Sign in to follow this  
Guest cw1011

Flight dynamics problem in Engine not hardware???

Recommended Posts

We have had many discussions here regarding flight dynamics, lack of feel, control forces.There have been many who have defended the flight dynamics engine and its capability blaming lack of feel and control forces on the multitude of control hardware devices rather than a deficiency in the flight dynamics engine itself.It crossed my mind that maybe we are not seeing the wood for the trees in our thinking and that the fault lies cleanely with the flight dynamic engine itself and not hardware, springs etc.As you know at slow speed an aircraft requires a lot of control movement for a given reaction compared to high speed.Flair an aircraft at slow speed to land and approaching the stall at touchdown you may have to pull the control column in a real aircraft to almost your chest.the column moves almost 9 inches in such a situation.The same goes for the rudder.Rotate an aircraft like the Bravo at VR and it takes a very hefty heave to pull the aircraft up to ten degrees, yet we blame hardware for this lack in our sim not a deficiency in the flight engine itself.At fast cruise the movement required can be as little as 1/4 inch to pitch 10 degrees from level flight.The faster you go the more control force is required to diplace the aircraft from its trimmed state and the less control movement that is required.I have often complained that many thousand hour pilots have problems flying FS2004 aircraft.We as sim pilots adjust to flying the sim making tiny movements and pitching to degrees on the A/H rather than by feel.It is this lack of capability on feel and simulating forces which is the missing ingredient.It should surely be theoretically possible to have a control column that has NO spring forces or resistance built in the hardware whatsover and yet have the aircraft responding throughout its envelope in a correct fashion because the flight engine itself controls the behaviour Not the stick or column.The stick or column should be just an input?These are just thoughts and need taking apart by the really clever people here who understand the flight dynamic capability more than I do?CheersPeterhttp://forums.avsim.net/user_files/94188.jpg

Share this post


Link to post
Share on other sites
Help AVSIM continue to serve you!
Please donate today!

Couldn't a module be written that will adjust the amount of what the sim reads as control surface deflection and makes it relative to a curve that is related to the speed envelope given in the aircraft.cfg file? Thus as the aircraft slows to stalling speed it take much larger control movements to affect the aircraft... it gets sluggish and control is not as easy to maintain.I'm with you Peter with what you're getting at. Both my father and cousin (high time pilots) got completely confused when they first got behind the sim and thought that it plain sucked! Citing the input device (a CH USB Yoke in this case) and the way the aircraft reacted to it as the main cause.Cheers,Zane

Share this post


Link to post
Share on other sites

Peter,I have no problem seperating aircraft flight dynamics from control forces. I never mix the two issues. One could even test the flight dynamics using keyboard alone.Very nice Bravo - looks spanking new!Michael J.WinXP-Home SP2,AMD64 3500+,Abit AV8,Radeon X800Pro,36GB Raptor,1GB PC3200,Audigy 2, Omega 2.7.90 (4xAA 16xAF)

Share this post


Link to post
Share on other sites

The problem is one flight simulator flight control peripheral may input twice the amount of control than another. Without a calibrated properly rigged control system, one doesn't know how much input is being made into the system for any given change in position of the flight control.Secondly, the flight simulator user doesn't have data available from actual aircraft to test if the flight simulator is operating correctly.W. Sieffert

Share this post


Link to post
Share on other sites

>>>>There have been many who have defended the flight dynamics engine and its capability blaming lack of feel and control forces on the multitude of control hardware devices rather than a deficiency in the flight dynamics engine itself.:)I try to stay out of these discussions about flight dynamics, there is so much misunderstanding and confusion (and not a little partisanship from some) about this subject. Information about MSFS flight dynamics is getting harder to come by with the increasing influence of payware and the proprietary baggage that goes with that territory, such as non-disclosure agreements, so I'll say a little here.The flight dynamics code that Microsoft uses is very sophisticated Peter. I have read that they purchased it from an outside source and adapted it to MSFS. That may be true, but what is undeniably true is Microsoft doesn't allocate enough resources to programming the stock aircraft to fly properly. I'm being kind here, an alternative viewpoint would be that they don't have a clue about aerodynamics or how to use this code to cause their aircraft to faithfully reflect the real-life counterparts, but these days I'm leaning toward the "that's good enough" scenario, recent events seem to confirm this. Regarding the control forces vs. airspeed, this is allowed for in the flight dynamics code, airfile tables 517, 518 and 519 (when present) provide the means to model the different stick forces experienced with elevator, ailerons and rudder over the complete range of the flight envelope. Competent third party flight dynamics will take advantage of this capability, but as Microsoft notes; the average gamer will not know the difference. You can't simulate the stick forces under your hand by programming the flight dynamics, only the movement required to produce the same result at different airspeeds can be modeled. For this reason I prefer to use these tables (517, 518 and 519) to model only the effect of reduced control response at lower airspeed, primarily the ailerons. You can imitate the effect of high stick forces at high airspeed by reducing the response, but to me it feels wrong, more like a cable failed or slipped off a pulley since you can move the stick as before, but with greatly reduced aircraft response. At low speed, it feels realistic, especially for aircraft with landing speeds under 80 KIAS. Some type of intelligent force feedback capability is what we need, and I'm sure it would be possible to model the air loads and stick forces in a realistic fashion, but now we're back to what the average gamer knows, and feeling the stick rattle now and then probably satisfies their thirst for realism. Microsoft is out of the hardware business now, so I have little hope for the realization of this concept, perhaps a competent third party developer could take up the challenge.That brings me to joysticks and their effect on flight dynamics. I won't address the joystick vs. flight dynamics subject in any great detail beyond what I've already written about force feedback. I've seen forum members here who will argue that the Microsoft flight dynamics programming is perfectly fine, and anyone who thinks otherwise needs a new joystick. I've used many joysticks over the years, and I've had hardware and software troubles (calibration) with some that left me briefly wondering what happened to the aerodynamics of my aircraft, but even with the cheapest controller ($2.00 electronics surplus) available, I've never seen degradation in aircraft response that was appreciable in contrast to using the most expensive joysticks. I have no expectation that the most finely machined custom joystick available, using the best components money can buy will be the cure for all or even some flight dynamics ills.There is some truth to the notion of the control interface being responsible for some of the problems with MSFS flight dynamics, the evidence lies in well known aerodynamic constants needing to be set differently than the real world values, but we have to work within the limitations of the current version of MSFS, and change the airfile to suit our needs.Also, you need to separate the flight dynamics from the flight model. The flight model is how the version of MSFS reads the flight dynamics code in the airfile. You can adjust it in the realism menu. I'm sure most people have heard that FS9 doesn't read the airfile code in the same way that FS2002 did, or the way FS2000 and CFS2 did. Angle of incidence of the wing and the aerodynamic effect of same is no longer read in FS9, and surprisingly, for a simulation with vintage aircraft using conventional landing gear, the effect of prop wash on the elevator is also among the missing. So the trend seems to be moving away from the flight dynamics code that allows for the most realism, to a simplified version that will please the average gamer that represents the greater portion of the customer base.I've been saying for quite some time that more realistic flight dynamics do not mean that the aircraft is harder to fly, quite the opposite. Avcomware stated the same thing in the recent topic you started on this subject, and I'm certain that as a pilot you're aware that this is true Peter. The question remains, does Microsoft? And if they do, do they care? Pleasing 1% of the customer base (the power users) is probably not a high priority item on their agenda, and perhaps bizarre and difficult flight dynamics (along with a slider to adjust things from easy to hard) is a concept more concerned with gaming than flight simulation, i.e. if you

Share this post


Link to post
Share on other sites

Hi, Peter.The main problem is that MS flight model is based on a number crunching engine, not air pressure on flight surfaces. This makes it difficult to scale the controls to the aircraft behavior, as we, pilots, know it. Some people seem to think that controls and flight dynamics are completely separate, these are the non pilots, or pilots that do not fly on regular basis, and they lack the

Share this post


Link to post
Share on other sites

MichaelThe Bravo is my new baby, my new toy :-)and I am well paid to fly her. I have been very lucky in getting a regular position flying this aircraft and we are flying interesting destinations.Yes she is almost new with 600 hrs total time and full EFIS primus 1000.CheersPeter

Share this post


Link to post
Share on other sites

DouglasThanks for your very detailed and informed views. If we take for simplification of description the control column/ rudder pedals totally devoid of any springing or artificial resistance there are two areas which spring to mind.Firstly as Zane mentioned in his reply above the control movement shouldnt be static for a given effect but variable to a speed graph.By this I mean at slow speed towards the stall there should be large control inputs required for a given control effect.As speed increases this should reduce to a minimal control movement for large effects.I ran some tests in a Seneca five and in the flair reducing to the stall there was several inches movement required for a desired effect.The same went with the rudder.speeding up to the cruise I attempted in this test to pitch the aircraft to 10 degrees.The control movement apart from becoming rock solid and very resistant to moving from its trimmed state only moved 1/4 of an inch.The same went for the rudder forces.Until this aspect is addressed we will never have a realistic flight sim experience yet it is in the realms of coding.Secondly the way the aircraft resists being moved from its trimmed state needs to also be coded.This should be apparent in the time the aircraft takes to move relative to its speed through the air.Pitch an aircraft from a relatively high speed and the initial response would be high forces plus a slow time rate to move from its trimmed state to say 10 degrees.As the speed drops, more control movement is required and the out of trim forces reduce.These effects are apparent through all controls yet poorly modelled in the sim.So we have two areas of "feel" control movement relative to speed and time to respond also relative to speed.Neither is realistically modelled yet both are vital if we want a realistic sim aircraft.Control hardware do need to offer resistance in the form of springs or rubbber bands but only to simulate out of trim forces.Only then will we get something approaching realism.I take your point on the conflict between the hardcore simmer and his expectations and Joe Bloggs average buyer who would nt give a T+SS or frankly know any better.For me Microsoft have done a wonderful job with the eye Candy in the sim but now need to address the structures under that Candy not only in flight dynamics but also in the weather and atmosphere as the relationship between the two is very important.Peter

Share this post


Link to post
Share on other sites

TvThanks too for your considered and experienced contribution one thing I would add to one of your comments which has cropped up before, relates to this extract."If a separation of flight controls from surfaces exists, be it due to software and or use of fly by wire, like the AirBus, and some fighter aircraft, the feel is lost."Regardless of whether the controls are direct linkage or indirect by fly by wire the forces are are still present and real.Ie with fly by wire the aircraft will still be reluctant to move from its trimmed state and large control v small control movements will still be required for a given effect what will be lacking is the out of trim forces which the motors which operate the control will still experience but the pilots controls will not.Best regardsPeter

Share this post


Link to post
Share on other sites

Hi, Peter.>Ie with fly by wire the aircraft will still be reluctant to move from its trimmed state and large control v small control movements will still be required for a given effect what will be lacking is the out of trim forces which the motors which operate the control will still experience but the pilots controls will not.

Share this post


Link to post
Share on other sites

>Some people seem to think>that controls and flight dynamics are completely separate,>these are the non pilots, I take issue with this sentence. I am one of those "some people" and I happen to be a pilot. I can seperate flight controls from aircraft dynamics simply because (in real life) I can provide some input to the aircraft and let go off the controls (or even better - fix them in one place with a vice or crazy glue) and watch the airplane do its thing so Peter's original statement about the controls being only the source of input is a good one. But anyone who knows physics (and I happen to be a physicist) knows very well what Newton's 3-rd law of motion says and that when you apply some force you get some feedback back. And Bob's expertly post above clearly demonstrates this force may not necessarily be coming from air pressure loads on flight controls surfaces.Michael J.WinXP-Home SP2,AMD64 3500+,Abit AV8,Radeon X800Pro,36GB Raptor,1GB PC3200,Audigy 2, Omega 2.7.90 (4xAA 16xAF)

Share this post


Link to post
Share on other sites

>"If a separation of flight controls from surfaces exists, be>it due to software and or use of fly by wire, like the AirBus,>and some fighter aircraft, the feel is lost.">>Regardless of whether the controls are direct linkage or>indirect by fly by wire the forces are are still present and>real.>Not necessarily. Many aircraft have hydraulic flight controls, and the controls themselves only move actuators. Any "feel" is produced artificially with springs...and any change in required control loading force due to air loads is also artificially fed back by use of variable spring tension. That feel is a direct function of the flight control system design, with some secondary affect arising from flight dynamics as it interacts with the control system (in those acft equipeed with a variable Q mechanism in the artificial feel system). Capturing this is not deterministic solely w/r/t flight dynamics, and thus can't be modeled by a one-size-fits-all FDE engine.>Ie with fly by wire the aircraft will still be reluctant to>move from its trimmed state and large control v small control>movements will still be required for a given effect what will>be lacking is the out of trim forces which the motors which>operate the control will still experience but the pilots>controls will not.No, I think you're confusing movement with force, and stability with feel. The F-16, for example, uses a nearly immovable stick with strain guages that measure the amount of *force* applied to the stick...not movement. Likewise, that airplane is inherently unstable and uses flight control computers to produce an artificial sort of neutral to slight positive stability. That airplane is not at all reluctant to move from its "trimmed" state (the concept of trim being something of an artificiality since it's really just an arbitrary flight condition which the FBW computer is working furiously to maintain for the pilot).I use a floor-mounted PFC yoke with a heavy spring loading system that does a credible job of producing the kinds of control forces I am used to in a medium to heavy jet transport with hydraulic flight controls. Given that the real planes also use springs to generate the control loading forces, it seems to follow that such a result is possible with a good design. For a direct-linkage FCS, any sort of loading feedback that would even halfway credibly simulate variable air loads on flight control surfaces is presently beyond the means of the casual simmer's budget. For that, we have to accept the limitations of the available controllers.RegardsBob ScottATP IMEL Gulfstream II-III-IV-V L-300Washington, DC

Share this post


Link to post
Share on other sites

Hi, Michael.If you are satisfied with flying with the keyboard, I think that you are wasting your time reading this thread. Over the years I

Share this post


Link to post
Share on other sites

Peter,I agree. I hope the oft-repeated rumor that the MSFS development team lurks in this forum is true, and that they see fit to at least consider some of these ideas. I also have to agree about the eye candy, it

Share this post


Link to post
Share on other sites

>If you are satisfied with flying with the keyboard... and where did I say that I am satisfied with flying using the keyboard? I never fly using the keyboard. I think you should go back and read again what I quoted you saying and what I was repying to - whether you can separate (to a degree) aircraft flight dynamics from control devices. Yes, I am learning too, some of the stuff Bob is talking about I clearly had no idea about.Michael J.WinXP-Home SP2,AMD64 3500+,Abit AV8,Radeon X800Pro,36GB Raptor,1GB PC3200,Audigy 2, Omega 2.7.90 (4xAA 16xAF)

Share this post


Link to post
Share on other sites

>Not necessarily. Many aircraft have hydraulic flight controls, and the controls themselves only move actuators. Any "feel" is produced artificially with springs...and any change in required control loading force due to air loads is also artificially fed back by use of variable spring tension. That feel is a direct function of the flight control system design, with some secondary affect arising from flight dynamics as it interacts with the control system (in those acft equipeed with a variable Q mechanism in the artificial feel system). Capturing this is not deterministic solely w/r/t flight dynamics, and thus can't be modeled by a one-size-fits-all FDE engine

Share this post


Link to post
Share on other sites

Hi.If I'm reading this correctly it appears that two different issues are getting twisted together.I think what was originally being said was that the *input* requirements were unrealistic. To get a certain change in attitude a user should have to move his/her joystick by X amount when in normal flight conditions, but X*3 (or whatever) amount when moving slowly. Currently the FS engine doesn't seem to do that.Joystick calibration will have an effect of course, but only on the *absolute* distance required; a stick with a tiny amount of travel should still require 3x the movement, even if the actual amount of movement is much smaller than with a different stick with a huge amount of travel. In other words, calibration affects *absolute* movement, not *relative* movement.Force-feedback is an entirely different question - it's output, not input. As somebody pointed out, the fact that you may have to work harder to move the stick a certain distance is irrelevant to the result on the controls. If (heaven forbid) the feedback system in a real A320 which somebody suggested was done by springs were to fail, the aircraft would (should?) respond in exactly the same way for a given input; the difference would be that the pilot wouldn't *feel* the same thing.We really need to separate this because even though it's not usually realistic to assume no feedback whatsoever, most FS users probably don't have force-feedback sticks. Given this, isn't it reasonable to assume they're flying in the situation where a spring has twonged off into the cargo holds? Their *input* to the stick should do the same thing to the aircraft even if the aircraft - whether by cables, springs or whatever - can't do the same thing to the stick.Just my 2p-worth :)Cheers,Jim Keir

Share this post


Link to post
Share on other sites

Very interesting thread. While we are at it, I wonder what sort of flight modelling full-size level D simulators use? And how much processing power they need to run? Do they base on a common software?And could those flight dynamics & weather engines replace those we got now in FS?Why should MS develope their own engines if there are very realistic ones out there in the airline industrie?There are 3 reasons, though, I can imagine this not to happen: processing power needed, money to buy it and the US goverment not wanting terrosist being able to buy a highly realistic sim for 50$.wondering,Claudio

Share this post


Link to post
Share on other sites

>There are 3 reasons, though, I can imagine this not to happen: processing power needed, money to buy it and the US goverment not wanting terrosist being able to buy a highly realistic sim for 50$.wondering,Claudio http://forums.avsim.net/user_files/94324.jpg

Share this post


Link to post
Share on other sites

Peter,I'll second that. Maybe we should send the Microsoft team a copy of Richard Collins' DVD series on advanced weather flying. Now if they could get MSFS to emulate what he teaches on that set we'd have some fun, challenging, and at times surprising weather to deal with. That screenshot in your last post is close to the view angle ROTW included with their latest Seneca release. Wish I was flying instead of here behind the computer!Take care Peter,ZanePS Curious if you tried out the T37 mentioned in an earlier thread? I bought it as a result of that discussion and was quite impressed, would love to hear your impressions.

Share this post


Link to post
Share on other sites

ZaneWeather in my opinion should be the flagship ingredient of the next version of MSFS.Microsoft have the visual candy down to a fine art now they need to add the meat to that candy.I would love to see downloadable weather systems as per real world with highs lows, fronts and their associated cloud formations, visibility, pressure chages,isobars, winds, turbulence, ice, temperatures. The list goes on :-)Then as I have outlined here variable control movement effect which is speed related and more control difference throughout the speed range.Nb the Seneca is going to be "work in progress" I believe ;-) superb work from all involved.Peter

Share this post


Link to post
Share on other sites

Peter,I've been using PS1 (www.aerowinx.com) up to v1.3a, and I find Hardy's approach rather well thought. Not being ready to program any sort of FF into the core of PS1, Hardy decided to "translate" force into "course", thus requiring biger control deflections whenever increased force should be used in reality.This approach results in a very convincing experience, even for me who have already had the chance to fly a Level D sim.On the other hand, there are so many more superficial things about the way MSFS and other sims approach the simulation engine that make me doubt about it's usefulness for, say, IFR training. Take the simple "Turn & Slip Indicator" as an example:- When you fly the default or even the best add-on GA aircraft for MSFS, you generaly find the turn & slip 2 minute marks for standard rate turns. Try to measure the time it takes to complete a 360 degrees turn, or even a 180 degrees one if you don't want to wait for so long... Some will do it in 1' 40'', other in 2' 33'', etc... Very few if any will do it in anything close to 2'', wich is what it was supposed to be the result...

Share this post


Link to post
Share on other sites

I think its an interesting topic. What I have to add is based on my time in the 777 simulator in Denver. I was quite surprised at the force required to flare, and hold the flare, that had been programmed into the fly-by-wire system that Boeing developed for the plane. You really had to pull hard to keep the nose up as the mains touched down. I have probably owned 10 joysticks, and only two give that physical, tactile sense. I have a PFC flight console (which I use for fs9)which includes one of their yokes (built into the console, not the jetliner stand-alone), and the Cougar HOTAS (for LOMAC and IL2) Despite some of the shortcomings in the Cougar (it's tendency to move easilly only in the cardinal directions and much harder when you try to both climb and turn), it's springs and resistance add a real element of realism to the programs I use it for. The PFC is also good, but I would like more even more resistance.I think the sim does a credible job modelling the differences at speeds. I have had many a landing where my trimming was not very good, and I have had to pull good and hard to keep the nose up. Same aircraft at high speeds, the plane moves with very small inputs. So, I think I am still in the "hardware is really important" camp on this. I am still really surprised that folks who are willing to plunk down $1,500 for the latest and greatest balk at spending anything more than $100 for controls.The single most important thing that has enhanced my enjoyment of the MSFS sim is the PFC control console. The heft, reasonably solid resistance, responsiveness and accuracy really help. And that is particularly true for the throttle quadrant, where you really can develop some muscle memory for throttle locations for certain settings.

Share this post


Link to post
Share on other sites

Hi, Colin.>So, I think I am still in the "hardware is really important" camp on this.>I agree with that, but of all the controllers available to date, and I tried most of them, none do a credible job. This would imply that one thing that could help would be MS / the engine. I have a set of PFCs that I installed at the FBO, and we use them with PCATDs and I still cannot get a realistic feel. Interestingly enough, the are, and were, some sims that have a much better controller response / feel. As was mentioned earlier, the feel has to be implemented by the maker of the sim, since the engine is basically a calculator. This, feel, problem exists in some real life aircraft also, since many of the them are hydraulic and some fly by wire. The difference is that they have other interface, software / hardware combination that brings back that feel. Some are doing a better job than others. I know some people that would rather not fly than fly some of the

Share this post


Link to post
Share on other sites

It get's even more complex then that. Every aircraft has a different control yoke set up and linkage, Even in two similar aircraft, the controls can move considerable more or less then the other. Even then, I have noticed some people expect similar body frames to behave identically. I know there are several piper cherokee models out there of various varients. My real world experience is limited to just the "Cadet" and "Warrior III" varients of the Cherokees. While they look exactly the same on the inside with the exact same engines and nearly the same empty weights, the "Warrior IIIs" have a much more noticble nose heaviness at neutral trim than the "Cadets". Another thing is that there are forces in flight that a pilot instincively responds to as well. I fly the real aircraft much better then the $60,000 Frasca simulator we have and have noticed this with my fellow students at our flight school as well. The Frascas sim is a near-exact reporduction of the Piper Warrior III and I know I'm moving the yoke the same, but I still have trouble maintaining stable flight. The hyper-sensitive instruments in the human body affect us in ways we don't even understand (just look at spacial disorientation incidents; no matter how hard you train for it, it still hits you).I know in the sims (MSFS and the FAA certified Frasca) that I have a tendancy to over control. I feel that this is simply from the lack of sensation in my butt that tells me I'm pulling or pushing or banking too quickly. Your body is more of a guage of how quickly you're imputing controls than the gauges are. The gauges really don't tell you how fast your imputting them. When you're looking at the gauge, you're looking to see where you want it stopped, you really don't pay much attention to how fast the needles moving there unless it's noticibly excessive. However, even the smallest movement that looks negligible on the gauges is felt quite well.I invite anyone to take up the Be A Pilot offer and get an introductory flight from a local flight school that accepts Be A Pilot certificates. You may be amazed at the sensations you feel flying the plane for real.----------------------------------------------------------------John S. MorganReal World: KGEG, UND Aerospace Spokane Satillite, Private 130+ hrs.Virtual: MSFS 2004"There is a feeling about an airport that no other piece of ground can have. No matter what the name of the country on whose land it lies, an airport is a place you can see and touch that leads to a reality that can only be thought and felt." - The Bridge Across Forever: A Love Story by Richard Bach

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this