Most Realistic flight modeling

[warchild]

welll, i would humbly submit... http://www.iasa.com.au/folders/Safet...nacleCRJ2.html
The Faa and NTSB had a field day with that one.. you can be sure of it..

Why Not use the main physics engine for every single plane??? because although the engine is quite capable of making use in the differences between wingspan wing root fuselage length and various other physical attributes, it doesnt take into account any of the myriad of differences that exist between each plane. For example, it doesnt model the difference between a CFM-56 and a JT-3. it treats both of them the same, if you dont have anything but zero's in your air file.. Thats why i use air files. I want it as true as i can possibly make it..

PS.. The CFM-56 is used on a 737, the JT-3 powered the SR-71 at over mach 3. ( for those who didnt know )

PPS.. I didnt call airline pilots over all idiots. I call this group of idiots idiots..

ppps. this discussion goes OT, but covers the above mentioned's causes and effects.. it is done by real pilots.. http://forums.jetcareers.com/technic...ow-cruise.html

[Panther_99FS]

Which one is the most FUN :ques:

[warchild]

Falcon 4.0..

[bstolle]

That's an easy one. Flying in a war scenario isn't fun at all, x-plane is still very sterile and FSX has easily the highest immersion factor!

[Major_Spittle]

My question was more of a " which FS models flight the best " in terms of start out with an airplane at 20,000ft sitting still with no power and drop it and see what happens.

Put a tail wind of 80 knots at 20,000 ft and drop it, side wind....... I'm talking more about accuracy in modeling the physics of flight, not duplicating a particular airplanes specs, stall speeds and such. I could care less if the cessna 172 is stalling at a faster airspeed or takes more runway to take off because I will never fly a cessna 172, but is the simulator duplicating flight? when there is a tail wind gusting does the plane react to it correctly, side wind, prop wash, drag from lowering landing gear, having fuel emptied in one wing and full in the other, speed/lack of airflow over control surfaces....... Which FS do you have to "fly" the airplane in it the most in accordance with real world physics.

To me X-Plane and ROF are better at this than FSX. FSX is prolly better (depending on aircraft) of duplicating " flying by the numbers flight " of particular aircraft. Just don't stall that airplane inverted with a tail wind and expect it to react in accordance with real world physics or your inputs to react in accordance with physics under those circumstances. (with any FS, but ESPECIALLY FSX).

And I disagree that plane feel like they are flying on rails. They feel stable if flying underpower and a speed in which their aerodynamics stablize them, they should feel like on rails while crabbing, slipping, stalling, flaring....... I have rode with a crop duster, I remember nothing better than the feeling of the planes wings washing during a hard pull back from a dive and then hearing/feeling the wings "pop" when they finally loaded with air. The feeling of stability from the airflow giving way to the balance of the aircraft at the top of a wing-over.

[fliger747]

I was somewhat involved in PMDG's development of their 747's for FSX. Having some 12 years flying captain on said airplane I do think they are pretty good as far as performance and flight manners. Aircraft with such a myriad of high lift devices are not particuarly easy to model in FSX.

Actually most of the flight dynamics are hidden in the aircraft module, using much wizardy to get around the limitations of the FS engine. Just looking at the .air and .cfg files won't tell much of the story.

Cheers: T

[Ken Stallings]

RoF feels "right" to me. Those early planes were quite the handful to fly. But RoF has only modeled the 1914-1918 era, so far. How would that sim represent a FW-109, or F-86? Or Ken's C-310R? Comparing "study sims" with "survey, or general" sims is inconclusive, at best.

I was about to mention that point myself, but figured I had already said enough in my first post!

But, you are right. Those World War I fighters (especially the Camel and Tridekker) were absolute death traps. I'm not sure a flight sim could make them unstable enough to be realistic!

Ken

[warchild]

I was about to mention that point myself, but figured I had already said enough in my first post!

But, you are right. Those World War I fighters (especially the Camel and Tridekker) were absolute death traps. I'm not sure a flight sim could make them unstable enough to be realistic!

Ken

get me drunk enough and i can make any plane fly like C%%p.
truth told, now that i have a grasp on the su-37 i can say that yes, the instabilities can be programmed in in fsx. i dont do anything with X plane so i dont know.. but with a little effort, those negative characteristics could be modeled,. it just isnt easy..

[Ken Stallings]

welll, i would humbly submit... http://www.iasa.com.au/folders/Safet...nacleCRJ2.html
The Faa and NTSB had a field day with that one.. you can be sure of it..

Why Not use the main physics engine for every single plane??? because although the engine is quite capable of making use in the differences between wingspan wing root fuselage length and various other physical attributes, it doesnt take into account any of the myriad of differences that exist between each plane. For example, it doesnt model the difference between a CFM-56 and a JT-3. it treats both of them the same, if you dont have anything but zero's in your air file.. Thats why i use air files. I want it as true as i can possibly make it..

PS.. The CFM-56 is used on a 737, the JT-3 powered the SR-71 at over mach 3. ( for those who didnt know )

PPS.. I didnt call airline pilots over all idiots. I call this group of idiots idiots..

ppps. this discussion goes OT, but covers the above mentioned's causes and effects.. it is done by real pilots.. http://forums.jetcareers.com/technic...ow-cruise.html

The talk about idiots suddenly reminded me of Johnny Damon's claim that the Red Sox were "just a bunch of idiots," because they didn't put any stock in the so-called "curse of the Bambino." Funny thing -- these idiots killed that curse by winning the World Series that year and putting the Yankees on notice that their mojo was nojo by coming back from a three game to zero deficit to win the ALCS.

But, anyway ... I digress!

I think the only way to make a particular aircraft accurate is to have pilots who fly that actual aircraft type and model to "wring it out," by flying it and making detailed observations about how well it matches the real airplane.

If you can get a good partnership between real-world pilots and talented FDE modelers, then you can achieve really nice results. But, while I don't mean any insult, the only way to know is to recruit those who fly the actual aircraft types and models.

It's not merely matching performance tables either. For two reasons this is true. First, let's face hard reality, manufacturers in the dog-eat-dog of GA often outright lie in the performance tables! Yeah, I know, sacreligious! But, it's true!

Also, even if you nail the real world numbers plum, there's also the point about doing it accurately! That requires a measurement of feel. One example is modeling what happens when the first notch of flaps are dropped at a given airspeed. You need a real-world pilot who has experienced the result to tell the FDE modeler if the reaction is right in immediate pitch-up (or down) and altitude plus airspeed change.

When the plane stalls, how does it react? That's not in a manual. It takes actual experience to measure the virtual replication. So, again, I say FSX can achieve very realistic results. But, it takes a lot of tweaks and tailoring. And yes, due to some limitations in the code for FSX, it also means perfection is not possible. You just try to make the compromises as small as possible and hope the real-world pilots don't complain too loudly!

Cheers,

Ken

[delta558]

I don't know that "death traps" would really be a suitable phrase for the early machines, it was probably more the inadequate training time during the first world war. Okay, they may not fly in such a straightforward manner as many modern machines but take, for example, the Shuttleworth collection - they have a large number of aircraft of that era and a very impressive safety record. Worst I can remember is engine cowlings detatching from the Sopwith Triplane.

[Ken Stallings]

I don't know that "death traps" would really be a suitable phrase for the early machines, it was probably more the inadequate training time during the first world war. Okay, they may not fly in such a straightforward manner as many modern machines but take, for example, the Shuttleworth collection - they have a large number of aircraft of that era and a very impressive safety record. Worst I can remember is engine cowlings detatching from the Sopwith Triplane.

To a degree, you are right. Those pilots were barely trained -- I don't even think the word "inadequate" is adequate!

But, speak with the handful of highly experienced pilots who today fly the replicas using the rotary engines. They will tell you they are very dangerous aircraft to fly. Some of these pilots simply will not fly the replicas with rotary engines on them, but instead use radial engines.

Yet, the replicas still use modern fittings and repair and maintenance methods. The originals had none of this working for them, including even the concept of quality and consistent construction!

Ken

[srgalahad]

It's not merely matching performance tables either. For two reasons this is true. First, let's face hard reality, manufacturers in the dog-eat-dog of GA often outright lie in the performance tables! Yeah, I know, sacreligious! But, it's true!

Ken

That's one of the essential points. Rarely does the modeler or FDE expert have access to sufficient and correct data. For the preponderance of modeled a/c most of the data is either limited or totally suspect.... not blueprints and engineering flight test reports but usually PR or propaganda (ours is PR, theirs is propaganda) - whether it came from the USAAC, the Reichsminister or around Churchill's cigar. Some people sneer at Wiki, but those same numbers appear in the 'history' books or sales brochures and are what was released, not necessarily what was found in test because that was 'secret' and a lot of files got tossed out as irrelevant afterward. Granted there are some a/c where the model is worked from the real data and I bet those FDE designers can attest to the bruises -in any sim.

The more limited the scope of the sim (RoF, CFS, etc.) the more possible -no guarantees - that the game engine can be optimized to fit a performance range and world environment. However age factors in... as aviation grew rapidly 1915-20 there was a certain lack of precision in note-taking, or knowledge about what detailed factors were coming into play. Add in the short careers of pilots and planes and something like RoF is, to me, suspect but the key is that all a/c in that small universe are likely to be 'comparative' so it's not obvious if there is a flaw - we just don't know. When you get into a broad-spectrum sim like MSFS it is inevitably headed for "You can't please all of the people all of the time" because the MS boys had to build within a production cycle, users system specs and marketability - knowing the vast majority didn't know manifold pressure from hydraulic pressure, or what stall speed was. Would they build a $50 sim for the 10,000 knowledgeable customers, or market to the mass 500K that just wanted a new computer challenge? In the end, technology and customer needs ruled. "pretty close guys! - box it and ship it"

Because I've been closely involved in aviation since I was 5, obtained a license, taught groundschool, studied the history and worked in the industry I can't accept flying a cartoon ...but I've also learned that 'close' is pretty good and I love it when it is "wow!". If I can do bits to make it "closer" then I've contributed. Based on the mass market users I've had to deal with, when you see that the biggest gripe is that a model is "too hard to start and to fly" and "things fail", I wonder if we are chasing our tails to a degree, but I also acknowledge that we do it because we have pride of craftmanship and love the challenge.

As to the specific question: "which FS models flight the best", it's an interesting question. I have yet to see any specific comparison tests to give an indication, let alone a definitive answer. Show me the money, honey!

As much as all of us (the small minority who really care and may have knowledge to judge) would like to be reassured that any one sim is 'more realistic', would we switch because of it?
Would we buy one for better flight dynamics with limited choices, and one for scenery, and one for greater variety?

So, for me, FS9/FSX is enough. It doesn't do any thing perfectly, but it covers all the bases satisfactorily.. it has scenery, it has an abundance of freeware and great payware, it has a whole populated world and covers the 107 years of powered flight so I can fit the session to my moods and desires. That whole overrules "best flight dynamics" any day when my credit card feels unloved. If my interest could be limited to a time period or genre, then it might be a bigger issue. As is, flight modeling is my first criterion, but can't be the only one.

From birth until death, life is a series of compromises.

[fliger747]

Sir Rob is correct about the uncertainty of much published data. One of the better sources of condensed and consolidated data on WWII fighters in America's Hundred Thousand. Interesting as it gives both mfgr and service test data, plus combat and mil power data. One does have to remember much of this data was aquired by a test pilot reading instruments and writing on a knee board..... The SOH A26, done for FS9 does pretty well performance wise in meeting the specs and flying well. Much formerly classified data is available now, but we do suffer from many limitations in mimicking piston engine altitude vrs power/thrust performance that dwell withing the limitations of the FS series. But still good approximations are possible through elaborate gauge programming.

I didn't see a response to my comments about PMDG. On the 747-400 project we had two high time 747-400 Captains, one of whom was a former MIG test pilot, a 747-400 maintaince engineer plus the whole gamut of programmers and artists and testers. I use the thing to practice for the real twice yearly bet your job events in the "big sim" including engine out on takeoff scenarios (V1 Cuts) etc anc consider it very positive training. It would be very difficult to plane jane this aircraft which has many automatic features such as "speed trim" and three different types of high lift devices.

Cheers: T

[warchild]

Sorry Fliger.. I didnt mean to seem that i was ignoring you.. I've been a little bit busythe last couple days..

I'm afraid that irregardless of what you use FS for, including the 747, that when reduced to its base components of nothing but the fsx executable, the aircraft config file and the air file, that FSX cannot accurately model all the aircraft its presented with by merely changing the variable data within the aircraft config file. Such issues as compressor stalls and windmill restarts cannot be addressed in the config file for example.. They CAN however be accounted for in the air file, but not if the airfile is a bunch of zeros.. Your PMDG model is a very very complex and systems dependent model, but I'll still always maintain that the FDE sux because in my opinion, the systems got too much attention and the fde was passed off like yesterdays news. Thats my opinion.. Sorry if it isnt popular..

[bkeske]

To Ken Stallings:

Thanks a bunch for the info to your 172 mod. I got your file, downloaded it, switched out the gauges to RXP units, and the radio units to my Friendly Panel 'nav pack' units. The difference between this and the default 172 is quite amazing, and very well done sir; Hats off to you....I kow have a very nice 172 perfect for both hand flying, and also hand flying with instruments, or even switching on the AP if I want.

BTW, what started as a fairly 'snarky' thread has turned into a very informative one. This has to be one of, if not the best flight sim forums we have. It is threads as this that reminds me why I love this place.

[fliger747]

In what way does the PMDG 747-400 not fly as the original? A someone who has been flying the 747 since 1996 I can't answer that question. To the point however is that within the confines of standard FSX programming accurate flight dynamics of such a plane are probabnly not possible. Hidden within the PMDG module are the things that make it fly, away from prying eyes.

I guess that the whole discussion is really about what can be done in the native flight sim. Most of the limitations of the fS series have been well documented,here and elsewhere.

Cheers: T

[Brett_Henderson]

I've always been intrigued by the possibilities of flight-modeling outside of the actual flight- modeling. The easiest path is to use "invisible" gauges that read/send parameters, and make certain things happen under certain conditions... like an XML file that will increase lift (even if by some bizarre mixture of thrust and flap functions) while in ground-effect. A creative modeler can even incorporate this stuff into model parts themselves.. The variables available for "read/write" are enormous.

Some of the payware modelers have done wonders.. My simming is a social focal point for one of the clubs I fly with... so I've had the luxury of breaking in a few testing pros. Once you get them comfortable with FSX itself.. they learn to interpret flight-models for which they're familiar, pretty darn well. The club's senior member is a retired AirForce C130 pilot and has probaly passed the 30,000 hour mark (most of it instructing), as I type this. He's utterly amazed by what CaptianSim has done with their C-130.. And for reference, a retired airline heavy pilot speaks highly of the PMDG 747.. especially compared to the defaul 747.

What I've learned from watching these guys grow to appreciate FSX, is my tenent that what REALLY matters, is what the sim-pilot experiences in front of the monitor, and in response to the controls and instruments... I doesn't matter if your air/cfg files look like someone threw random number in there.. or how many XML files you have running to "fool" FSX.. so long as the sim-pilot gets predictable, accurate performance across the entire performance envelope..

[bstolle]

So far if it's the correct take off distance, climb rate etc.. I haven't found any serious things to be impossible by just using the air and cfg files. E.g. not even the questionable 'brake stand' take off needs an 'external gauge' like in the A2A Cub, just careful airfile design.

[Brett_Henderson]

So far if it's the correct take off distance, climb rate etc.. I haven't found any serious things to be impossible by just using the air and cfg files. E.g. not even the questionable 'brake stand' take off needs an 'external gauge' like in the A2A Cub, just careful airfile design.

Agreed ... In fact, you can get pretty close without even goin into the air file.. unless of course you're modeling something that has no comparible air file..

Here's a example routine :

Raising thrust to the point where model will rotate and climb (per its loading), will almost certainly give it thrust that will yield unrealistically high, level speed.. and probably radically high climb performace. This is where wing-efficiency and the two drag types come into play. I've found that wing efficiency effect is proportional to airspeed.. so you'd start by reducing it. At low takeoff speeds, lift isn't altered as much as at cruise speed. So.. your model now uses more of its thrust in level flight to maintain altitude, hence lowering cruise speed, without dramatically altering rotation speed. Samefor the two drag types... one is more a function of airspeed, the other more a function of AoA.

Between those 4 parameters (thrust, wing-efficiency, P-drag, I-drag) you can get a model more accurate than most simmers even care about, or even know..

Going further into the air file can fine-tune things, to a point... but if you started with an air file that resembles your model.. all you'd likely have to change in the air file, is things like gear-drag, and subtle engine stuff (mostly gauge responses).

[bstolle]

Interesting, I'm just doing the opposite. I use the cfg file for just the most basic stuff like engine location, contact points and geometry. 99% of the 'real' FDE design I'm doing within the air file.

[Brett_Henderson]

Interesting, I'm just doing the opposite. I use the cfg file for just the most basic stuff like engine location, contact points and geometry. 99% of the 'real' FDE design I'm doing within the air file.

That works too.. but it's going about it in a way where you've got variables affecting variables.

Let's say you set up the geometry accurately in the cfg file. Aside from interaction tables, any geometric tweaks in the air file, still give way to any corresponding cfg entries. And things like 'effectiveness' can be more sensibly altered in the cgf file via control-surface areas, locations and deflections.. and if need-be, 'Flight_Tuning'.

And the generic, interaction tables in the air file are as good a 'starting point' as any. I mean you ARE gonna change things.. might as well limit the tweaking to the self-documenting, and easily edited cfg file. Chasing your tail in the one file or the other is hard enough.. let alone BOTH file chasing each other... :isadizzy:

What I'm getting at, is a kind of like a , "The Emperor Has No Clothes", deal. You could sit down with a well-versed Aerodynamics Engineer.. have him go through the air-file and meticulously edit all the tables and variables with 'real' data.... and then have an engine expert plug in all the power-plant data... and what you'll end up with, is an airplane not even CLOSE to realistic performance accross the envelope... it might not even be flyable. And by flyable, I mean that if you had an experienced test pilot for that model (who obviously understands the limitations and use of FSX) testing it, he will be able to not only fly it, but predict the performance to the point where he can use real performance charts and realistically plan things like takeoff distance, climb-performance, appraoch handling with different flap-settings.. landing distances..etc.... no surprises.

If that type of flight-model is your goal, then starting with anything more than, 'close enough', is a waste of time. By the time the model would pass a performance checklist like I use.. the cfg (and even air) files will be bizzare compromises...

In one of my models.. in order to get harmony between the elevator/ailerons/rudder for everything from gentle, cruising turns, to X-wind approaches.. the elevator and rudder weren't even near each other (cfg apexes)... and the MOI/control-surface-deflection entries might leave you scratching your head..lol

[Major_Spittle]

For those that haven't played Rise of Flight. For the cost if you like WWI aircraft it is worth the money. You don't fly these planes without using the rudder.

[fliger747]

I don't know if I was the "retired" heavy pilot mentioned above, still actively flying the "whale" and just back from (yet another) Round the World in four plus days trip....

I once made a test Corsair FD that had "two" engines (co-located), in an effort to try to get around some issues with the way FS models Superchargers. A fruitless avenue of adventure, however experimentally tweaking the power and drag upwards one still got the correct level speeds, but talk about rate of climb!!

Overall we can do surprisingly well with the MS FS series. That we can produce any sort of plane we want and make it fly (more or less) and fly it anywhere is one of the great attractions of the open system.

Is FSX the end of the line? In the year XX will we still be using flight sim X? Will the hardware still run it?

Cheers: T

[Brett_Henderson]

I don't know if I was the "retired" heavy pilot mentioned above

No.. this is a guy in our club. Funny side-story, is that he's the guy who landed gear-up , first time he flew our 210..

"I'm used to having a first-officer"..

[Bone]

The Little Bombardier RJ-200 stalls just above 40000 feet. The engines flame out. If a specific set of conditions arent met ( turbine speed etc ) those engines will not start back up. More than one RJ has been lost because of that, and it keeps happening because there's idiots out there who think pushing the plane above 40000 feet is some kind of macho milestone.. Kinda like the Mile High Club..one you like. Pam

PPS.. I didnt call airline pilots over all idiots. I call this group of idiots idiots..

Those are pretty ignorant things to say, judging all who've gone to FL410 in a CRJ-200 just because that Pinnacle crew totally screwed up. And FL410 is in the POH as max operating altitude...not FL400, and it doesn't flame out above FL400 as a matter of course. I've read the transcripts from the Pinnacle flight, they would have melted the engines even if they had stopped at FL380. They crashed because they were acting like idiots, if I may borrow from your lexicon. But there's plenty who went to FL410 and are still around because they did it the right way...not like idiots. I'm only aware of one incident like the Pinnacle debacle, but you say it keeps happenning...OK, I'll say you don't know what you're talking about.

I flew the CRJ-200 from 1999 to 2004, and took it up to FL410 twice. We were fairly light and it was in the winter (cold temps), we also referred to our altitude capability charts before we initiated the climb, and we didn't climb in a way that killed off our airspeed, either. And, we did it to top the weather, not to join the "club".

Seriously, man.