Auto-rudder?

[Alec 370]

I like where this is going. Exactly the kind of thinking I was expecting to hear.In the end it's all about getting as close as possible to the real deal. Unlike Formula 1 cars, which they have to create from zero, with no model as base until it's finished, we can get actual data from the airplanes in every situation we need, and then translate that data to the actual Sim, tweaking the values to get as near as possible to the real thing. It's not about just adding the values from the real aircraft to the simulator, that would NEVER work, not in a thousand years. Nature and it's physics are FAR too complex for any computer to get near to calculating it exactly right, and even if it does get close, that gap can be enough to make the airplane react nothing like it should.I don't like the artificial stability, it reminds me too much of FS, where the aircraft feels not flying though air. Seems it's getting all the codes that were put into the FM and just trimming the ends to get only the middle part. But I think X-Plane is the opposite of that, it's just too wild from what I have tried to get near the actual reactions of a plane, and it will never be able to simulate an aircraft only using Math and physics, it has to be some comparison data to say how it should react under a situation that has been tested in RL.Final point: It's about getting as near as possible to the aircraft which served as model. It's not about letting the physics do it's thing and expect it to correctly give an end result that reacts in every way like the aircraft would.

Edited February 22, 2012 by Alec

[LAdamson 240]

I don't like the artificial stability, it reminds me too much of FS, where the aircraft feels not flying though air.

This is a point, I do not agree on. Many times, FS has reminded me very much of flying through real air. You don't need to be constantly yawing or rolling........to simulate flight.I've experienced flight, in many real life airplanes, ranging from aerobatic, to WWII fighters. Gliders, hot air ballooons, and a Ford Tri-motor too. Sometimes, this yawing/rolling action in X-Plane, would irratate me enough, that I'd usually quit after five minutes. It's as if I'm doing constant corrections with the stick or yoke, to maintain smoothness. In effect, I'm just "chasing it" with the stick. Real life flight, isn't like that. Not even with an aerobatic plane, that just requires thinking of moving the stick, to get it done. If it's adding a bit more "artificial", to what's artificial anyway, then I have no problem with it. Any way you look at it, X-Plane is NOT a real airplane. I believe that some, seem to confuse that issue..........to the point of believing it is.L.Adamson Edited February 22, 2012 by LAdamson

[GoranM 3,764]

Well, I find this to be a TRAVESTY of epic proportions. Carenado should REMOVE THEMSELVES IMMEDIATELY from all X-Plane development and never look back, as they aren't worthy enough to develop functional airplanes without shortcuts and cheats.I, for one, am shocked (SHOCKED!) that such a fine and upstanding development group as Carenado would not at all know what they were doing and try to sell their wares to unsuspecting customers. And these customers were ever so duped into believing that they were flying good products! Even our own Geofa, who flies a Bonanza himself, was duped into thinking that the Carenado X-Plane version of his aircraft felt and flew closer to the real thing than anything that has come before it for a desktop simulator. Clearly he was wrong and his impressions of the product were grossly incorrect, considering Carenado used cheats and shortcuts to achieve what we now learn is nothing more than a hollow achievement.I'm calling the Better Business Bureau posthaste. They will rue the day they tried to sell us poorly written flight simulation add ons! After all, we're told that we cannot judge X-Plane by the aircraft that are included in the demo/box because they are poorly written, and now we learn that one of the premier examples of how great X-Plane can fly is also a poorly written piece of festering cheat-code. Umbrage! Outrage!!(exasperated sigh...)(Editing to add that this post contains a ridiculous heaping of sarcasm, and in no way am I actually calling out Carenado or peoples impressions of the product. I simply find it very interesting that even when you release a product that can be strongly argued to be one of the best representations of the real thing, you can get criticized for apparently not doing it correctly.)

No one said what Carenado (or the x plane developer doing the conversions) were doing was BAD. Some people prefer following the right "roads" to their destination. Using artificial stability is like cutting through the "park".I must admit, the sarcasm was surprising. You usually make very intelligent posts. This one was out of character for you.

[60North 120]

Just FYI, many real aircraft USE artificial stability. A couple of examples;- Yaw dampers- Fly By Wire- CWS mode om APBtw, airflow, I took this interesting shot the other day at CPH. At first glanceit just looks as a dirty A321...

Edited February 24, 2012 by MortenM

[Alec 370]

MortenM, you're the Guru of X-Plane Flight Modeling from what I know. What's your opinion of Artificial Stability in X-Plane? Do you use it with your addons?

[60North 120]

Sure, on the airliners we use it for the yawdamper. Other than that it's mostly suited for aircraft that are extremely fast, extremely light or unstable (fighters, ultralights, RC etc)For CWS we use a plugin. M

[Alec 370]

But do you think it's necessary in order for an aircraft addon like a Bonanza to fly accurately like it's real world counterpart?

[60North 120]

No, I wouldn't use it for GA's (unless it has a yawdamper in real). AI will try to stabilize the aircraft both with andwithout control input so it will feel and behave weird (artificial)M

[GoranM 3,764]

Always good to have Morten corroborate and clarify a technical x plane point. :(

[800xp 0]

This is a point, I do not agree on. Many times, FS has reminded me very much of flying through real air. You don't need to be constantly yawing or rolling........to simulate flight.I've experienced flight, in many real life airplanes, ranging from aerobatic, to WWII fighters. Gliders, hot air ballooons, and a Ford Tri-motor too. Sometimes, this yawing/rolling action in X-Plane, would irratate me enough, that I'd usually quit after five minutes. It's as if I'm doing constant corrections with the stick or yoke, to maintain smoothness. In effect, I'm just "chasing it" with the stick. Real life flight, isn't like that. Not even with an aerobatic plane, that just requires thinking of moving the stick, to get it done. If it's adding a bit more "artificial", to what's artificial anyway, then I have no problem with it. Any way you look at it, X-Plane is NOT a real airplane. I believe that some, seem to confuse that issue..........to the point of believing it is.L.Adamson

Can you please send me some recommended artificial stability settings? Real world pilot feedback is the most valuable tool one can use to evaluate a and tweak a low level (non-professional, not Level D) computer flight model. I am an aerospace engineer in real life and one of my older posts about the limitations and large errors of coarse finite element analysis and blade element theory got deleted because the thread referred to FSX. I always thought my posts on X-Plane.org will be deleted first because i provide feedback about the flight models over there as well, so I'm a bit surprised it happened on this forum. We have the luxury of real flight test data and real world pilot feedback and X-Plane chooses to ignore all of that and solve for everything all over again using a very rough and basic approximation. Great for a conceptual aircraft design tool (with a "back of napkin" accuracy, 30%-50% error) but not so great for a high fidelity computer flight simulator. My main point is you should use all available tricks in the book, whether it is artificial stability, ghost flying surfaces, whatever.. so that the final product matches real world flight test data, real world dynamic and static stability, feeling of inertia, roll rates etc.. and gets a stamp of approval by a real world pilot. The X-Plane simulation engine introduces so much error that lots of tweaking is needed to get to an acceptable level of realism. Edited February 25, 2012 by 800xp

[800xp 0]

I'm not sure I understand the argument that reappears from time to time that because X-Plane attempts to model airflow over a surface, it's inherently accurate and not to be tampered with.As Larry Adamson points out above, X-Plane doesn't move real air over real surfaces - it represents airflow and surfaces mathematically. A mathematical model involves assumptions, and assumptions can introduce errors.In Formula 1 racing - another pursuit I keep track of - there are constant problems with the modeling of airflow. Modeling results don't match wind tunnel results, and wind tunnel results don't translate to the track. Honda's 2008 wind tunnel was calibrated incorrectly - the outcome was a disaster and led directly to Honda's withdrawal from Formula 1. Virgin Racing designed its first car using only computer modeling, because they coudn't finance a wind tunnel. The car they produced was abysmal.So clearly, very skilled, very highly paid people in an intensely competitve meritocracy failed when they tried to generate accurate results using computer models of aerodynamics.Isn't it likely that there are similar errors - not only in the flight dynamics of particular X-Plane aircraft, but in the basic airflow modeling of X-Plane itself?And if that's so, why shouldn't an aircraft developer correct the errors? An adjustment that addresses a flaw in the airflow modeling might produce an aircraft that performs more accurately.The adjustment is only a "cheat" if you assume that the X-Plane airflow model is, to quote another computer, "foolproof and incapable of error."Am I missing something?

You are right on the money, as an aerospace engineer i can confirm that wind tunnels and structural test rigs are here to stay for decades. It will take many decades until CFD is sophisticated enough to simulate a full aircraft flying to the air in real time. At this point we can simulate a small part at a time, in a steady state condition, or a few seconds of a dynamic condition, and there is error that must be accounted for in professional, sophisticated CFD analysis. This should give you an idea as to how rough the X-Plane approximation is. Just because it solves a bunch of complicated looking equations does not make it completely accurate, although i can see how the uneducated might think that it is "superior".I recently got in touch with one of my classmates who now works for a Level D simulator company. Out of curiosity i asked to learn more about the simulation software in a Level D sim. I wanted to see if it is closer to X-Plane (geometric model, solving for forces on aircraft) or closer to FSX (real world flight test data in a table). Guess what, it resembled FSX, it did not even attempt to perform the conceptual force analysis based on the geometry of the aircraft model. A high fidelity flight seem does not need to solve for all forces in real time. Edited February 25, 2012 by 800xp

[LAdamson 240]

Can you please send me some recommended artificial stability settings? Real world pilot feedback is the most valuable tool one can use to evaluate a and tweak a low level (non-professional, not Level D) computer flight model.

I can't really recommend settings, that would be across the board. It's depended on the model, as well as a particular build of X-Plane. I also wouldn't know if it's built into the model or not. I've never cared to really look into plane maker.

[GoranM 3,764]

I recently got in touch with one of my classmates who now works for a Level D simulator company. Out of curiosity i asked to learn more about the simulation software in a Level D sim. I wanted to see if it is closer to X-Plane (geometric model, solving for forces on aircraft) or closer to FSX (real world flight test data in a table). Guess what, it resembled FSX, it did not even attempt to perform the conceptual force analysis based on the geometry of the aircraft model. A high fidelity flight seem does not need to solve for all forces in real time.

So as an engineer, you're saying that linear look up tables are more accurate than non-linear blade element theory and physics flight sims?And also, of course level d sims more closely resembles FSX. Level D Sims use look up tables. They always have. I could have told you that level d sims more closely resemble FSX WITHOUT even flying in one purely based on this fact. They use the same tech. Could you imagine reprogramming the software for these level d flight sims to use BET? Logistical nightmareDoes it mean they are more accurate? Well, that's up for debate. Personally, I prefer physics over look up tables any day of the week. As do quite a few aircraft manufacturers who depend on x plane for flight testing.

[800xp 0]

So as an engineer, you're saying that linear look up tables are more accurate than non-linear blade element theory and physics flight sims?And also, of course level d sims more closely resembles FSX. Level D Sims use look up tables. They always have. I could have told you that level d sims more closely resemble FSX WITHOUT even flying in one purely based on this fact. They use the same tech. Could you imagine reprogramming the software for these level d flight sims to use BET? Logistical nightmareDoes it mean they are more accurate? Well, that's up for debate. Personally, I prefer physics over look up tables any day of the week. As do quite a few aircraft manufacturers who depend on x plane for flight testing.

The point I am trying to get across is that X-Plane it is not real physics, but a rough approximation of real physics. It solves for a gross approximation of what the real aircraft is actually doing. And I am not familiar with the table based model but suspect that these tables are not as simple and linear as one might expect. Austin certainly describes the table based model as a simple linear table, but I can bet putting real life test data into a Level D sim is more involved than a simple liner table. What I am trying to say is that solving for the "physics" is not necessary, unless you are making a simulator for an aircraft that has never flown before. As for real aircraft manufacturers I can see them using this at the earliest, conceptual stages of aircraft design. It is a great conceptual design tool and solves for aircraft flight within the loose conceptual design tolerances. X-Plane is worth the 80 bucks just for that capability.But it may not be the most efficient solution when it comes to simulating flight on aircraft that already have real life counterparts that are flying. We will be able to solve for physics that resemble "real physics" when CFD becomes sophisticated enough to render wind tunnels obsolete and computing power increases to the point where we can run CFD in real time on the full aircraft. I estimate it will take about 100 years until that can happen given the current state of CFD and computing power.

I can't really recommend settings, that would be across the board. It's depended on the model, as well as a particular build of X-Plane. I also wouldn't know if it's built into the model or not. I've never cared to really look into plane maker.

I see, how about the stability settings for the X-Plane model that you fly in real life in the current build of X-Plane?

[LAdamson 240]

The point I am trying to get across is that X-Plane it is not real physics, but a rough approximation of real physics. It solves for a gross approximation of what the real aircraft is actually doing. And I am not familiar with the table based model but suspect that these tables are not as simple and linear as one might expect. Austin certainly describes the table based model as a simple linear table, but I can bet putting real life test data into a Level D sim is more involved than a simple liner table. What I am trying to say is that solving for the "physics" is not necessary, unless you are making a simulator for an aircraft that has never flown before.As for real aircraft manufacturers I can see them using this at the earliest, conceptual stages of aircraft design. It is a great conceptual design tool and solves for aircraft flight within the loose conceptual design tolerances. X-Plane is worth the 80 bucks just for that capability.But it may not be the most efficient solution when it comes to simulating flight on aircraft that already have real life counterparts that are flying.We will be able to solve for physics that resemble "real physics" when CFD becomes sophisticated enough to render wind tunnels obsolete and computing power increases to the point where we can run CFD in real time on the full aircraft. I estimate it will take about 100 years until that can happen given the current state of CFD and computing power.

You nailed that perfectly! And it's a fact, that the table lookup modeling, as used in MSFS is far from simple. I never got into programming for MSFS, but worked with different programmers to get results. It appears to be anything but easy, and the best modelers have years of experience. Same applies to X-Plane too. Turns out that the rough product, churned out by plane maker............takes a lot of know how to get it to hit the numbers, as well as fly accurately. Takes a lot more, than a few minutes during a coffee break... :(

I see, how about the stability settings for the X-Plane model that you fly in real life in the current build of X-Plane?

The X-Plane models of RVs that I fly in real life, are from old versions. They are terrible, and not worth the trouble of doing anything with.There is a much better 3rd party model for FSX.......by Baytower. The closest to the RV for X-Plane is the 3rd party Falco. I just leave the settings at full realism for that one.The Falco and RV are similar in size and engine horsepower...........as well as having sliding canopies. I did mess with settings for the default 172, and BTW--- Morton's (free) Archer seemed much like a real life landing at the 9000+ altitude at Telluride, Colorado this afternoon (XP-9). It was one of those "good" moments. Edited February 25, 2012 by LAdamson