User avatar
Buyantukhaa
Posts: 2288
Joined: Thu May 13, 2004 5:33 am

How To Make Sim Flight Characteristics Realistic?

Sat May 26, 2007 6:32 pm

I was just reading the "How do your students try to kill you" thread and one question came up: one can program a flight simulator to display just about any behaviour possible, but how do you know that the programmed behaviour actually reflects the real behaviour? Especially with aircraft types that are new to the market. In that student thread someone said that because of flaring, the CRJ refused to land and touched down really late. So somebody must have flown a CRJ in the beginning and observed this before it could be programmed in the sim, or not? And then the question is: how do you know you really got it right? Especially with extreme situations (simulated damage to the aircraft etc.), how do you know that sim flight characteristics are a correct representation of the real thing?
I scratch my head, therefore I am.
 
Arniepie
Posts: 1428
Joined: Thu Aug 04, 2005 11:00 pm

RE: How To Make Sim Flight Characteristics Realistic?

Sat May 26, 2007 8:27 pm

I think if you want to use MSFS, the most popular simulator on the market, you can forget about having real life characteristics on most of the models because it needs
-A real life pilot to advice and test
-An expert in tweaking the cfg and air ed files
-better weather effects and dynamics
-very high end force feedback rudder and yoke
The way MSFS works makes it way too hard to get all these things just right and real life pilots tell me that it is good for learning your instruments but not so for creating the feel for flying.
Strangely many things in MSFS are harder than real life (if you only use VC or 2D) certainly when it comes to situational awareness , depth view, feel for pitch/roll/yaw, maintaining level flight while hand-flying, landing and flying in bad weather.
[edit post]
 
David L
Posts: 8547
Joined: Tue May 18, 1999 2:26 am

RE: How To Make Sim Flight Characteristics Realistic?

Sat May 26, 2007 9:29 pm

Just a guess but I think BuyantUkhaa is referring to the commercial sims used by the airlines.
 
Arniepie
Posts: 1428
Joined: Thu Aug 04, 2005 11:00 pm

RE: How To Make Sim Flight Characteristics Realistic?

Sat May 26, 2007 9:54 pm

Sorry about that, have been playing way too long with MSFS and therefore I just assumed he was referring to the game.

As for the big boys sim I guess it really is "as real as it gets", most pilots would feel the difference immediately when they go from sim to real plane the first time, I assume.
[edit post]
 
David L
Posts: 8547
Joined: Tue May 18, 1999 2:26 am

RE: How To Make Sim Flight Characteristics Realistic?

Sat May 26, 2007 10:06 pm

Quoting ArniePie (Reply 3):
As for the big boys sim I guess it really is "as real as it gets", most pilots would feel the difference immediately when they go from sim to real plane the first time, I assume.

Yes, but the question is about how many of the sim's characteristics are known to be realistic and how many are assumed to be realistic from theory - something I've often wondered about but was too lazy to ask.  Smile
 
3DPlanes
Posts: 167
Joined: Thu Apr 27, 2006 11:12 pm

RE: How To Make Sim Flight Characteristics Realistic?

Sat May 26, 2007 11:42 pm

It depends on how things are simulated...

In something like MSFS, it is driven by what are known as "look up tables" - basically meaning that for a given set of parameters it goes into a table and finds the result. So at 120 kts, with full power and 30 nose up in a 172, the speed will bleed of at X rate. This is similar to how the lower end PC-ATDs and FTDs work.

At the higher end, the simulation is done by actually computing what the airflow is doing over various parts of the airframe. This is much better than using LUTs, but still not 100% accurate since they only sample at discreet intervals - say every 2 ft or so along the wing. More samples means more computing power - more than you could put in a sim. On a full-size sim, this performance can be compared to the flight test data from the manufacturer.

On the PC market, X-plane is supposed to use this more advanced approach, and there are discussions of people using it to test fly designs to see how they perform and how various changes effect that performance.
"Simplicate and add lightness." - Ed Heinemann
 
David L
Posts: 8547
Joined: Tue May 18, 1999 2:26 am

RE: How To Make Sim Flight Characteristics Realistic?

Sun May 27, 2007 12:07 am

Quoting 3DPlanes (Reply 5):
On a full-size sim, this performance can be compared to the flight test data from the manufacturer.

Yes but, as asked in the OP, how much actually is tuned to match flight test data, how much remains theoretical and, where appropriate, how accurate is the theory estimated to be?

I don't think the OP was asking about PC sims.  Smile
 
User avatar
Jetlagged
Posts: 2562
Joined: Sun Jan 23, 2005 3:00 pm

RE: How To Make Sim Flight Characteristics Realistic?

Sun May 27, 2007 9:28 am

Quoting 3DPlanes (Reply 5):
At the higher end, the simulation is done by actually computing what the airflow is doing over various parts of the airframe. This is much better than using LUTs, but still not 100% accurate since they only sample at discreet intervals - say every 2 ft or so along the wing. More samples means more computing power - more than you could put in a sim. On a full-size sim, this performance can be compared to the flight test data from the manufacturer.

I design commercial training sims for a living and can assure you that computing airflow over the surfaces is not part of the deal. That may just happen in high-end research sims. Such computations would require extremely powerful computers to complete in real time. The aerodynamic model consists of many functions which compute the various coefficients and derivatives according to values of AOA, Mach, beta, flap angle, etc. The model builds up the coefficients, converts them to forces and moments then integrates the effects to calculate velocities and rotations. New values of input variables result and the process repeats, 30 or 60 times a second. It's a maths model, rather than an aerodynamic calculation.

Surprising as it may seem, such an aero model has much more in common with MSFS than it has with X-Plane for example.

For aircraft which have not yet been flown what happens is this:

The aircraft manufacturer produces a predictive aero maths model, based on it's research, expected performance and non-real time simulations using CFD methods. This is how the manufacturer expects the aerodynamic coefficients and stability derivatives to be.

The sim designer programs this model for use in training until flight test data becomes available.

When the aircraft flies, it is flight tested, and the aircraft manufacturer adjusts the predictive model to match real world data. Predictive models used to be poor, but these days they are pretty accurate. There are 787 and A380 sims being built right now using this kind of data.

There are also mathematical approximations to predict basic coefficients and derivatives simply from aircraft dimensions and weights. These techniques can be used even for MSFS models. It's like reverse engineering the model.

For any commercial training sim, it must be "validated" against flight test data (or predictive data) to within prescribed tolerances. All tests specified must pass to the satisfaction of the FAA, etc. When this is done, and the FAA pilots are happy, the sim is qualified. A sim qualified to the highest standard (Level D) will be extremely accurate, but even a Level B sim will be very nearly as close. It will have the same aero model, the difference maybe only a lower standard of visual, motion or sound simulation.

Lower end FTD's use a similar aero maths model to full flight sims, however manufacturers data is expensive, so approximate performance may be obtained by the tuning of a generic airliner aero model to observed data.

Whatever it is, it is not magic. The "gurus" who produce flight dynamics for MSFS like you to think some mystical process is going on, but if you can understand basic flight performance calculations there is no great trick to it. You can calculate the basic numbers to first order approximations on the proverbial back of an envelope.

Hopefully I won't now be expelled from the "Magic Circle" for revealing this!
The glass isn't half empty, or half full, it's twice as big as it needs to be.
 
N231YE
Posts: 2620
Joined: Tue Jul 25, 2006 9:24 am

RE: How To Make Sim Flight Characteristics Realistic?

Sun May 27, 2007 9:44 am

Quoting ArniePie (Reply 3):
As for the big boys sim I guess it really is "as real as it gets", most pilots would feel the difference immediately when they go from sim to real plane the first time, I assume.

Somewhat. However, there are flight schools out there that do not own a particular aircraft, but the simulators they possess are so realistic that a pilot can be fully certified to fly that particular aircraft, even if they have never put their hands on the yoke/joystick of the "real" aircraft.

I believe Flight Safety International is an example of one of these schools.
 
User avatar
Jetlagged
Posts: 2562
Joined: Sun Jan 23, 2005 3:00 pm

RE: How To Make Sim Flight Characteristics Realistic?

Sun May 27, 2007 10:44 am

Quoting David L (Reply 6):
Yes but, as asked in the OP, how much actually is tuned to match flight test data, how much remains theoretical and, where appropriate, how accurate is the theory estimated to be?

It's all matched to flight test data in a full flight sim. Engineering sim data is usable, to tighter tolerances. No theoretical stuff allowed.  Smile

Tuning is not really the right word. These days, simulator data packages from the major manufacturers are usually so good that tuning is rarely if ever necessary.
The glass isn't half empty, or half full, it's twice as big as it needs to be.
 
Fly2HMO
Posts: 7207
Joined: Sat Jan 24, 2004 12:14 pm

RE: How To Make Sim Flight Characteristics Realistic?

Sun May 27, 2007 1:17 pm

Quoting Jetlagged (Reply 7):
the process repeats, 30 or 60 times a second.

Is the number of times this calculation is done vary from say, a level D to a level 6? What about MSFS to X-plane?
 
David L
Posts: 8547
Joined: Tue May 18, 1999 2:26 am

RE: How To Make Sim Flight Characteristics Realistic?

Sun May 27, 2007 9:51 pm

Quoting Jetlagged (Reply 7):

Excellent. Your name was the first to spring to mind when I read the question.  Smile

Quoting Jetlagged (Reply 9):
It's all matched to flight test data in a full flight sim. Engineering sim data is usable, to tighter tolerances. No theoretical stuff allowed.

I didn't think there'd be much but I just wondered if there's anything done in the simulator that doesn't have full test flight data, e.g. because it's too dangerous to test.
 
saab2000
Posts: 1216
Joined: Thu Jun 21, 2001 6:19 pm

RE: How To Make Sim Flight Characteristics Realistic?

Sun May 27, 2007 10:45 pm

I have spent considerable time flying two level "D" simulators. The SAAB2000 and the CRJ-200. Neither are especially realistic to a seasoned pilot, but both provide enough realism for the job at hand - training procedures and profiles. And both will bite you if you don't fly them correctly. Same as the airplane. FWIW, the CRJ sim is more like the real airplane than the SAAB2000 simulator. For a non-pilot I suppose they would both be pretty realistic.

The landings in both are not very realistic. Just be on speed and aim for the centerline and touchdown zone. Speed control is the cause of most landing problems AFAIK.

But the biggest difference between sim training is that all the weather all the time is IMC, down to or below mins. Lots of time just staring at the instruments. Also, the CRJ is very sensitive to turbulence in real life. The simulator does not simulate this very well. A real engine failure in real turbulence and bad weather would likely be pretty challenging. But absolutely controllable in both the sim and the real airplane.

The sims are a safe and efficient way to provide an experience that is about 95% realistic. But landings are not really realistic. That is why some sims are certified for type-ratings, but the pilots are still required to do some landings in the actual airplane.

I would guess that simulators will become ever more realistic as computer processors increase in speed, providing better graphics and more realistic control response.
smrtrthnu
 
User avatar
Buyantukhaa
Posts: 2288
Joined: Thu May 13, 2004 5:33 am

RE: How To Make Sim Flight Characteristics Realistic?

Mon May 28, 2007 6:31 am

Quoting ArniePie (Reply 3):
Sorry about that, have been playing way too long with MSFS and therefore I just assumed he was referring to the game.

No problem  Smile

Quoting Jetlagged (Reply 7):
Hopefully I won't now be expelled from the "Magic Circle" for revealing this!

Thanks for the explanation! Always nice to see that for virtually question, there is somebody here that knows the answer.

Makes sense really, come up with a prediction (the aero model) and then calibrate it against observed behaviour. I have done some modelling myself (although definitely not aviation-related!) and the (most basic) principles we've used were similar.

Quoting Jetlagged (Reply 7):
All tests specified must pass to the satisfaction of the FAA, etc. When this is done, and the FAA pilots are happy, the sim is qualified.

How would that be done? Make a real flight, read out the DFDR, fly the "same" flight on a sim and check parameters? Or using a pilot who is very familiar with the real plane and let him/her fly the sim?
I scratch my head, therefore I am.
 
User avatar
Starlionblue
Posts: 17058
Joined: Fri Feb 27, 2004 9:54 pm

RE: How To Make Sim Flight Characteristics Realist

Mon May 28, 2007 10:46 am

Quoting Saab2000 (Reply 12):
I would guess that simulators will become ever more realistic as computer processors increase in speed, providing better graphics and more realistic control response.

IIRC the CAA 767 simulator I "flew" has a couple of racks filled with computers just for graphics, and the graphics were rather, ahem, crummy compared to even some previous MSFS. My single Radeon x1950 card could easily chug all that out. But that's what 15-20 years of development means. The F28 simulators in the same building presumably didn't show much more than wireframe graphics, and didn't move at all. But they could still be used for training.

However, the field of view and all the instruments and controls were the same as in the real plane, and that really made it a good training aid as opposed to a PC with Yoke, Throttles and Pedals.

As Saab2000 says, in the future simulations will become better because the computation required to generate it will become cheaper and cheaper.

BTW where is SK973? He should weigh in on this one.

[Edited 2007-05-28 03:48:10]
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
Pihero
Posts: 4196
Joined: Mon Jan 31, 2005 5:11 am

RE: How To Make Sim Flight Characteristics Realistic?

Wed May 30, 2007 5:04 am

Quoting Jetlagged (Reply 7):
aircraft manufacturer produces a predictive aero maths model, based on it's research, expected performance and non-real time simulations using CFD methods. This is how the manufacturer expects the aerodynamic coefficients and stability derivatives to be.

I'm glad you're in this, Jetlagged, as I have thought of the same question, mine being about the influence -or not- of some aircraft mass and inertia (how is it computed) and whether you consider some sort of interaction / vibration period / structural resonnance...between the plane and it's flight control surfaces.
I was told by a friend some time ago that the main problem (ok, the"first" main problem !) the 380 engineers encountered was that the plane's own osciullation was just aboutr 1 cycle/s, which induced nausea, increased fatigue...and so forth.( By the way, that's how he explained why it was difficult to reduce the weight as so much tightening was required ).
Apparently, they found all that on the sim, months before the first flight.
Contrail designer
 
User avatar
Starlionblue
Posts: 17058
Joined: Fri Feb 27, 2004 9:54 pm

RE: How To Make Sim Flight Characteristics Realistic?

Wed May 30, 2007 5:39 am

Quoting Pihero (Reply 15):
I was told by a friend some time ago that the main problem (ok, the"first" main problem !) the 380 engineers encountered was that the plane's own osciullation was just aboutr 1 cycle/s, which induced nausea, increased fatigue...and so forth.( By the way, that's how he explained why it was difficult to reduce the weight as so much tightening was required ).

Wow. IIRC the 345 and 346 both actuate the elevators automatically to dampen oscillations.
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
FredT
Posts: 2166
Joined: Thu Feb 07, 2002 9:51 pm

RE: How To Make Sim Flight Characteristics Realistic?

Wed May 30, 2007 6:27 am

Quoting Jetlagged (Reply 7):
The aircraft manufacturer produces a predictive aero maths model, based on it's research, expected performance and non-real time simulations using CFD methods. This is how the manufacturer expects the aerodynamic coefficients and stability derivatives to be.

Very good post, Jetlagged. I'd just like to add that tens of kgs of wind tunnel data also goes into the predicted flight model of a new design. Flying qualities is an interesting field, but the processing of wind tunnel data is a good reason to stay away from it...  Wink

Quoting Pihero (Reply 15):
I'm glad you're in this, Jetlagged, as I have thought of the same question, mine being about the influence -or not- of some aircraft mass and inertia (how is it computed)

Mass, center of gravity and the moments/products of inertia are all calculated early on in the design process and then continuously refined as the design phase proceeds. By the time you are ready to get a sim running, the mass properties model will be a good representation of the actual aircraft.

Quoting Pihero (Reply 15):
and whether you consider some sort of interaction / vibration period / structural resonnance...between the plane and it's flight control surfaces.

Are you talking about flutter? If so, it is an abnormal condition and will be simulated if it is required to do simulator training to deal with flutter. The simulators will do the common failure modes the pilots can be expected to encounter and which they should be able to cope with. If control surface flutter is one of those modes, well... hopefully it isn't but you never know.  Smile

If you are talking about dutch roll, dynamic instability and the like then yes, that will occur in the simulator just as in real life and for the same reasons.
I thought I was doing good trying to avoid those airport hotels... and look at me now.
 
Pihero
Posts: 4196
Joined: Mon Jan 31, 2005 5:11 am

RE: How To Make Sim Flight Characteristics Realistic?

Wed May 30, 2007 4:18 pm

Quoting FredT (Reply 17):
Are you talking about flutter?

No, but you've answered another question I wanted to ask.

Quoting FredT (Reply 17):
If you are talking about dutch roll, dynamic instability and the like then yes, that will occur in the simulator just as in real life and for the same reasons.

Thanks.
I've been very impressed by the accuracy of modern simulators and I've used them a lot in my career to improve my flying ( for instance, work out engine settings vs aircraft mass and configurations... gives a very smooth / precise handling...).

The only aspect that sometimes annoyed me was the sync between sim movement and the "visual". There was one among the dozens we have which took months to caliber...and certify with DGCA.

Once again, thanks for the infos.
Contrail designer
 
User avatar
moriarty
Posts: 172
Joined: Tue Jan 10, 2006 1:05 am

RE: How To Make Sim Flight Characteristics Realistic?

Wed May 30, 2007 7:48 pm

Hello,

This might not be an actual answer to the question but I cannot resist to write something anyway!

I am part of a project (my part is very small but never the less) that builds an AJ-37 Viggen simulator. This includes both cockpit and software. We have the ambition to make a 100% accurate simulation, from the cockpit to the flight characteristics.

I won't go into detail about the cockpit but we have the interior of an actual aircraft so it's pretty safe to say it's accurate. Software wise then... We use Flight Simulator X (migrating from 2004) as a base for the software. Why invent the wheel once again...

Now, the delicate part is to try to make the Viggen inside MFS to behave like the real thing. First of all, the project owner/leader (that also builds the cockpit and all associated hardware) is a pilot and flight instructor. This enables him to have a quite good guess about the basic handling of the aircraft.

The magic part however, is when all known (and unknown) real life parameters are entered into the configuration files of MFS. The guy that does this has done a tremendous work, reading documentation and asking people and in the end calculated and entered numbers into MFS. Part from documentation, we do have quite some very useful connections with people actually working (well, worked) with Viggen. Of course this is crucial in order to save effort and increase accuracy.

To this point everything can be considered to be qualified guesses though. What we do in the end to verify what we have implemented, is what most people do (or wanna do I guess): get the people that work with the real thing to fly the thing as much as possible, trying out every aspect of the simulation. In other words, we need a test pilot with real life experience.

And luckily, we've had several Viggen pilots flying our simulator and their feedback has been invaluable. They do not only provide us with feedback on specific details of the simulation, we also get a "general" impression from them. As one of them said after his first ride: "The first impression of the Viggen was that it really smells like a Viggen!". We then know we're pretty close. And when some other Viggen pilot says "This is as close you can get" we're really happy!

How do we perform a test flight? We're not professional aircraft manufacturers but we try to hang around the simulator, noting (mentally often) every thing the pilot observes. It can be anything from "that gauge should not accelerate in that way when I do this" to "the aircraft feels a little bit fast when applying air brakes in this speed range".

Iterating this process as much as possible hopefully (we think so anyway) has taken us so far so that it comes down to opinions rather than facts. Of course, there are so many things to simulate that we cannot get everything perfect or even near perfect. But the ambition is there anyway. And we're not done by any means (probably never will be) - there's always more functionality around the corner to implement!

The ironic part of it is, every now and then we do something that forces us to go back a few steps and do it all over again. For instance, going from 2004 to FSX forces us to revise all flight data and aircraft behavior since the latest iteration of MFS behaves a little different than the previous one.

In the end: how do we know what *is* correct and what *seems* correct? We don't make a difference. Since this is for amusement rides only (not education) we don't need to verify all the way. A real life pilots words that it feels accurate is enough for us.

To make a real simulator accurate is another story I guess. But the basic idea is the same, even though there's probably more formal methods of doing it.

cheers

Moriarty
Proud to part of www.novelair.com.
 
User avatar
Starlionblue
Posts: 17058
Joined: Fri Feb 27, 2004 9:54 pm

RE: How To Make Sim Flight Characteristics Realistic?

Wed May 30, 2007 10:32 pm

Interesting post Moriarty. After the initial setup, what you seem to be doing is iterative reverse engineering from empirical data.
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
User avatar
moriarty
Posts: 172
Joined: Tue Jan 10, 2006 1:05 am

RE: How To Make Sim Flight Characteristics Realistic?

Wed May 30, 2007 10:57 pm

Pretty much yes. But we do have some nifty declassified documentation. But then again, we build a simulator, not an airplane. So in the end we need to adjust all information to a simulation. And real life aircraft has some more "parameters" than a simulation.

But it's amazing how far you can get using a (relative professional simulators) cheap simulator like MFS.
Proud to part of www.novelair.com.
 
User avatar
Starlionblue
Posts: 17058
Joined: Fri Feb 27, 2004 9:54 pm

RE: How To Make Sim Flight Characteristics Realistic?

Wed May 30, 2007 11:23 pm

Quoting Moriarty (Reply 21):
But it's amazing how far you can get using a (relative professional simulators) cheap simulator like MFS.

As with a lot of gaming software nowadays, the configuration/programming framework is very important. MSFS is highly customizable and has an open, published interface. World of Warcraft and Half-Life come to mind. So your constraints move from the software itself to the hardware it runs on. "How many computations can this tin actually perform?" And even that constraint is dissapearing as computing power in PCs recedes at an ever increasing pace.
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
User avatar
Jetlagged
Posts: 2562
Joined: Sun Jan 23, 2005 3:00 pm

RE: How To Make Sim Flight Characteristics Realistic?

Thu May 31, 2007 9:48 am

Quoting FLY2HMO (Reply 10):
Is the number of times this calculation is done vary from say, a level D to a level 6? What about MSFS to X-plane?

30 Hz is the minimum these days in practice, and with modern computing it is easily achieved. There really isn't any point in running slower. Level 6 FTD's have the possibility to be upgraded to FFSs so the computing power has to be sufficient. MSFS seems to run at a variable rate, as fast as the PC can cope with unless you cap the maximum FPS (strongly recommended). I don't have any knowledge of how X-Plane frames are divided up. Generally it is better to have slower, evenly spaced frames rather than a higher frame rate with irregular timing.

Computers relating to things like control loading and motion run much faster for servo stability (at least 1KHz for control loading, less for motion).

Quoting David L (Reply 11):
I didn't think there'd be much but I just wondered if there's anything done in the simulator that doesn't have full test flight data, e.g. because it's too dangerous to test.

The only one that springs to mind is the Dynamic Engine Failure After Takeoff test. Often done on the engineering simulator, not the real aircraft, for some reason.  Smile If you get flight test data, you'll usually see it was flown at a safe height, rather than just after takeoff.

The VMCG test must have a high "pucker factor" too. You sometimes see the corrective rudder going in as the engine is cut, rather in response to the yaw, but I can't say I blame the pilots! VMCG is not flown for every airframe/engine combination, so often the test has to be run by matching the thrust of the engines to those of the test aircraft.

Quoting BuyantUkhaa (Reply 13):
How would that be done? Make a real flight, read out the DFDR, fly the "same" flight on a sim and check parameters? Or using a pilot who is very familiar with the real plane and let him/her fly the sim?

The flight test data comes from specially instrumented tests flights done during aircraft certification. The simulator comparison is done by flying the test in the simulator and comparing the result with the original flight test. The tests are "flown" automatically, reproducing test pilot inpiuts. Most simulators "overplot" the flight test data on the sim response so a visual comparison is easy. Most modern sims use automatic tolerance checking too, so it's a simple pass/fail indication. All the tests must be run every year to maintain qualification.

For simulators where flight test data is not available (it's very expensive to get if not provided by the manufacturer), a "pilot familiar with the real plane" is commonly called upon! In fact, subject assessment is a major part of all simulator initial qualifications.
The glass isn't half empty, or half full, it's twice as big as it needs to be.
 
User avatar
jetmech
Posts: 2316
Joined: Wed Mar 29, 2006 2:14 am

RE: How To Make Sim Flight Characteristics Realistic?

Thu May 31, 2007 10:43 am

Very informative post Jetlagged, thanks for that  bigthumbsup  !

Quoting Jetlagged (Reply 7):
expected performance and non-real time simulations using CFD methods......The sim designer programs this model for use in training until flight test data becomes available.

Urghh.... CFD and programming. I am currently struggling through an assignment where we have to write a CFD program in MATLAB. The programming doesn't seem to be too hard. Understanding the maths behind it is much worse  cry .

Regards, JetMech
JetMech split the back of his pants. He can feel the wind in his hair :shock: .
 
User avatar
Jetlagged
Posts: 2562
Joined: Sun Jan 23, 2005 3:00 pm

RE: How To Make Sim Flight Characteristics Realistic?

Thu May 31, 2007 11:44 am

Quoting Pihero (Reply 15):
I have thought of the same question, mine being about the influence -or not- of some aircraft mass and inertia (how is it computed) and whether you consider some sort of interaction / vibration period / structural resonnance...between the plane and it's flight control surfaces.

As FredT said, mass properties are well defined by data, early in the aircraft design stage. Each airline may have it's own unique mass properties data, depending on its configuration. This is mostly overkill from a sim design point of view as it's unnecessary to consider passenger distribution. All the instructor needs is to do is set up the weight, then adjust CG as required, which is done by moving the CG of the payload so subsequent CG changes due to fuel usage are always reflected correctly. Fuel slosh effects on CG must also included.

As far as inertia modelling goes, the empty aircraft inertia is defined by data. Other inertia terms vary with mass distribution, again as defined in the data.

Aeroelasticity is modelled, but really only in a static way, often not accounting for g loads for example. Structural modes, if significant, could be built in if data was made available.

You'll have gathered by now that a huge amount of data is provided by the manufacturer these days, leaving little room for subjective interpretation and "tuning".
The glass isn't half empty, or half full, it's twice as big as it needs to be.
 
cptspeaking
Posts: 567
Joined: Mon Apr 24, 2006 2:49 pm

RE: How To Make Sim Flight Characteristics Realistic?

Thu May 31, 2007 12:34 pm

Quoting David L (Reply 11):
but I just wondered if there's anything done in the simulator that doesn't have full test flight data, e.g. because it's too dangerous to test.

A professor of mine is a former check pilot for a large regional airline that flies the CRJ series (mostly). They teach during training that the CRJ is NOT recoverable from a stall, and they are to be avoided like the plague. In fact, the CRJ is sometimes recoverable from a stall (i.e. that Pinnacle disaster flight...it wasn't the stall that killed them...), but is never recoverable from a spin or deep stall (yes, there is a difference, and it isn't how big the crater that you make is).

Why is this relevant? I'm told that the simulators for the CRJ are never recoverable from a stall. Bombardier does not publish the stalling characteristics of the airplane for insurance reasons, so the sims cannot accurately model a stall and therefore they don't at all. Basically, Bombardier does not want to be held responsible in any way for some bozo going out trying to stall the CRJ just because he did it in the sim and it turned out alright. So, in order to prevent this, the data is not used and the training is that you can't recover it. This doesn't mean that if it is stalled accidentally that the pilots won't have the proper knowledge to be able to attempt a recovery, but they don't practice it on the sim.

It's not that this was too dangerous to test...they did it with a drogue chute in flight testing, but as explained above, the manufacturer doesn't publish the results and therefore it isn't duplicated in the sim.

Your CptSpeaking
...and don't call me Shirley!!
 
User avatar
Jetlagged
Posts: 2562
Joined: Sun Jan 23, 2005 3:00 pm

RE: How To Make Sim Flight Characteristics Realistic?

Thu May 31, 2007 12:45 pm

Quoting CptSpeaking (Reply 26):
A professor of mine is a former check pilot for a large regional airline that flies the CRJ series (mostly). They teach during training that the CRJ is NOT recoverable from a stall, and they are to be avoided like the plague. In fact, the CRJ is sometimes recoverable from a stall (i.e. that Pinnacle disaster flight...it wasn't the stall that killed them...), but is never recoverable from a spin or deep stall (yes, there is a difference, and it isn't how big the crater that you make is).

Sounds a bit like a war story to me. Deep stall is irrecoverable on any type, other stalls can be recovered, but obviously are not encouraged in training.

It is rare to train AOA beyond stick shake, certainly in North America, so airliner manufacturers do not provide much stall data. Nobody spins airliners (most airliner aero models do not even simulate it).
The glass isn't half empty, or half full, it's twice as big as it needs to be.

Who is online

Users browsing this forum: No registered users and 10 guests