reply!
It would be an honor for me to have an actual pilot testing how the aircraft performs (turn rates, forward accelerations, etc.) according to the CL and CD tables that I can provide (optionally thrust tables). I am a glider pilot myself and I understand quite well too how a fighter aircraft should behave in certain conditions, so if there is someone who might have RL experience flying a fighter plane up to it’s limits (yet not exactly necessarily), he would be a great help for our simulator and I’m willing to share thoughts and help making BMS get better aero tables for all flyable aircraft, at least.
Thanks for giving me a hint of what aircraft might immediately need better aero tables, and indeed the MIG-29A/S, which I’ve been thinking about for quite a while, would need to get more realistic lift and drag tables, so when a Falcon driver would meet an AI MIG-29A/S or an opponent player MIG-29, he or they should get a more realistic experience.
I’ll start working on both the MIG-29A and S versions asap and come back with the final charts (as I did with the F-4 and MIG-23) and their specific aero and thrust tables.
Best wishes!
P.S.:About the graphic charts and why do the lift and drag coefs vary the way they do (if some might want to ask): As the aerodynamicists know, the CL does have a law of variation not only with AoA, but also with Mach, and so does the CD, Cm and the other 3 left coefs., and these laws depend on both airfoil section shape as well as the 3D wing’s shape. Now, although these laws/functions of variations might look very similar to one another from different aircraft, their values differ widely as it is clearly noticeable. With these CL and CD variations, the aircraft should be able to experience a more realistic CL increase with Mach (at constant AoA) until a “shock stall” occurs (from where it drops somewhere below the value at Mach 0) and drops before rising back to an X value where the airflow is almost completely supersonic over the wing and also about where the CD might have a peak.
-
Been a while since I posted last good to see you all again
MaverickF22 have you tested for departure condition’s? or have any documentation for such planes that you are working on?
Cheers
Steve -
Hi A.S and thanks for the reply!
It would be an honor for me to have an actual pilot testing how the aircraft performs…
I am not MaverickF22. Just 13 years of virtual flying and alot reading and sometimes too much dedication to this hobby ….nothing more.
-
@A.S:
I am not MaverickF22. Just 13 years of virtual flying and alot reading and sometimes too much dedication to this hobby ….nothing more.
Hello again A.S,
I can tell about the same of myself too, including that I am an aviation engineer (aerodynamicist) and active glider pilot since 2007 and I’ve started playing Open Falcon (this is what I started with from all other Falcon 4.0 series) since 2006 until today. Very lately though I’ve found the data files and after seeing the aero tables I understood that the game actually listens to that data in order to have the lift, drag, sideforce, etc. simulated on the aircraft and I thought that I could try to give those values a more appropriate/realistic measure and thus contribute to get the Falcon (now BMS) a greater experience from either a real fighter pilot who had flown these planes or to those who are curious how they perform when competing against each other. That’s where the aero tables “win their money” to say it that way and I’m pretty obsessed to have them as accurate as possible and thus be able to smile every time I dogfight against another aircraft and know that it performs just like in reality, or at least very close to it.
Now I’ll keep working on the 29 (mostly when I have the time for it) and can’t wait to bring in it’s updated CL and CD data.
-
Hello again A.S,
Now I’ll keep working on the 29 (mostly when I have the time for it) and can’t wait to bring in it’s updated CL and CD data.
nice
-
Hello again,
I’ve uploaded the MIG-29A and S data files! The only difference between the 2 is the AoA limiter, otherwise they share the same aero data and engines (I couldn’t find differences on the engine models used between A and S, so it’s possibly the same engine).
As comparison between the initial (default) aero tables and thrust data and the ones calculated, now the MIG-29 has slightly more acceleration, higher drag coef. at higher AoA and a slightly higher lift slope (hence the maximum CL). So, now it’s a bit easier to win a guns only 1vs1 (while flying the F-16C) against a MIG-29 when it turns (due to MIG’s higher speed bleed), but on the vertical fight it’s slightly above the F-16 and can beat it easier than before.
Here are the data files:
Here are the obtained charts:
Take a time to verify how it performs.
Cheers and good day!
-
Nice one Maverick.
Having a passion for the SU-27/30 and working long to improve its FM/AFM I look forward to results of your work.
Much Thanks……Shadow
-
Hi Shadow,
I’ll find some time to work on the 27 as well. I also love this beautiful fighter, and in fact I love them all, but the Flanker is special. I was already having in plan to put the CFD and equations at work for each flyable plane in BMS, one at a time and now we have a grood priority ahead of all!
I must remind that the general form/shape of the functions (both aero and thrust) like CL vs AoA, CD vs AoA, CL vs CD, CL&CD vs Mach may not provide a mean accuracy with less than 5% error in some cases, yet overall I believe that the aero tables that I can work on will make these birds fly much more like the real ones, at least this is my goal! When I’ll have the computer capability and time (for using powerful CFD software that can “do it all” and depend less on theoretical equations that complete the CFD results in specific areas), I’ll enhance the aero tables from top to bottom again for each aircraft, just to make sure they’re as realistic as possible, but by then only time should tell!
All the best…!
-
I’m back with the Su-27 completed and with slightly modified MIG-29 (A&S) for the engine thrust tables!
Here are the data files:
Here are the tables for Su-27:
I only hope I could’ve brought them better (more realistic) than the default values!
Cheers, good day!