Sign in to follow this  
Avcomware

Comments on aircraft.cfg

Recommended Posts

What is important in some of the aircraft.cfg lines and what has no effect unless FSEdit is used to FUBAR the AIR file. One can save a backup of the air file, then write it over any FSEdit messed up air file if FSEdit is used to edit aircraft.cfg. I'll only cover some sections of aircraft.cfg. The offsets below are based on Reference Datum offset = 0, 0, 0. Otherwise, the distances along the fuselage change.[airplane_geometry]wing_area=1700.0 80 80 << Set near real turbine inlet area. Higher values increase ram drag, less net thrust at high speeds.rated_N2_rpm=12250 <

Share this post


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

Thanks, Ron!I think I know why I wasn't seeing the Cdm on the 747. I believe if any of the scalar's are not set to one Herve ASDF doesn't display the data. I'm using a downloaded 747 air file that has quite a few mods to the para's and scalars.I tried ASDF on an aircraft with the scalars set to one and all the data appeared.

Share this post


Link to post
Share on other sites

I will ask my fool questions now.> What is important in some of the aircraft.cfg lines and >what has no effect unless FSEdit is used to FUBAR the AIR >file.We know that certain settings in aircraft.cfg are used by FSEdit to create the air file. Some of these settings are input to the "engine" to generate values stored in the air file. But not all, only those pertaining to flight dynamics (we are told).But are some of the aircraft.cfg settings used in "real time" or during air file initialization at run time to drive aircraft performance or air file values? >> One can save a backup of the air file, then write it over >any FSEdit messed up air file if FSEdit is used to edit >aircraft.cfg. You are suggesting that we can use FSEdit to edit values in aircraft.cfg, using the nice tab interface, then let it overwrite the air file, but because it is botched, we replace the air file with the old one.This would mean using FSEdit only as a convenient way to edit those values that are not processed into the air file, correct?>> I'll only cover some sections of aircraft.cfg. >> The offsets below are based on Reference Datum offset = 0, >0, 0. Otherwise, the distances along the fuselage change. >>[airplane_geometry] >wing_area=1700.0 << Most important >wing_span=108.0 << Also Important >wing_root_chord=15.7 << SET to MAC=area/span, NOT root >length What does MAC stand for? Mean Chord? And do you mean the area or the span (distance)?I've set mine to the actual root chord, the length of the wing section parallel to the fueslage where it meets the fuselage as measured on the visual model. I do not have the actual mean chord for this section having to take it from a drawing as the only source---well I may be able to look up the specs for this aircraft and find it.>wing_dihedral=3.0 << No effect This does not have an effect on the air file?>wing_incidence=0.70 << This, and twist, affect cruise >attitude This is in degrees, correct?>wing_twist=-3.0 Also called washout I think.> However, there is a parameter in the AIR file that should >be changed when either of the above are changed. Better to >change 'cruise_lift_scalar' (below) to adjust cruise pitch. Hmm, I set my wing incidence to the angle of the a/c I was modelling, but I did not set anything in the air file. I'd still like to set the correct incidence given that is one number I do have.>>oswald_efficiency_factor=0.80 <to 0.85 >wing_winglets_flag=0 <<1=T Reduces Induced Drag - >possibly too much I think these increase thrust counterbalancing induced drag. But I could be wrong. Off topic anyway.>wing_sweep=0.0 <wing_pos_apex_lon=3.93 <wing_pos_apex_vert=0.0 <htail_span=35.75 << No Effect >htail_pos_lon=-65.9 << No Effect >htail_pos_vert=19.29 << No Effect >>htail_incidence=-2.0 <<0.0 to -3.0 may improve pitch >stability This is the angle the horizontal stabilizer makes with a line perpendicular to the fuselage, correct?>htail_sweep=0.0 << No Effect >vtail_area=356.0 << No Effect >vtail_span=21.5 << No Effect >vtail_sweep=0.0 << No Effect >vtail_pos_lon=-46.2 << No Effect >vtail_pos_vert=15.1 << No Effect >elevator_area=108.0 << No Effect >aileron_area=67.4 << No Effect >rudder_area=66.2 << No Effect >elevator_up_limit=27.0 <elevator_down_limit=17.0 <aileron_up_limit=17.0 <aileron_down_limit=17.0 <rudder_limit=25.0 <elevator_trim_limit=13.0 <spoiler_limit=60.0 <spoilerons_available=1 << 1=T, 0=F >aileron_to_spoileron_gain=2.6 <deflection to aileron deflection >min_ailerons_for_spoilerons=8.5 <min_flaps_for_spoilerons=0.0 <spoilerons >>[Reference Speeds] >flaps_up_stall_speed=165 <full_flaps_stall_speed=132 <cruise_speed=434.8 > I think this should be in IAS cruise speed if used as an AI >aircraft. So it doesn't fly too fast. That would be about >260 kts for jets. I am trying to discover what defines the overspeed for an a/c. Is it this cruise speed. You're saying cruise speed is in true knots airspeed not indicated?>>[flight_tuning] Most a/c seem to leave these all at 1.0.>cruise_lift_scalar = 1.0 <pitch >parasite_drag_scalar = 1.0 <0.7 might be appropriate. Set for appropriate PPH or TAS >but see "TurbineEngine" >induced_drag_scalar = 1.0 <<1.0 should be good if >'Oswald' right >elevator_effectiveness = 1.0 <desired >aileron_effectiveness = 1.0 >rudder_effectiveness = 1.0 >pitch_stability = 1.5 <roll_stability = 1.0 >yaw_stability = 1.3 <in Yaw >elevator_trim_effectiveness = 1.0 <almost stalls at full UP trim >aileron_trim_effectiveness = 1.0 <better >rudder_trim_effectiveness = 1.0 <better >>>[TurbineEngineData] >fuel_flow_gain=0.004 <rate inlet_area=9.80 << Set near real turbine inlet area. >Higher values increase ram drag, less net thrust at high >speeds. >rated_N2_rpm=12250 <static_thrust=17000 <Often this should be lower than TO rating, full throttle >usually adds a lot to this number. >afterburner_available=0 >reverser_available=1 >>>[GeneralEngineData] >engine_type= 1 >Engine.0=-49.7,-9.3, 2.00 //set up for thrust vector >Engine.1=-65.1, 0.0, 3.00 //approx center of thrust >Engine.2=-49.7, 9.3, 2.00 //1.00 >> These locations are for the "center of thrust", not the >engine! Prop location or about 2/3 of the way back along a >turbine. Up/Down will add a pitching moment with power >changes. Okay, that is closer to what I finally discovered, that this is another 'virtual' value representing a force, not the engine location. But I had little idea where to put it besides equal to the prop location.>>fuel_flow_scalar=1.60 >> This should always be above 1.0 for turbines. >SFC=fuel_flow_scalar/2. Typical SFC cruise values range >from 0.55 0.90. So, fuel_flow_scalar would be 1.1 to 1.8. >What is SFC?I really had no idea what to set this one at. I understand it is some non-unit reprsentation of fuel flow. I just had to play with it until the fuel flow looked right.Steve

Share this post


Link to post
Share on other sites

>Thanks, Ron! >>I think I know why I wasn't seeing the Cdm on the 747. I >believe if any of the scalar's are not set to one Herve ASDF >doesn't display the data. I'm using a downloaded 747 air >file that has quite a few mods to the para's and scalars. >>I tried ASDF on an aircraft with the scalars set to one and >all the data appeared. Ah, yes. I'd forgotten about that since I always set the drag_scalar to 1.0. Herve' will account for those scalars in the future. For now, simply set them to 1.0 and read the values in AFSD. Then, reset to the original values and multiply Cdo (or, probably Cdp = Cdo+Cdm+Cdg) by the 'drag_scalar'. It should be similar for other scalars that affect AFSD. Many don't. Also, the [flaps..] section(s) will be accounted for in the future. I explained to Herve' how I think they work. There is still one detail to work out concerning how flaps add both parasitic and induced drag. One of my in-progress jets shows Cdm=0.0005 at Mach 0.80, and it increases to 0.0025 at Mach 0.90. Note this is much lower than in 'other jets'. While the MS 747-400 shows a negative 'Mach Drag': -0.0117 at Mach 0.80. Cdo = '31' = 0.0151 in the AIR file, so Cdo + Cdm = 0.0034 at Mach 0.80. Almost low enough for perpetual motion. ;) Since the UL of AFSD 1.30 to the AVSIM Library Herve has added "Ice Weight" to the W&B section. The effects of Ice are also apparent in 'L/D', 'CD', 'Cdi' and 'Prop Efficiency'. My C177RG picked up only 20 lb of ice, but AoA increased a lot and so did Induced Drag when I flew through an icing layer. It's not the weight, it's the effect of Ice on the wing's Lift Slope that gets you. Also, the prop efficiency dropped from 81% to 62%, which doesn't help either. Ron

Share this post


Link to post
Share on other sites

>We know that certain settings in aircraft.cfg are used by >FSEdit to create the air file. Some of these settings are >input to the "engine" to generate values stored in the air >file. But not all, only those pertaining to flight dynamics >(we are told). >>But are some of the aircraft.cfg settings used in "real >time" or during air file initialization at run time to drive >aircraft performance or air file values? The lines I marked 'Not used' are used by FSEdit to set new records of parameters at the end of an AIR file. These replace most of the parameters in REC 1101 (if they exist). Also, new versions of TBL 403 and 473 are created by FSEdit. these are especially messed up. So are the other lines which are also used to set some of the flight model parameters when an AC is loaded. That is, 'Wing Area' replaces the 'wing area' set in FS2K and earlier AIR files, and it is also used, in combination with wing span and 'chord' to adjust the new parameters. I base my comments on an understanding of the FS flight model, and in general, how important wing parameters affect most every other 'final coefficient' scale in the run time flight model. Also, on having set up a few dozen AC for FS2K2. Which still leaves a possibility of being wrong here and there. ;)>> One can save a backup of the air file, then write it over >>any FSEdit messed up air file if FSEdit is used to edit >>aircraft.cfg. >>You are suggesting that we can use FSEdit to edit values in >aircraft.cfg, using the nice tab interface, then let it >overwrite the air file, but because it is botched, we >replace the air file with the old one. That would seem to be a way to get rid of unwanted changes in the AIR file. I just used Aired to Delete the new records when I had acciently created them. I think any time FSEdit is used to open aircraft.cfg, when it is closed the new records are set. That is, I don't think one even needs to change anything. Or, just changing one LG dimension will result in a modified AIR file when FSEdit is closed.>This would mean using FSEdit only as a convenient way to >edit those values that are not processed into the air file, >correct? FSEdit is convenient for setting the Description, etc. since one doesn't have to use '/t', for 'tab', etc. However, it still reformates all my nicely edited lines, replacing things like '1', with '1.00000'. Which is really crude, especially when '1' only signifies a particular sound file.>> The offsets below are based on Reference Datum offset = 0, >>0, 0. Otherwise, the distances along the fuselage change. >>>>[airplane_geometry] >>wing_area=1700.0 << Most important >>wing_span=108.0 << Also Important >>wing_root_chord=15.7 << SET to MAC=area/span, NOT root >>length >>What does MAC stand for? Mean Chord? And do you mean the >area or the span (distance)? "Mean Aerodynamic Chord". It is approximately Area / Span.>>wing_dihedral=3.0 << No effect >>This does not have an effect on the air file? Only if FSEdit creates new records. Then, it would be used to calculate 'Cl_beta -- Dihederal Effect'. Since $8500 professional programs may not calculte this well I have little faith that FSEdit will. But, it may set a half decent approximate value. One could then change the 'wing_dihederal' with FSEdit to increase or decrease the effect to get what is appropriate in the end.>>wing_incidence=0.70 << This, and twist, affect cruise >>attitude >>This is in degrees, correct? Yes, many FS2K2 aircraft.cfg files have comments that explain these lines. Though, sometimes the explaination is useless. Similar to what a GS kid might write in an assignment. ;(>>wing_twist=-3.0 >Also called washout I think. And, incidence + twist/2 = effective incidence. That is, if incidence is +1.0 deg and twist is -2.0 deg, the end result is 1.0 + 1/2 (-2) = 0.0 degrees. However, the twist does add a small component to total parasitic drag. Aired shows REC 1101 "Fuselage AoA at Min Induced Drag" and the Info for that parameter gives a formula for setting it. If 'incidence + twist/2' changes in aircraft.cfg then that parameter should be changed. It definetely needs to be changed to keep the Induced Drag parabola the same. Rather than change incidence and twist it is probably better to change 'cruise_lift_scalar'. I never have since I know about the parameter in REC 1101 that has to be set correctly. I didn't figure out how all this works but would estimate 2000 hours of flight testing and thinking by others was required to figure out that one formula.>> However, there is a parameter in the AIR file that should >>be changed when either of the above are changed. Better to >>change 'cruise_lift_scalar' (below) to adjust cruise pitch. >>Hmm, I set my wing incidence to the angle of the a/c I was >modelling, but I did not set anything in the air file. I'd >still like to set the correct incidence given that is one >number I do have. You can change the wing_incidence as long as the combination of incidence + twist/2 is constant. However, the only result will be a small, probably immeasurable, change in drag. Further, I do not set the the actual wing_incidence for FS. In the real AC, there is downwash ahead of and behind the wing. Some wing incidence is required to correct for downwash. Perhaps 1 degree. FS does not know about downwash, so I set the wing incidence to get the appropriate cruise pitch. I have also edited the Lift Curve to change the lift at wing AoA=0, which has a similar effect to changing incidence + twist/2. Typical cambered airfoils develop a CL ~ 0.20 at AoAwing = 0 degrees. The downwash also affects the horiz_tail. When set to 0.0 deg, it sees no downwash in FS, but acts as if it were at about -1 deg (LE down) in a real AC. OTOH, the H. Stab likely has camber, but perhaps 'upside down'. That would result in a down force on the tail, which is the usual situation. In FS, I have to set the H. Stab to +1 degree (LE Up) to just approximate the effect of downwash on a non-cambered tail. But more like +2 to +3 degrees to approximate the effect in a real AC. Note the sign convention is confusing. For MS FS2K2 AIR files a postive setting pushes the tail down. >>>oswald_efficiency_factor=0.80 <>to 0.85 >>wing_winglets_flag=0 <<1=T Reduces Induced Drag - >>possibly too much >>I think these increase thrust counterbalancing induced drag. >But I could be wrong. Off topic anyway. Winglets help keep induced drag generating vortices from shedding off the wing tips, they have a similar effect to a longer wing (but, same area).>>wing_sweep=0.0 <>wing_pos_apex_lon=3.93 <>wing_pos_apex_vert=0.0 <>This is the "compromise" point of lift force on the wing. Is >that Meacn Chord? 1/4 root chord is suggested in the SDK, >somewhere it was suggested using 1/3 root chord. The standard center of lift is considered to be 1/4 of the distance back from the LE of a rectangular wing. Though, this is often 50% in a supersonic AC. Around 53% in the Concorde. Note when the wing_pos_apex_lon is set 1/4 of the MAC ahead of 0.00, then the Center of Lift comes out at 0.00. Which is normally where the CG is set to. Any small difference from 25% MAC for the Center of Lift is considered a pitching moment, and that is taken out by Trim. However, I see FSEdit sets an offset in the Center of Lift when it sets the new records. Perhaps to fix other effects of using that stupid 'wing_positioin_apex_lon' setting. However, I know setting the wing as above works well, I have set up the wing and tanks, and payload locations and seen 25.0% MAC in AFSD as soon as I loaded the AC. The Percent CG variable AFSD reports appears to be off when other ways are used to set the wing, etc. Some AC appear to be well balanced but show -10% or so. I avoid such confusion by setting things up in a standard way to start.>>htail_incidence=-2.0 <<0.0 to -3.0 may improve pitch >>stability >>This is the angle the horizontal stabilizer makes with a >line perpendicular to the fuselage, correct? An horizontal like that corresponds to pitch = 0 degrees. It is confusing, since +2.0 means the trailing end of the H. Stab is up 2.0 degrees, which 2.0 wing incidence would mean the TE is down. And, it depends on two variables in the AIR file. So, if one wants to set the H. Stab deflection just verify that it is set with a sign that forces the tail down, the nose up. Which means more Down Trim.>>elevator_up_limit=27.0 <>Will have to look for this data. Try . One can DL pdf files for all Type Certificate Data Sheets they have. Which covers just about any AC ever licenced in the US but for military AC. >>[Reference Speeds] >>flaps_up_stall_speed=165 <>full_flaps_stall_speed=132 <>cruise_speed=434.8 >> I think this should be in IAS cruise speed if used as an AI >>aircraft. So it doesn't fly too fast. That would be about >>260 kts for jets. >>I am trying to discover what defines the overspeed for an >a/c. Is it this cruise speed. You're saying cruise speed is >in true knots airspeed not indicated? The overspeed warning is set in the AIR file. You can hit F2 in Aired and search for 'overspeed'. Hit F3 to 'search again'. I just found the two settings, the first in in Mach, and the second is in Record 1101 and is in IAS. The overspeed warning goes off when the first overspeed is hit. At lower altitudes is is generally the IAS limit, for jets at FL 350 it is usually the Mach number. However, a good flight model may not be able to hit the Mach limit at higer altitudes and weights. The Concorde airframe type (set in REC 1101) sets different speed limits and the previous ones have no effect. FS2K2 uses the FS2K Concorde AIR file for its supersonic prototype, so don't expect the normal overspeed limits to work. At least over much of the range. I think the Mach limit may have an effect for the Concorde in Cruise at over FL 500. I've read in AVSIM Forums that 'cruise_speed' is in TAS, but should be set in IAS for 'automatic AC' to fly at the correct speed. Otherwise, they fly too fast! The cruise_speed= line will be set with a value when an AC is loaded if it was blank. That should be near the expected TAS cruise speed. However, if one sets it in kts, I doubt it has any effect on the FS AC. But, should fix AI AC.>>>>[TurbineEngineData] >>fuel_flow_gain=0.004 <>rate inlet_area=9.80 << Set near real turbine inlet area. >>Higher values increase ram drag, less net thrust at high >>speeds. >>rated_N2_rpm=12250 <>static_thrust=17000 <>Often this should be lower than TO rating, full throttle >>usually adds a lot to this number. >>afterburner_available=0 >>reverser_available=1 >>>>>>[GeneralEngineData] >>engine_type= 1 >>Engine.0=-49.7,-9.3, 2.00 //set up for thrust vector >>Engine.1=-65.1, 0.0, 3.00 //approx center of thrust >>Engine.2=-49.7, 9.3, 2.00 //1.00 >>>> These locations are for the "center of thrust", not the >>engine! Prop location or about 2/3 of the way back along a >>turbine. Up/Down will add a pitching moment with power >>changes. >>Okay, that is closer to what I finally discovered, that this >is another 'virtual' value representing a force, not the >engine location. But I had little idea where to put it >besides equal to the prop location. Both the turbine inlet 'suction' and jet thrust have an effect on net thrust. I simply assumed the result is about 2/3 of the distance along the turbine nacelle. Generally, this is not critical. FS doesn't know anything about the engine weight or location. Those elements are included in the empty weight and MoI's. Clearly, the location of the prop is what is important. Again, fore/aft doesn't have a big effect, but one might as well set the location approprietly.>>>>>>fuel_flow_scalar=1.60 >>>> This should always be above 1.0 for turbines. >>SFC=fuel_flow_scalar/2. Typical SFC cruise values range >>from 0.55 0.90. So, fuel_flow_scalar would be 1.1 to 1.8. >>>>What is SFC? Specific Fuel Consumption. It is given in lb(thrust)/lb per hour. A JT8D has an SFC of about 0.82 in high altitude cruise. When the scalar is set to 1.0 the SFC comes out at 0.500. Too low. Advanced turbines in 747-400: SFC ~ 0.55. ERJ 135/145: SFC ~ 0.58. JT8D in 727: ~ 0.82. When the SFC is set appropriately, and Cdo (or the parasitic_drag_scalar) set to get the correct fuel flow at a given cruise condition one will find that jets don't slow down excessively fast. For example, I see about 90 seconds required to slow from 350 kts IAS to 220 kts at 10,000 ft in one of my jets. My FS2K2 AC was within 5% the times a full motion sim of the same AC gave. At 10000 ft 'Mach Drag' should be neglible. In fact, it shouild be neglible in most any swept wing AC up to Mach 0.8. >I really had no idea what to set this one at. I understand >it is some non-unit reprsentation of fuel flow. I just had >to play with it until the fuel flow looked right. >Steve I gave more info above. There is a parameter for FS98 AC in the AIR file that has a similar effect, however it is not linear. Ron

Share this post


Link to post
Share on other sites

Thank you Ron for increasing my understanding of these aerodynamic terms and and clarity of what FS is really doing with the values I give it. It will be a while before I can take in all of your information, but I understand why I am having trouble getting this aircraft to fly correctly.I revisited the aerodynamics 101 site this weekend and found the Control section helpful. Especially the part about Center of Gravity and Aerodynamic Center. I can see better how these relate to what FS calls wing apex. The previous definition of Wing Apex I had implied that it was created only because the sim could not represent the lift force over the entire wing, but it appears that it represents aerodynamic center.http://142.26.194.131/aerodynamics1/Stability/Page7.htmlAs you mention, the page states that the ac for most subsonic aircraft is very close to 25% of chord. Now I know where MS gets that 1/4 chord value from. And as you note, 50% for supersonic aircraft.Finally, I have some understanding of dihedral and a clue about why the Cessna's have a high wing and why they fly the way they do. Why they fly differently than the low wing aircraft and why I probably have the trouble I do with the low wing. Also, it explains why the Extra is so touchy. It is designed that way so it does not have a lot of resistance to rolling for aerobatics purpose.http://142.26.194.131/aerodynamics1/Stability/Page5.htmlI don't pretend to understand everything I have read, but it does help to see the reality behind the FS fog. The movies are cool too.I can clearly see why my aircraft was terribly unstable with the CoG aft of center. In fact, it tried to go straight up like a helicopter. I had spent a lot of time wondering where the CoG should be, where the wing chord lift point is in relation to the CoG, where in relation to the wheels, whether FS models the wheel contact points as a pivot point for balance on the ground, etc. I wondered how having a high wing versus a low wing affected CoG. Because I knew I had a problem with balance. Now I have a much better idea of these things.I also experimented with a tailwheel version of the aircraft. I discovered that I had to move the wheels forward before it would settle on its tail instead of its nose. Checking other tailwheel aircraft I found that they follow this plan, with the wheels are placed closer to the front, creating a more stable triangle of contact points than they would be if you leave the main wheels were they might be with tricycle gear.This is really more than I wanted to get into but it seems the only way for an aircraft there really is no published data for and probably difficult to find someone who has flown one like it to confirm the FS model.Steve

Share this post


Link to post
Share on other sites

>Thank you Ron for increasing my understanding of these >aerodynamic terms and and clarity of what FS is really doing >with the values I give it. It will be a while before I can >take in all of your information, but I understand why I am >having trouble getting this aircraft to fly correctly. I've had problems myself, and came up with the approach I gave for setting the 'wing_apex_position_lon' and 'wing_root_chord'. The Apex position may be easy to scale from a drawing, but has little aerodynamic significance by itself. The important thing is to set the wing location so the 'apex' is 1/4 the length of the MAC ahead of the FS_Reference. Which given as 1/4 the MAC length behind the LE. The MAC can be calculated as Area/Span. Again, this assumes FSEdit is NOT used to generate new records in the AIR file. Now FSEdit may set 90% of the new parameters reasonably well, but one has to know what is reasonable and what needs to be changed. I have gone into this in the avhistory.org AIR file Forum. In general, it is better not to use FSEdit. ;)>I revisited the aerodynamics 101 site this weekend and found >the Control section helpful. Especially the part about >Center of Gravity and Aerodynamic Center. Yes, I had looked over the site again a while ago. >definition of Wing Apex I had implied that it was created >only because the sim could not represent the lift force over >the entire wing, but it appears that it represents >aerodynamic center. The Aerodynamic Center of Lift is NOT at the Apex. It is behind the leading edge of the "Mean Chord", which is well behind the apex of a swept wing.>As you mention, the page states that the ac for most >subsonic aircraft is very close to 25% of chord. Now I know >where MS gets that 1/4 chord value from. And as you note, >50% for supersonic aircraft. I set one supersonic Fighter to balance at 48%. >Finally, I have some understanding of dihedral and a clue >about why the Cessna's have a high wing and why they fly the >way they do. Why they fly differently than the low wing >aircraft and why I probably have the trouble I do with the >low wing. I found that setting the wing at +2.0 ft for a high wing Cessna gave more wing leveling, and I reduced the 'Dihederal Effect' parameter in the AIR file. In contrast, a low wing AC is less stable, and has more physical dihederal, which means a higher 'Dihederal Effect' set in the AIR file. However, I generally change that parameter to give reasonable wing leveling in my flight tests. >Also, it explains why the Extra is so touchy. It >is designed that way so it does not have a lot of resistance >to rolling for aerobatics purpose. The Extra has a symmetrical wing with no dihederal. Which makes it less desirable for making long trips in.>I don't pretend to understand everything I have read, but it >does help to see the reality behind the FS fog. The movies >are cool too. I've checked many aerodynamics sites over the years and slowly pick up these concepts. >I can clearly see why my aircraft was terribly unstable with >the CoG aft of center. In fact, it tried to go straight up >like a helicopter. I had spent a lot of time wondering where >the CoG should be, where the wing chord lift point is in >relation to the CoG, where in relation to the wheels, >whether FS models the wheel contact points as a pivot point >for balance on the ground, etc. I've noted that the main gear is typically 75% of the MAC behind the LE. That is, about 1/4 the length of the MAC ahead of the trailing edge. Normally, the CG varies about the MAC/4 point since the Center of Lift is normally close to MAC/4. This results in an AC that can be rotated fairly easy for takeoff. Too far forward and the AC can tip on its tail. Also, the weight on the nose wheel decreases and the front tire skids when you try to turn. I figure the real AC is designed with all these things taken into account so I work to set my FS AC the same way. As long as FS models the effects correctly the critical locations should be set to what the real AC is.>This is really more than I wanted to get into but it seems >the only way for an aircraft there really is no published >data for and probably difficult to find someone who has >flown one like it to confirm the FS model. >>Steve I've learned most of what I know about AC though MSFS. ;)Ron

Share this post


Link to post
Share on other sites

RonThis information is greatly appreciated, it's obvious that you have devoted an incredible amount of time to this complex and difficult subject and I can't tell you how grateful I am that you took the time to post it here. I had to work all weekend and I am just getting around to reviewing it, but already I can see that I will be busy with it for some time to come. I am also reminded that my earlier suspicions were correct, I really know very little about modifiying the flight simulator flight dynamics! I wish I could contribute 10% of what you have so kindly provided here, but unfortunately I find I have more questions than answers at this point. In working on a new flight model for the Microsoft Dash 8-100, and I have encountered difficulties with the pitch attitude on final approach with flaps extended, the aicraft pitch begins to increase towards nose up as the speed falls below approximately 115 KIAS and the aircraft approaches the runway nose high and "behind the power curve". To solve this, I have edited the X & Y values in TBLDB 401 *Wing Cl vs Mach Factor. By increasing the Y value of the second pair from 1 to 1.8, and moving the X value to the left (towards what seems to be the lower end of the a/c speed range) by reducing its value from .2 to 1.75 I have managed to achieve a pitch angle of 4 degrees nose down with full flaps extended at 95 KIAS, on a 3 degree glideslope (about 500 fpm rate of descent). TBLDB 400 ground effect graph required editing after this change to compensate for the increased lift. I reduced the Y values of the remaining pairs (.51-.30) to give the aircraft an operational ceiling of about FL280 with a 300 fpm rate of climb. The aircrafts speed envelope seems quite small in relation to this table, perhaps the first 3 or 4 pairs are used (I could be wrong!), and my concern is that I have this huge spike at what is essentially the 30% position of the operational speed envelope, with steep ramps leading down on either side of it and any proper L/D curve is hopelessly skewed. While the flight characteristics exhibit no problems while hand flying or using the autopilot, I wonder if there might be another way to solve these problems. As far as the operational/service ceiling is concerned, perhaps by limiting the engine power at altitude somehow ( something like the "critical altitude" CFG entry for turbocharged recips)? After suffering through the loss of some of my earlier work when FSEdit corrupted the AIR/CFG files beyond repair I removed it and all related files from my hardrive. With each sucessive incarnation of FS2002 (whenever I reinstall the program) I remove FSedit before loading any aircraft that I care about, and I haven't suffered any ill effects as yet. Once again, thanks for the info!

Share this post


Link to post
Share on other sites

>Ron >>This information is greatly appreciated, it's obvious that >you have devoted an incredible amount of time to this >complex and difficult subject and I can't tell you how >grateful I am that you took the time to post it here. Numerous individuals have worked on figuring out the MSFS AIR file etc. over the years. Unfortunately, MS' concept of FSEdit wasn't correctly realised with FS2K2. Rather than simplify matters, it just makes things worse. I expect there is an FSEdit included in CFS3, but I don't know if it's been been fixed for not.>I had >to work all weekend and I am just getting around to >reviewing it, but already I can see that I will be busy with >it for some time to come. I am also reminded that my earlier >suspicions were correct, I really know very little about >modifiying the flight simulator flight dynamics! Ideally, one should be able to get a reasonable AC using just basic information such as set in FSEdit and in the AIR file. Unfortunately, it didn't turn out this way. Nor did MS bother to explain things I figured out over the last year. >could contribute 10% of what you have so kindly provided >here, but unfortunately I find I have more questions than >answers at this point. In working on a new flight model for >the Microsoft Dash 8-100, and I have encountered >difficulties with the pitch attitude on final approach with >flaps extended, the aicraft pitch begins to increase towards >nose up as the speed falls below approximately 115 KIAS and >the aircraft approaches the runway nose high and "behind the >power curve". There are various basic matters that could cause that. First, I always want to get the CG in the correct place. And, have the wing (actually, the virtual rectangular wing) in the correct longitudinal location. I outlined how to do this. I also use Herve' Sors' AFSD. Or, a panel test gauge. Jerry Beckwith has been working on this stuff for CFS AC for years and has a test gauge, "JWB.gau" available at his site: . I see he also has a new, 'combined' test gauge at: . In the past the 'jet' gauge didn't work in FS2K2, but perhaps the new, "combined" gauge will. Even though it says "FS2000/CFS2" I use the "Original" gauge's prop gauge in FS2K2 to show CG as %MAC, Trim, Pitch, AoA, and many other things. There is more than one needs in the small DL's. Many things are for CFS2. Just put the JWB.gau or the newer version in GAUGES and add the appropirate window to any panel you want to use it with. You do not need to use Jerry's test panel -- it is mainly for CFS2. ---------------------- > To solve this, I have edited the X & Y values >in TBLDB 401 *Wing Cl vs Mach Factor. By increasing the Y >value of the second pair from 1 to 1.8, and moving the X >value to the left (towards what seems to be the lower end of >the a/c speed range) by reducing its value from .2 to 1.75 I >have managed to achieve a pitch angle of 4 degrees nose down >with full flaps extended at 95 KIAS, on a 3 degree >glideslope (about 500 fpm rate of descent). TBL 401 affects the 'slope' of TBL 403 as Mach Changes. Sorry, it is not the appropriate table to change to fix your problem. As far as Jet AC go, TBL 403 is fairly reasonable in the MS 777 AIR file. One can use Aired to copy it to a clipboard, then 'Replace' TBL 404 in another AIR file with one from any other AIR file. But, I found a better set of TBL's for you (latter).> TBLDB 400 ground >effect graph required editing after this change to >compensate for the increased lift. Normally, TBL 400 only has an effect when altitude is within one wing span of the ground. And, the effect is very small at that altitude, it only becomes significant at about 0.1 the span.>....... While the flight characteristics exhibit no problems >while hand flying or using the autopilot, I wonder if there >might be another way to solve these problems. Looking at the MS Dash 8 AIR file, I see TBL 478 is inappropriate. This sets the AC pitching momement vs Body AoA (what the test gauge, etc. show). The table is inverted relative to the normal convention in MS AIR files; that is OK. However, I see a '45 deg' slope in the middle, that drops at AoA = 0.1745 radians (10 deg). Then, it increases at a lower rate. The effect of this is that beyond 10 deg AoA, there would be less 'nose down' effect, and AC trimmed at 9 deg AoA would tend to nose up when the AoA increased past 10 degrees. Now 10 deg AoA is too high for landing, I see TBL 403 shows a Lift Coefficient of 1.62 at 10 deg, and the stall is at an excessive CL of 2.0 at 14.0 degrees. Ah, the MS Beech_King_Air_350 AIR file has reasonable TBL's 403 and 473. I'd suggest you replace the ones in the Dash 8 with the King Air tables. Also, put TBL's 400 and 401 back to what they were. Or, just replace then with those from the King Air. I can't guarantee anything, Donna. But, these changes should at least get some things appropriate in your AIR file. > As far as the >operational/service ceiling is concerned, perhaps by >limiting the engine power at altitude somehow ( something >like the "critical altitude" CFG entry for turbocharged >recips)? That doesn't affect turboprops. There are ways to change the turboprop tables to change the power with altitude and Mach number, but you don't want to hear. ;) I haven't been completly successful myself in setting up turboprops. However, AFSD V 1.3 (in AVSIM Lib) will show turboprop Shaft HP, Equivalent SHP, fuel flow, thrust, and other things. So, you can at least see what the current HP is and change the 'turboprop torque' (something like that) in aircraft.cfg to set the appropriate TO HP. This should give you something to keep you busy for a while. ;) If you get your Dash 8 flying reasonably well, then we can go on to setting drags, etc. Ron

Share this post


Link to post
Share on other sites

Hi Donna,Whilst you can delete FSEdit from your system, the FS2002.exe will still try to put your house in order. The example is usually an FS98/2000 imported aircraft that gets FS2002-ized upon selection. I am not sure when and how FS2002.exe knows not to mess with a file after it have been FS2002-ized, maybe just because it sees new areas in the aircraft.cfg or *.air files.As far as I know, FS2002.exe doesn't call FSEdit.exe to make the changes since FSEdit doesn't exist in the FS2002 STD addition.

Share this post


Link to post
Share on other sites

Hi, Ron.Thanks again for your information and dedication. While other factors are important in the Feel of a simulation, for Real, Serious, and Knowledgeable Pilots the FLIGHT DYNAMICS has to be and IS the most important factor in a simulation. It

Share this post


Link to post
Share on other sites

> I also use Herve' Sors' AFSD. Or, a panel test gauge. >Jerry Beckwith has been working on this stuff for CFS AC for Last night I used AFSD for the first time to monitor flight of the aircraft I am designing. I am beginning to see the value of this application because it displays the Center of Gravity as computed in real time. Using the Empty Weight Position (which I assume is a _static_ center of gravity excluding variable weights such as fuel and pilot and baggage) and adding in the current remaining fuel, pilot weight and other defined station loads.Maybe I am wrong on this, but it finally occurred to me that the aircraft CoG needs to be in a certain place in relation to the center of lift for the aircraft to fly safely and optimally. The airframe, etc. has to be designed so the CoG is where it needs to be, not the other way around. It really does not matter if the aircraft itself is well balanced as much as the CoG in relation to the wing.I think that I saw this when playing with balsa gliders as a kid. We used to tape a penny to the nose to get them to fly with more stability. I assume this put the CoG fore of the center of lift, making the wing more stable. Aft CoG makes a wing unstable. Of course, once the airspeed feel off the glider would nose down more readily to the floor with the penny than without.You could probably get the glider to balance perfectly with your finger supporting the stick fuselage, but I doubt that would be the right CoG for flight. You would want to balance it on your finger at the CoL point of the wing for flight.AFSD allows you to see where the CoG is in relation to the CoL when setup correctly. WRT the above example, although the wing is more stable the more fore the CoG, you reach a limit where there is too much nose down. So there have to be fore and aft GoG limits and you can watch the CoG in AFSD while adding and removing fuel and station loads to be sure CoG is within limits.Also, I can see where if the aircraft design needs the CoG to be a bit forward for stability, that trim can compensate for this to keep cruise level.Right now, my Empty Weight CoG is behind the CoL slightly. I placed this by trail and error adjusting until the aircraft did not try to tip back vertically. I can tell it is still a little unstable and now I know why, with the aft CoG.However, I have discovered through AFSD that TRIM is active in the simulation all the time, even when I do not want it for this aircraft that has no trim and was defaulting to -14 incredible! Now my CoG is all wrong because of this.I see a setting in AirEd for Elevator Trim TRUE/FALSE. Will this get rid of trim on my aircraft?I have one more question. I am confused about where the Visual Model Center and Reference Datum should be. The VMC is at 0,0,0 in the GMAX scene. However, I can move the visual model anywhere in relation to this point. Where should it be? Should I drag the visual model around so that the Empty Weight CoG is at 0,0,0? Should the Reference Datum be located at the Empty Weight CoG position? I think you mentioned that the 0,0,0 is usually the center of gravity location (Empty Weight or in flight?).The MS docs talk about setting Reference Datum to the published Reference Datum of the aircraft being modeled so that published locations for loading can be used. To me this implies that Reference Datum is at the same location as Empty Weight CoG.Okay, I have* Wing Apex: line perpendicular to where the extreme leading edges join at the centerline (in rectangular or swept wing).* Leading Edge: for purposes of our calculation either rectangular wing or the extreme leading edge at apex for swept wing* Apex is placed 1/4 MAC ahead of Reference Datum* Reference Datum is placed 1/4 MAX behind leading edge* These are two ways of stating the same location and distance* In this case, Reference Datum corresponds to standard Center of Lift?* Where is Empty Weight Center of Gravity located and where should I expect In Flight Center of Gravity to be located?I assume that if Reference Datum is Center of Lift then you do not want CoG aft of it because the wing will be unstable. And you do not want fuel and etc. to push it too far fore of CoG although still stable. You want to have "dynamic" CoG fall in a range between CoL and some point forward where trim can still keep the aircraft from nose down excessively.I think I am getting closer, but still have a lot to sort out. You've probably already answered many of these questions, but they still have not clicked with me.Steve

Share this post


Link to post
Share on other sites

>Hi, Ron. >Thanks again for your information and dedication. While >other factors are important in the Feel of a simulation, for >Real, Serious, and Knowledgeable Pilots the FLIGHT DYNAMICS >has to be and IS the most important factor in a simulation. Certainly. While I work on details such as fuel consumption, it doesn't make that much difference in most cases. I never fly the several hours it takes to run the tanks dry. Many aspects of a good flight model are settable in aircraft.cfg. One should be able to use a good, similar AIR file to the AC he wants to create and change only aircraft.cfg to get a reasonable end result. By changing the wing area, weight, MoI's, power, etc. in aircraft.cfg he scales the AC to a new variation without having to touch the AIR file. This is becaues many of the parameters in the AIR file are normalized, and are often about the same in all aircraft. Microsoft moved the scaling factors such as weight, wing area, power, etc. to aircraft.cfg. One can also adjust parameters set in the AIR file with various 'scalars' in aircraft.cfg.>It

Share this post


Link to post
Share on other sites

>Hi Donna, >>Whilst you can delete FSEdit from your system, the >FS2002.exe will still try to put your house in order. Ah, correct. I found that out when I set the wing_sweep parameter in aircraft.cfg to other than 0.0 deg after importing an FS2K2 AC. When the AC was loaded the AIR file was modified as if it had been touched with FSEdit. I don't think setting the wing_sweep angle always does this. Regardless, I now always set to to 0.0. The effects of sweep are already accounted for in good AIR files. >The example is usually an FS98/2000 imported aircraft that gets >FS2002-ized upon selection. I am not sure when and how >As far as I know, FS2002.exe doesn't call FSEdit.exe to make >the changes since FSEdit doesn't exist in the FS2002 STD >addition. I think one of the FS2K2 DLL's does the work. Or, perhaps sim1.dll itself. Both FSEdit and loading an AC can result in the changes. Fortunately, it doesn't occur unless one sets certain parameters in aircraft.cfg. Another reason to copy the AIR file to a BU such as C185.org. Then, you can "copy C185.org C185.AIR" to get your original AIR file back. The AIR file will have a larger file size if it has been changed. Find the parameter that resulted in the unwanted change and set it to zero (or, perhaps delete it). The ";) Ron

Share this post


Link to post
Share on other sites

>Last night I used AFSD for the first time to monitor flight >of the aircraft I am designing. I am beginning to see the >value of this application because it displays the Center of >Gravity as computed in real time. Using the Empty Weight >Position (which I assume is a _static_ center of gravity >excluding variable weights such as fuel and pilot and >baggage) and adding in the current remaining fuel, pilot >weight and other defined station loads. This has turned out to be another long reply. I hope it isn't too confusing. AFSD and test gauges display an FS2K+ Token Variable that is the current CG relative the the MAC (Chord of virtual rectangular wing). The only panel that used this was the FS2K2 Concorde, but knowing the variable is very useful for an AC. Real AC often have balance limits defined relative to the MAC. 25% is taken to be the Center of Lift in most cases. Since most real wings are close to 1/4 the distance back from the leading edge. The range for transport aircraft is typically 15% to 35% MAC. Though sometimes it might be different. In part, due to how the location of the MAC is defined. Light AC are more confusing. Usually the Balance is given as an acceptable range for a 'Moment Arm'. Private Pilots are required to pass an FAA written test that includes problems of Weight and Balance. Most of us soon forget how to do it after the test. ;) >Maybe I am wrong on this, but it finally occurred to me that >the aircraft CoG needs to be in a certain place in relation >to the center of lift for the aircraft to fly safely and >optimally. The airframe, etc. has to be designed so the CoG >is where it needs to be, not the other way around. It really >does not matter if the aircraft itself is well balanced as >much as the CoG in relation to the wing. Absolutely Correct! An AC may not be stable if empty. But, with a pilot it should be appropriately balanced. The old J-3 Cub has two seats, one ahead of the other. If only the pilot is in the AC he normally sits in the rear seat to get appropriate balance.>I think that I saw this when playing with balsa gliders as a >kid. We used to tape a penny to the nose to get them to fly >with more stability. I assume this put the CoG fore of the >center of lift, making the wing more stable. Aft CoG makes a >wing unstable. Of course, once the airspeed feel off the >glider would nose down more readily to the floor with the >penny than without. The more the CG is set forward, the more stable the AC is in Pitch. However, one doesn't want it so stable it is hard to rotate for TO, etc. An aft CG makes the AC more responsive in Pitch, but when one gets too far back it won't even be stable without constant control by the pilot. Actually, it's the "Neutral Point" that determins the static pitch stability. This point is normally behind the 25% MAC location. Perhaps at 50% MAC. I've acciently set payloads incorrectly and had AC that were balanced too far aft. Such AC are not pleasant to fly. Whether simulated or real. In all cases it is assumed one can Trim the pitch so no control force is needed. But, that doesn't mean the AC is stable.>You could probably get the glider to balance perfectly with >your finger supporting the stick fuselage, but I doubt that >would be the right CoG for flight. You would want to balance >it on your finger at the CoL point of the wing for flight. And, it is common to use two fingers, one at each end of the wing, to get the balance about 1/4 of the way back from the wing's LE. That's a start, small adjustments will change the final flight dynamics.>AFSD allows you to see where the CoG is in relation to the >CoL when setup correctly. WRT the above example, although >the wing is more stable the more fore the CoG, you reach a >limit where there is too much nose down. So there have to be >fore and aft GoG limits and you can watch the CoG in AFSD >while adding and removing fuel and station loads to be sure >CoG is within limits. The only problem is some FS AC are stable even though the CG is way off. It's due to other things. For example, FSEdit sets a new CoL offset for the Wing that may be "-2.5" (ft). That may be to adjust for setting the wing_apex in aircraft.cfg, rather than LE of the MAC. And, may apply more to swept wings. Regardless, it is confusing and another reason to not let FSEdit do things behinds one's back. Setting AC as I suggested in my 'Comments on aircraft.cfg' appears to give appropriate balance and AFSD, etc. gives appropriate % MAC values. This is also the way REAL AIRCRAFT are defined. By the virtual wing, rather than a tricky wing with sweep, taper (ratio of tip to root chord), etc. >Also, I can see where if the aircraft design needs the CoG >to be a bit forward for stability, that trim can compensate >for this to keep cruise level. In fact, it is common to set the Horiz Stabilizer so it pushes the tail down in normal flight. Note that means the CG will then have to be forward where it would otherwise be. I've read that the down force on the tail is typically 5% of the load the wing supports and keep this in mind when setting 'horiz_tail_incidence'. AFSD shows this Horiz Stab Force as a Coefficient of Lift (CL). AFSD also shows the Wing CL. Which is typically 0.35 in cruise. 5% of 0.35 = 0.017 and I typically see values around -0.025 in my AC. Note the sign is negative, since the lift is downward. This down force on the tail adds to the load on the wing. Typically 5%. Now I've seen pretty stable FS AC with no tail incidence. That is only one part of what determings 'Static Pitch Stability'. Which relates to how well and AC stays at constant climb, descent, or cruise pitch. Normally, changing the incidence of the H. Tail will also change the trim. However, I've noted most MS FS2K2 AIR files have set the pitching moment due to H. Stab incidence to zero. Leaving the H. Stab 'Lift'. It's the 'Lift' that affects the pitch stablility, and this 'trick' eliminates the need to change the internal 'trim' offset. This can be confusing, since most AIR files are set so changing the horiz_tail_incidence will change the pitch trim. There is a paramter in REC 1101 of the AIR file which mentions "(trim)", and that can be adjusted so the panel trim is near zero in cruise. However, setting the horiz_tail_incidence to 0.00 often still results in stable AC, so one might only get into this if he is willing to experiment and understand the details. Conversly, FSEdited AC require a lot of trim. Due to an error in the way the new parameters are set. ;( In such cases, one might be able to change the horiz_tail_incidence to move the panel trim closer to the center. I think imported AC get the horiz_tail_indcence set from the value in the AIR file. Which is where it is set in pre-FS2K2 AC.>Right now, my Empty Weight CoG is behind the CoL slightly. I >placed this by trail and error adjusting until the aircraft >did not try to tip back vertically. I can tell it is still a >little unstable and now I know why, with the aft CoG. I found that the FAA specs sometimes give no 'empty weight' CG. It's the loaded CG that counts, and I'd assume that one pilot would be enough to move the CG within the acceptable limits. However, when there is an 'empty weight' CG spec, one can adjust "empty_weight_CG_position=x,0,0" in aircraft.cfg to get it to the official % MAC (or, the equivalent balance arm moment). Just comment out any station loads (you need station_load.0 set to 0.0 lbs or it will be added and set to 0.000 0.000, 0.000, 0.000) and set zero fuel in the tanks. Ifreference_datum_position=0.00, 0.00, 0.00empty_weight_CG_position=0.00, 0.00, 0.00Then the[WEIGHT_AND_BALANCE]empty_weight =110000 Will be set exactly at the reference_datum_position. In this case, offset in x, y, and z by 0.00 ft from the FS Reference. Which is at 1/4 MAC, etc. One can also change that 'empty weight' location to change the loaded balance of the AC. Since it essentially moves the empty weight setting by x, y, z ft (long, lateral, vertical). In fact, with the FS datum at the vertical offset of the wing (which can be above the vertical CG in a high wing AC) one might set the empty weight to 0.0, 0.0, -2.0 to move the CG down. Which may improve or cure runway jittering. >However, I have discovered through AFSD that TRIM is active >in the simulation all the time, even when I do not want it >for this aircraft that has no trim and was defaulting to -14 >incredible! Now my CoG is all wrong because of this. Pitch Trim comes up at bad negative values when an AC is loaded in FS2K2. Annoying. It needs to be set to a positive (up) value for TO, but should be near 0 degrees in cruise. One way to kill the pitch trim would be to set elevator_trim_effectiveness=0.0 However, your AC will have to be balanced correctly so it is reasonably trimmed in flight. 15% to 35% MAC may be acceptable; the most desirable value can be determined by experiment. Which means some pitch trim is needed, even though it can't be changed. Changing horiz_tail_incidence +/- 2 deg may trim it, but the appropriate way is to change "Pitch Moment Coefficient at AoA=0 (trim)" in REC 1101 of the AIR file. Typically it comes out in the range of "+/- 100". If an AIR file sets aileron trim and one doesn't want it for a simple AC he can set: aileron_trim_effectiveness=0.0>I see a setting in AirEd for Elevator Trim TRUE/FALSE. Will >this get rid of trim on my aircraft? It may. Though REC 320 seems to be replaced by lines in aircraft.cfg now.>I have one more question. I am confused about where the >Visual Model Center and Reference Datum should be. The VMC >is at 0,0,0 in the GMAX scene. However, I can move the >visual model anywhere in relation to this point. Where >should it be? Should I drag the visual model around so that >the Empty Weight CoG is at 0,0,0? The .mdl should probably be set to 0, 0, 0. Then one knows where it is in the AC flight model. 1/4 MAC behind the LE. For swept wings this point isn't obvious in the spot view, but I eyeball it. Or, the CoV could be placed at the actual CG of the AC. Which would be below the wing in a high wing AC, a bit above for a low wing. Remember that wings with dihederal have the CoL above the root. For big jets it appears to be near the vertical center of the fuselage, especially after the wings have flexed up in flight. This isn't all that critical, but it does affect wing leveling. The FS2K Concorde is a good example to study. It places the CoL 63 ft behind the CoV. Note there is a Center of Lift Record in the AIR file that sets the CoL at -63.0. While aircraft.cfg has the references set for the official factory reference. The idea is to have the AC rotate about the correct point in the spot view. And, have the pilot's eye at the correct point so the runway falls away appropriately in rotation. I check the spot view while I move the JS. > Should the Reference Datum >be located at the Empty Weight CoG position? I think you >mentioned that the 0,0,0 is usually the center of gravity >location (Empty Weight or in flight?). The Reference_datum is often set to 0, 0, 0. Everything is referenced from it. However, I think it might be good to set it something like 0.0, 0.0, -2.0 for a high wing Cessna. Below the CoL and CV, but balanced along the fuselage And, set empty_weight_cg_position = 0.0, 0.0, 0.0. Perhaps directly in line with the prop shaft. Then, the prop thrust will be directly ahead of the CG and the passenger station loads can be at 0.0 vertically. >>The MS docs talk about setting Reference Datum to the >published Reference Datum of the aircraft being modeled so >that published locations for loading can be used. To me this >implies that Reference Datum is at the same location as >Empty Weight CoG. No, that means it might be at +6.0, 0.0, -2.0 for something like a C172. Often the offical Reference Datum is at the prop spinner tip. In this case, one would set the empty_weight_cg_position=-6.0, 0.0, 0.0. Note the FS Cessna 182's have that CG set slightly behind the CoL of the wing. Apparently where the Real AC's CG is when empty.>* Wing Apex: line perpendicular to where the extreme leading >edges join at the centerline (in rectangular or swept wing). >* Leading Edge: for purposes of our calculation either >rectangular wing or the extreme leading edge at apex for >swept wing The Wing Apex is meant by MS to be the LE apex of a swept wing, or the LE of a rectangular wing. Note I always assume a rectangular wing and set the 'wing_apex' to the LE of such a wing. MAC/4 ahead of the CoL (often 0.0, 0.0, 0.0. Thus, my setting for this wing apex is not at the visual apex. >* Apex is placed 1/4 MAC ahead of Reference Datum >* Reference Datum is placed 1/4 MAX behind leading edge >* These are two ways of stating the same location and >distance >* In this case, Reference Datum corresponds to standard >Center of Lift? Yes.>* Where is Empty Weight Center of Gravity located and where >should I expect In Flight Center of Gravity to be located? Relative to the reference_datum, at or near the FS Reference. Often 0.0, 0.0, 0.0. But -X, 0, 0 if the reference_datum is at +X, 0, 0. >I assume that if Reference Datum is Center of Lift then you >do not want CoG aft of it because the wing will be unstable. I like to see 25% MAC CG on a test gauge when an AC is fully loaded. Or, at a typical flight condition. However, this can vary some if one sets realistic station loads. 22% to 28% MAC doesn't seem to make much differene in pitch stability as long as the AC can be trimmed for the current CG. Since you don't want to have a pitch trim, you need to have the AC interally trimmed so the CG can be adjusted so balance is near 25% MAC. See my comment near the top about the parameter in REC 1101 of the AIR file that can adjust pitch trim. >And you do not want fuel and etc. to push it too far fore of >CoG although still stable. You want to have "dynamic" CoG >fall in a range between CoL and some point forward where >trim can still keep the aircraft from nose down excessively. I've noted fuel tanks are often a bit behind the CoL. As they drain the weight decreases, but so does the down force behind the wing. The end effect is little change in trim is needed to account for fuel use. I expect this is by design. In big jets, the center tank is somewhat forward of the CoL, and the wing tanks are behind it. I've set them so the CG is at 25% MAC when full, though the real AC may be a bit different. Also, the wing tanks should be set about 1/3 of the distance to a wing tip if they fill most of each wing. Note the wing is usually tapered and most of the fuel is inboard from the center of each wing. Wing tip tanks should be placed where they really are. Half the wing span from the fuselage center for each one.>I think I am getting closer, but still have a lot to sort >out. You've probably already answered many of these >questions, but they still have not clicked with me. >Steve I have gone though all this confusing stuff many times and often set something wrong. I see the % MAC is way off. So, I comment out the station loads, set zero fuel in the tanks and see if the CG is now right. Then, look thought my settings to see which is wrong. I found I'd set one of two station loads at the wrong location yesterday, after I fixed that CG came out at 25%. While I don't think much of MS AIR files, they seem to get the stuff in aircraft.cfg right. At least consistent for the correct wing location, etc. Look at different MS AC to see the variations (thought, the CoL offset in an AIR file may be offset for the aircraft.cfg settings to work. i.e. Concorde). Those who do the G-Max model can set many things, such as the LG, the same as in G-Max. Though some adjustments on 'compression', etc. may be needed to make an AC stable on the runway and the articulation may not end up completly consistent with what works well for the dynamics. Ron

Share this post


Link to post
Share on other sites

Ron, You haven't been able to see any benefit from the wing dihedral setting in the cfg? With my current projects: B-1B, F-14D, F-16C, and F/A-18C". I can see a change once I change this variable for testing purposes. I use the setting as per the specs for the aircraft. Great info that you have provided. take careJulian

Share this post


Link to post
Share on other sites

Ron,thanks for you all your help. The aircraft is now flying with greatly improved stability and controllability. Before last nights changes the aircraft would take off, nose up, begin oscillating in pitch, increasing in amplitude until it stalled. I suppose you would call this negative static longitudinal stability. Does not make for a very enjoyable flight!The first move I made was to throw out the original Ref Dat Long. The original location was leftover from the Cessna inspired SingleProp aircraft configuration I based my aircraft on. It was 3.6ft long and zero lateral, zero vertical. I did not really know what I was doing, so although I suspected it might be easier to set this to zero, so instead I dragged my visual model over to match this location measured from the visual model center (GMAX scene center).However, by last night is was clear the calculations were just getting too messy and that this offset was creating more confusion than anything else it might be good for. Trying to find 1/4 MAC back from a ref datum that was 3.6 feet forward of visual center is just much for this inexperienced designer. So I set all three components of the reference datum to zero.Then I moved the visual model so the wing model leading edge matched the new measured leading edge matching the revised reference datum. Of course, I will have to recalculate the positions of all my lights and so forth.To start with a new flight model, I calculated 1/4 MAC according to your instructions and placed the apex 1/4 MAC ahead of the new reference datum of zero. I then cleared out my old empty weight and set all three empty weight components to zero. This places the empty weight center of gravity at the same position as the reference datum longitudinal position and the center of lift there also.Once this was complete, the aircraft was able to takeoff without stalling too badly if I worked the controls and remain steady in flight with quite a bit of nose down elevator.What I discovered, as I said before, through AFSD, was that the default elevator trim was set to 19.5 degrees on aircraft initialization. Now wonder I was experiencing trouble on takeoff and confused about the center of gravity!So I set out to turn trim off or set it to neutral. I put a trim gauge on the panel, but moving it from 19.50 to 0 every time was becoming tiresome. So I went looking in aircraft.cfg for the trim setting. I found elevator_trim_limit= 0.500I discovered through experimenting that this value is not exactly what you might expect. It is both the _upper_ and _lower_ limit for the trim input. So 10 degrees limit is 10 nose up and 10 nose down. I set it to zero in a bid to get rid of trim altogether or keep it locked at neutral, but the aircraft seemed to respond badly to that setting. So I chose 0.5 half a degree as close enough to zero. It worked! No more trim problems.(for anyone reading, why I do not want trim is that my aircraft is like a Tiger Moth or other early aircraft that did not have trim)Level flight seemed much more stable and flew hands off for a half hour once stabilized.With this and the new reference and empty weight, center of gravity is reported as 1.3 by AFSD with 50 lbs of fuel in two equidistant wing tanks a foot behind reference datum and a FAA standard pilot weight of 170lbs temporarily set to 0,0,0 position from reference datum. I will probably have to move the pilot location a bit to match the seat position later.The flight performance was still not what anyone would call home about.I decided to look at the Cessna 172 to see how it might differ from my aircraft configuration.The Reference Datum is 3.6,0,0The Wing Apex is -2.4 Long, 0 VertThe Empty Weight is -3, 0, 0I did not expect the Center of Gravity (which I believe is what Empty Weight location represents), to be equal with the wing height. It would seem to me this would make the aircraft top heavy.However, if you look at the high wing as suspending the weight of the aircraft like a pendulum from a string, it may not appear so strange.Looking at the H Tail, which is very influential on longitudinal stability (pitching), I see that the Cessna 172H Tail Apex Vertical is zero.To me, this places the Apex Vertical at the same height as the Ref datum and the main wing.H Tail Apex Vertical = Wing Apex Vertical = 0It is clear from the visual model that the C172's main wing is higher than the horizontal stabilizer wing. Odd. But apparently the zero location works in flight.If you look at the SDK the H Tail Apex Vertical is shown located at the middle of the H Tail (vertical stab on the example).The aircraft seemed to handle a little better. I tried setting the H Tail further behind the aircraft for more leverage, maybe more stability, but no improvement. So I went back to the original distance.Next, I decided on a hunch to look into the wing incidence. The aircraft is controllable, but still seems to exhibit too much pitch instability and pitch up on takeoff. The incidence was one of the few figures I had on the original aircraft, 3 degrees.So I got the Tiger Moth out and did a full throttle takeoff, yes there is a little nose up tendency, but it will climb hands off with any pitch oscillations self-damped.I tried setting incidence to zero. The takeoff began to look more like the Tiger Moth. Very good!The aircraft still exhibited the same pitch instability. While digging through the air file I discovered what I think is the real culprit behind all these problems. The H Tail incidence had been naively set to 3, from the original Cessna configuration. I set it to -2.0 according to your annotations on the aircraft.cfg file and the aircraft is much more pitch stable on takeoff and climb to cruise. In fact flight characteristics are nearly where I expected they should be, not perfect on takeoff and I have not tried landing, but I think it is where I want it for now.So anyway, I jumped the gun with my previous message. You were right to suspect something else was wrong.StevePS: The landing seems less stable than it was. Tends to sink too quickly on final. This aircraft always seems difficult to line up with the runway (at least using the keyboard---my yoke is disconnected).

Share this post


Link to post
Share on other sites

Hi, Ron and everyone. While I consider myself somewhat knowledgeable in the world of Aviation, sufficient to instruct to the CFI level, I am constantly, pleasantly surprised, of the amount of in depth understanding and knowledge that some of you poses. I tried to achieve a

Share this post


Link to post
Share on other sites

>Ron, >>You haven't been able to see any benefit from the wing >dihedral setting in the cfg? Actually, I don't think I tried changing it. Since there is a parameter in REC 1101 in the AIR file for Dihedral Effect.>With my current projects: B-1B, F-14D, F-16C, and F/A-18C". >I can see a change once I change this variable for testing >purposes. I use the setting as per the specs for the >aircraft. >>Julian I'll try changing it some time. Are you sure you didn't use FSEdit to set the new AIR file records from 'Sweep', 'Dihederal', etc? Ron

Share this post


Link to post
Share on other sites

>Ron, >thanks for you all your help. The aircraft is now flying >with greatly improved stability and controllability. Good. So my approch really works. ;)>The first move I made was to throw out the original Ref Dat >Long. .........>However, by last night is was clear the calculations were >just getting too messy and that this offset was creating >more confusion than anything else it might be good for. >Trying to find 1/4 MAC back from a ref datum that was 3.6 >feet forward of visual center is just much for this >inexperienced designer. So I set all three components of the >reference datum to zero. Yes, I usually start with 0, 0, 0. When everthing is balanced well I eventually set the reference_datum to some positive number to set the real location. Then, subtract that offset from all the longitudinal entries for LG, lights, Wing, etc. It's also a lot easier to calculate moments for the passengers, fuel tanks, etc. with the reference at zero. >Then I moved the visual model so the wing model leading edge >matched the new measured leading edge matching the revised >reference datum. Of course, I will have to recalculate the >positions of all my lights and so forth. >To start with a new flight model, I calculated 1/4 MAC >according to your instructions and placed the apex 1/4 MAC >ahead of the new reference datum of zero. I then cleared out >my old empty weight and set all three empty weight >components to zero. This places the empty weight center of >gravity at the same position as the reference datum >longitudinal position and the center of lift there also. Which is appropriate. Unless the empty weight is elsewhere. In which case the station loads, fuel tanks, etc. would likely be unbalanced enought to bring the AC within the CG limits with just the pilot aboard. Regardless, the empty_weight CG should be near the CoL, only if one wants to model an AC exactly is there much reason to set it otherwise. One does want the CG to be within specs with different payload and fuel loadings, but that can be checked by watching the CG reading with only the front seats loaded, or only the pilot in front and heavy passengers and baggage aft. Or, move the empty_CG so CG is at or near 25% with a full payload and fuel.>Once this was complete, the aircraft was able to takeoff >without stalling too badly if I worked the controls and >>What I discovered, as I said before, through AFSD, was that >the default elevator trim was set to 19.5 degrees on >aircraft initialization. Now wonder I was experiencing >trouble on takeoff and confused about the center of gravity! That's the normal MS setting for SEL's. I set it to the TCDS, which may be 15.0 degrees. Or, 10 to 13 degrees for Horiz Stab Trim. As you noted, there is only one limit, so I have to set +10 degrees even if the H. Stab range is 0 to 10 degrees. >So I set out to turn trim off or set it to neutral. I put a >trim gauge on the panel, but moving it from 19.50 to 0 every >time was becoming tiresome. If you save a situation with pitch trim correct for TO, it will be set that way when loaded. Same for flaps, rudder trim, etc. But, if you reload just the AC, it goes to some negative number. Very annoying during flight tests, where I change a parameter, save aircraft.cfg or the AIR file, then reload the aC in flight. For jets I generally have to Pause the sim before reloading, then adjust the trim to it's previous flight setting before unpausing the sim. FS2K didn't have this defect.> So I went looking in >aircraft.cfg for the trim setting. I found >>elevator_trim_limit= 0.500 >>I discovered through experimenting that this value is not >exactly what you might expect. It is both the _upper_ and >_lower_ limit for the trim input. Right.>... So I chose 0.5 half a degree as close enough to zero. It worked! >No more trim problems. >>(for anyone reading, why I do not want trim is that my >aircraft is like a Tiger Moth or other early aircraft that >did not have trim) OK when payload doesn't change and fuel tanks are near CG. >Level flight seemed much more stable and flew hands off for >a half hour once stabilized. That's how most real AC fly. Though, they may turn slowly.>With this and the new reference and empty weight, center of >gravity is reported as 1.3 by AFSD with 50 lbs of fuel in >two equidistant wing tanks a foot behind reference datum and >a FAA standard pilot weight of 170lbs temporarily set to >0,0,0 position from reference datum. I will probably have to >move the pilot location a bit to match the seat position >later. Sometimes I add 20 lbs of baggage at a location which balances and AC. Note the EYEPOINT will be at the same location as the pilot's seat, but perhaps 2 ft higher. >The flight performance was still not what anyone would call >home about. >>I decided to look at the Cessna 172 to see how it might >differ from my aircraft configuration. >The Reference Datum is 3.6,0,0 >The Wing Apex is -2.4 Long, 0 Vert >The Empty Weight is -3, 0, 0 With it's rectangular wing, the LE would be at the 'Apex' setting. The 172 chord is about 4.8 ft. 4.8/4 = 1.2 ft. So, the 25% MAC (CoL) would be at 3.6 - 2.4 - 1.2 = -3.6 ft. That came out right!>I did not expect the Center of Gravity (which I believe is >what Empty Weight location represents), to be equal with the >wing height. It would seem to me this would make the >aircraft top heavy. I just realised that yesterday. The aircraft.cfg file puts the vertical CG at the level of the wing. I'd suggest trying: "3.6, 0.0, -2.0" for the reference_datum. Then, the prop will be below the wing and so will the vertical CG's of the passegers, etc. However, the vertical CG isn't that critical. The main effect appears to be that when it is below the wing, the AC tends to level it's wings more. While for a SEL low wing AC setting the vertical CG say, 1 ft high makes the wings level less. I then adjust the 'Dihederal Effect' parameter in REC 1101 of the AIR file to get reasonable wing leveling. Which generally means 'not too much'. However, I've also suspected AC that jitter on the runway, especially when braked, have the vertical CG too high. This is a problem with many smaller AC. While one can adjust the LG 'static compression' values and 'damping factor' and sometimes improve this jitter, it would be nice if setting an appropriate vertical CG took care of this problem. That would be at about the middle of the fuselage in many AC, but making it lower would be acceptable if it fixed the jitters. I forgot if moving the vertical CG down fixed any of my AC, but the DF C177 didn't have jitters, so I may not have tried to move the CG. Actually, the Visual Center was set near the vertical CG (not on the wing vertically), soI would set the 'Wing_Apex_Vertical' setting to +2.0 ft. The prop shaft was not exactly on the vertical CG.[GeneralEngineData]engine_type=0Engine.0=-1.75, 0.00, 0.70 Note I set it 0.70 ft above the reference datum, which was at the vertical CG. Based on a C177 drawing. This means increasing power will also tend to pitch the AC down slightly. It will speed up a bit relative to idle power IAS. Typical of many real AC. As an option, one could set this line instead://Engine.0=-1.75, -0.40, 0.70 //Set Prop sliders to 1/2 or more In this case the prop shaft is set 0.40 ft to the left of AC lateral center. With prop sliders set to include prop effect, increasing the throttle changes rudder trim much less than if the offset were not present. One needs about 1.2 deg of right rudder trim, but hardly has to change it at all, wheter in a climb or cruising. The Real C177 may set the engine mounting so the shaft is a couple of degrees CCW of the AC body. Setting a lateral offset is similar. But, if the prop sliders (mainly the "P-Factor") are set at zero the offset causes the AC to yaw to the right with increased throttle. 50% to 100% works fine. >However, if you look at the high wing as suspending the >weight of the aircraft like a pendulum from a string, it may >not appear so strange. That's called the 'pendulum effect'. It is not the correct explanation for the change in wing leveling realative to wing location, but it is easy to remember. I haven't tried to understand the 'correct explanation'.>Looking at the H Tail, which is very influential on >longitudinal stability (pitching), I see that the Cessna 172 >H Tail Apex Vertical is zero. >To me, this places the Apex Vertical at the same height as >the Ref datum and the main wing. >>H Tail Apex Vertical = Wing Apex Vertical = 0 If it had an effect without FSEdit, the appropriate height would be at about the vertical center of the fin. Half way from the bottom to the top. Or, a bit lower, since most fins are wider on the bottom than the top.>It is clear from the visual model that the C172's main wing >is higher than the horizontal stabilizer wing. Odd. But >apparently the zero location works in flight. FSEdit should should set the new parameters in the AIR file in a way to account for the high tail. But, messes up so much it isn't of much use. The FS C182's were set with many realistic parmeters at least as far back as FS98 (probably based orignally on Bruce Artick's verson) and the complete AC, including the location of the tail is more or less accounted for. The FS2K2 C172SP was much worse. >If you look at the SDK the H Tail Apex Vertical is shown >located at the middle of the H Tail (vertical stab on the >example). >>The aircraft seemed to handle a little better. I tried >setting the H Tail further behind the aircraft for more >leverage, maybe more stability, but no improvement. So I >went back to the original distance. Placebo effect!>Next, I decided on a hunch to look into the wing incidence. >The aircraft is controllable, but still seems to exhibit too >much pitch instability and pitch up on takeoff. The >incidence was one of the few figures I had on the original >aircraft, 3 degrees. >>So I got the Tiger Moth out and did a full throttle takeoff, >yes there is a little nose up tendency, but it will climb >hands off with any pitch oscillations self-damped. >>I tried setting incidence to zero. The takeoff began to look >more like the Tiger Moth. Very good! I said before: don't set Wing Incidence to the full physical value. ;) Though, TBL 404 can be shifted to have the same effect as changing incidence. But, changing incidence is easier. "cruise_lift_scalar=1.0" could also be dropped to perhaps 0.8 to have a similar effect to reducing incidence.Don't assume that increasing "pitch_stability=1.0" to a higher value will give more desirable pitch stability. An overdamped AC takes forever to trim. Also, the MoI's should be at least approximately correct. At least for a start.>The aircraft still exhibited the same pitch instability. >While digging through the air file I discovered what I think >is the real culprit behind all these problems. The H Tail >incidence had been naively set to 3, from the original >Cessna configuration. I set it to -2.0 according to your >annotations on the aircraft.cfg file and the aircraft is >much more pitch stable on takeoff and climb to cruise. Oh oh. Looking at my commented aircraft.cfg I see I had used -2.0 deg for horiz_tail_incidence. But, I'd reversed the signs of two REC 1101 parameters in the AIR file. Normally, that is 0.0 to +3.0 degrees. It depends on the AIR file. Regardless, the idea is to set the sign of that parameter to push the tail down. If changing it to -2.0 degrees helped pitch stability, then that sounds reasonable to me. I don't completly understand the advantage of a down force on the tail. I think in part it means the CG then has to be more forward to balance at a given trim. >In fact flight characteristics are nearly where I expected they >should be, not perfect on takeoff and I have not tried >landing, but I think it is where I want it for now. >So anyway, I jumped the gun with my previous message. You >were right to suspect something else was wrong. >Steve There are a large number of things which interact. However, if one starts with a good, similar AIR file, then getting the CG, LG, etc. in the correct places goes a long way to having a good flight model. >PS: The landing seems less stable than it was. Tends to sink >too quickly on final. This aircraft always seems difficult >to line up with the runway (at least using the keyboard---my >yoke is disconnected). If it is too stable in pitch with a forward CG, it would seem it might also be too stable in Yaw. Making it harder to turn rapidly. I just thought of this. However, you might try reducing "roll_stability" (which also requires reducing "aileron_effectivness" or going the opposite way. Also, try adjusting "yaw_stablity". The problem is that unless the basics are roughtly correct, changing one or two parameters to improve one characteristic may make others worse. I always set CG so it is near 25% MAC since that is normally optimum.Ron

Share this post


Link to post
Share on other sites

My original message had this line:htail_incidence=-2.0 <<0.0 to -3.0 may improve pitch stability For most MS FS2K2 AIR files, that would behtail_incidence=2.0 <<0.0 to 3.0 may improve pitch stability The sign depends on the signs of two H. Stab parameters in REC 1101 of the AIR file. The idea is to set the incidence of the H. Stab so it pushes the tail (or CG with MS AIR files) down a bit. Requiring nose down trim to compensate. Often, AC seem stable in static pitch with this angle set to zero. I think I got into this more in my Dash 8 message.Ron

Share this post


Link to post
Share on other sites

To Avcomware,Some comments on your previous messages:< Hi, Ron.Thanks again for your information and dedication. While other factors are important in the Feel of a simulation, for Real, Serious, and Knowledgeable Pilots the FLIGHT DYNAMICS has to be and IS the most important factor in a simulation. It

Share this post


Link to post
Share on other sites

I hear very litle discussion of how much fidelity to real counterparts the aircraft have in flight. It appears that if it has a pretty model and flies reasonably well that is enough. I'd like to know that if I succesfully fly the simulated version I would have a reasonable chance of flying the real aircraft. It would be nice if these differences were collected somewhere.Steve

Share this post


Link to post
Share on other sites

> I'll try changing it some time. Are you sure you didn't >use FSEdit to set the new AIR file records from 'Sweep', >'Dihederal', etc? >> Ron Ron, I refused to use FSEdit given my experiences in the past. :)You have a good one. Julian

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