In my many years of FS experience I've always seen these MOI values in all FS9 aircraft.cfg files but have never touched them, simply because I've never understood what they were there for ......... untill I downloaded a very well made model which bounced around violently on the ground while taxiing, taking off and landing, even directly after starting the engines.The plane's author said he had had no such problems during his testing phase and he advised me to increase the three MOI (pitch, yaw and roll) values until the problem disappeared. I tried this but it didn't work, most probably because I didn't increase them far enough. So I copied/pasted the three lines (containing values which were very much higher) from another similar looking aircraft and that seems to have solved the problem.The plane now handles quite well .......BUT ...... I'm one of those guys who has always wanted to know why/how something works and especially why/how something does NOT work, when it should. So ........Judging only by the name "Moments of inertia", I assume that when these values are (far) too high, the plane will fly like a brick, i.e. will be extremely sluggish while, when too low, the plane will be too sensitive to steering input. E.g. if the MOI roll value is too low for a twin engined plane, it will unrealistically snap into a roll or turn and not take the engine weights into account. Similarly, if the MOI pitch value is too high, it will be difficult to e.g. flair just before landing. But these are only assumptions on my side.Does anyone know how these values should be realistically calculated or is it just a case of trial and error, specific only to simming ?Is there any relation between the aircraft's empty weight and some entries in the [Flight_tuning] section, with these MOI values ? E.g. The aileron, elevator and rudder effectiveness entries.I hope one or more of you technical experts can help me get to the bottom of this "problem".Thanks in advance for any trouble taken.Hans

From the SDK:

Moments of InertiaA moment of inertia (MOI) defines the mass distribution about an axis of an aircraft. A moment of inertia for a particular axis is increased as mass is increased and/or as the given mass is distributed farther from the axis. This is largely what determines the inertial characteristics of the aircraft. The following weight and balance parameters define the MOIs of the empty aircraft, so the values should not reflect fuel, passengers or baggage. The simulation engine determines the total MOIs with these additional, and variable, influences. The units are slugs per foot squared. Omission of a parameter will result in the use of a default value set in the .air file, if one exists.These values can be estimated with the following formula:MOI = EmptyWeight * (D^2 / K) Where:

		   Pitch				   Roll					Yaw D =			  Length (feet)		Wingspan (feet)   0.5* (Length+Wingspan)  K =			  810					 1870				   770

This formula yields only rough estimates. Actual values vary based on aircraft material, installed equipment, and number of engines and their positions.

Thanks n4gix for your quick response. I'll give the formula a try with different aircraft and will post my findings.Hans

A couple of notes:1 - MOI values are based on an empty aircraft in the .cfg file.2 - The formula ACES provides isn't very accurate. :(

2 - The formula ACES provides isn't very accurate. :(

Presumably, that's why they wrote: "This formula yields only rough estimates." :(

Yes I agree, the formula is not very accurate because large variations between the "calculated" MOIs and any manually "tried out" ones (above certain minimums), do not seem to make any percievable difference in a flyable plane's handling characteristics. However, some of my AI planes "wag their tails" just before beginning their take off rolls and during taxiing, so I suspect that their yaw MOIs could be too low. OK, for the commedians amongst you, yes, they could be just happy to see me but I'm definitely going to have a go at increasing their yaw MOIs to see if this would dampen their wagging.I will report my findings.Hans

This is from a article posted by Tom Goodrick on his site. The post covers a multitude ofitems. Here is the relevant info regarding MOI. Just use the moments for the aircrafttype nearest yours and grind out the MOI. The method below uses maximum takeoff weightso for FS you could substitute empty weight without losing very much. "Dr Jan Roskam, has published an 8-volume set of texts on aircraft design. He has taught courses in aircraft design for some time at the University of Kansas and has developed a method for estimating MOI's that uses only basic aircraft weight and dimensions combined with a set of coefficients of radii of gyration for several specific aircraft. We simply determine which aircraft is similar to the design we are concerned with, take the coefficients for each axis and the approriate dimension and calculate MOI's. The formulae are: Ix=(W/g)*(Rx*b/2)^2 Iy=(W/g)*(Ry*d/2)^2 Iz=(W/g)*(Rz*e/2)^2 where g=32.2 (gravity acceleration) W=max takeoff weight (MTOW, lb) b=span (feet) d=length (feet) e=(b+d)/2 and Rx, Ry and Rz are chosen from the table for the aircraft type. Some of Dr Roskam

Thanks for all your input guys.Tweaking MOI values for some "tail wagging" AI aircraft, does not seem to work in any way, even with grossly exaggerated values. In that respect I'm back to square one.Ideas anyone ?Hans

..The method below uses maximum takeoff weight so for FS you could substitute empty weight without losing very much. Jim

I think you could lose quite a lot. Moment of Inertia is defined the product of mass and a distance squared. If the aircraft carries most of its fuel in the wings then the difference in roll MoI between an empty and fully-fueled aircraft would be signifcant because of the distance factor. It would be much more significant that the difference for the same aircraft between empty and with a full payload. This is because the payload is in the fuselage and so the distance factor would be much smaller.

I think you could lose quite a lot. Moment of Inertia is defined the product of mass and a distance squared. If the aircraft carries most of its fuel in the wings then the difference in roll MoI between an empty and fully-fueled aircraft would be signifcant because of the distance factor. It would be much more significant that the difference for the same aircraft between empty and with a full payload. This is because the payload is in the fuselage and so the distance factor would be much smaller.

The MOI of all aircraft change as they burn fuel. The moments given above are for specific aircraft configurations.FS modifies the MOI as the aircraft is loaded, and when it burns fuel. The weight contribution to MOI is close tolinear I've read. Roskam's moments will allow you to determine very reasonable MOI which is all you need in FS.I have found most MOI are too high even in some commercial packages.Indeed, you are correct, the MOI varies directly with weight and with square of the distance. Distance doesn't changewith a given aircraft, only weight, and that is linear which is why you can sub empty weight for gross weight. The onlydifference is you are calculating the empty weight MOI, not gross weight MOI. FS does the rest otherwise you would have to continuously change your MOI in flight which is not feasable.One thing I should mention, the above data are moments, not MOI. You have calculate the MOI from the moments.To do this you need reasonable moments, and Roskam provides them.Jim

MOI doen't vary directly with aircraft weight.Take the default FSX B747 as an example. From the aircraft.cfg file, the empty weight is 394,008 lb and the maximum gross weight is 875,000 lb..The empty roll MOI is 13,352,310 slugs/ft2. From data given in the fuel and station.loads sections the additional MOI due to load is 4,268,094 slugs/ft2 giving a total roll MOI of 17,620,446 slugs/ft2. Thus the empty weight is 45% of the gross weight and the roll MOI at empty weight is 76% of the MOI at gross weight.This shows that MOI doesn't scale with weight. Scaling the gross MOI by 45% (weight) gives an empty MOI of 7,936,007 instead of Mircrosoft's value of 13,352,310 slugs/f2 - a 40% difference.Incidentally using Microsoft's data for the B747 results in a value of 0.241 for Rx. This is pretty close to the values given in the table for jet airliners. Maybe Microsoft didn't do just a bad job after all!

See if I can sum this up a bit more simply. The originalquestion was to how to get good MOI for FS.FS modifies the MOI by multiplying the empty wgt MOI bythe difference in weight - current wgt/empty weight. Thisis a linear method. If MOI deviated far from linear then thiswould not work and planes would not fly in FS very well.The problem then becomes one of getting a reasonable setof MOI for the plane. Roskam gives you the moments to dothis for gross weight. Therefore, in FS you need a set of MOIthat will scale linearly to give the Roskam MOI at gross weight.This is why you can substitute empty weight in the formula asthose MOI will, by definition, scale linearly to the gross wgt MOI.This is what FS does with them.To be accurate, we are not dealing with weight, but mass. W/Gis the mass (slugs). But since we are dealing with the ratio of themwe can use weight as well. In addition to Roskam, there are othersources such as DATCOM, Dr. Phillips and Lifting Line theory forthose interested. I find Roskam very simple to use, that's all.Furthermore, to produce a flight model that indeed behaved justlike the real thing, one would need to have the sim model theMOI as it changes in real life. Any one interested in this I'd pointyou to the AvHistory.org forum. They specialize in producing 1%models for CFS. These are aircraft that are intended to fly in CFSto within 1% of the real plane. They have good discussion on allphases of flight dynamics in CFS/FS. But at present, I know of noone using anything other that the above mentioned methods of calculating MOI. See http://avhistory.orgLastly, the air file should be built around good MOI. Typically weare left with modifying the MOI after the air file is established andredoing air files is more often than not a labor of love affair.Jim

MOI doen't vary directly with aircraft weight.Take the default FSX B747 as an example. From the aircraft.cfg file, the empty weight is 394,008 lb and the maximum gross weight is 875,000 lb..The empty roll MOI is 13,352,310 slugs/ft2. From data given in the fuel and station.loads sections the additional MOI due to load is 4,268,094 slugs/ft2 giving a total roll MOI of 17,620,446 slugs/ft2. Thus the empty weight is 45% of the gross weight and the roll MOI at empty weight is 76% of the MOI at gross weight.This shows that MOI doesn't scale with weight. Scaling the gross MOI by 45% (weight) gives an empty MOI of 7,936,007 instead of Mircrosoft's value of 13,352,310 slugs/f2 - a 40% difference.Incidentally using Microsoft's data for the B747 results in a value of 0.241 for Rx. This is pretty close to the values given in the table for jet airliners. Maybe Microsoft didn't do just a bad job after all!

Fascinating. What you have shown for FX is that the scale factor isn't 1. You would need other points to show it not to be linear.I'm using FS 9 and this is what I get for a DC9-31.Roskam empty gross roll 226,437.0 430,133.0 pitch 819,294.0 1,637,769.0 yaw 1,043,639.0 2,086,234.0FS empty roll 1,087,183.0 pitch 2,675,439.0 yaw 2,937,024.0I don't know how to see what FS would give for gross weight MOI but it would have to be higher than empty wgt. As youcan see FS empty wgt MOI are very much higher than Roskam gross wgt and the ratios between the MOI are vastly different.You would expect a DC9 to have a low roll MOI compared to pitch which Roskam gives.I cannot comment on FSX. FS9 however appears to be quite at odds with realistic MOI. You don't perchance have the FSXnumbers for DC9 do you? It would be interesting to compare.Jim

Fascinating. What you have shown for FX is that the scale factor isn't 1. You would need other points to show it not to be linear.I'm using FS 9 and this is what I get for a DC9-31.Roskam empty grossroll 226,437.0 430,133.0pitch 819,294.0 1,637,769.0yaw 1,043,639.0 2,086,234.0FS emptyroll 1,087,183.0pitch 2,675,439.0yaw 2,937,024.0I don't know how to see what FS would give for gross weight MOI but it would have to be higher than empty wgt. As youcan see FS empty wgt MOI are very much higher than Roskam gross wgt and the ratios between the MOI are vastly different.You would expect a DC9 to have a low roll MOI compared to pitch which Roskam gives.I cannot comment on FSX. FS9 however appears to be quite at odds with realistic MOI. You don't perchance have the FSXnumbers for DC9 do you? It would be interesting to compare.Jim

It can't be linear with weight. A pound of fuel in a B747 wing tip tank, which is 74 ft from the centre-line, clearly has a greater contribution to roll MOI than a lb of fuel in a tank on the centre-line.Sorry, but I don't have figures for the DC 9.

MOI doen't vary directly with aircraft weight.Take the default FSX B747 as an example. From the aircraft.cfg file, the empty weight is 394,008 lb and the maximum gross weight is 875,000 lb..The empty roll MOI is 13,352,310 slugs/ft2. From data given in the fuel and station.loads sections the additional MOI due to load is 4,268,094 slugs/ft2 giving a total roll MOI of 17,620,446 slugs/ft2. Thus the empty weight is 45% of the gross weight and the roll MOI at empty weight is 76% of the MOI at gross weight.This shows that MOI doesn't scale with weight. Scaling the gross MOI by 45% (weight) gives an empty MOI of 7,936,007 instead of Mircrosoft's value of 13,352,310 slugs/f2 - a 40% difference.Incidentally using Microsoft's data for the B747 results in a value of 0.241 for Rx. This is pretty close to the values given in the table for jet airliners. Maybe Microsoft didn't do just a bad job after all!

All my Aircraft.cfg section shows for empty weight moi is the that are input there or are read from the air file.There is no mention of what those moi are from the station loads. Where do you find that? I'm in FS9, not FSX.For a 747-200 Roskam gives: empty:roll 12,037,944 gross wgt: roll 25,921,179 pitch 18,669,671 pitch 40,201,704 yaw 29,777,080 yaw 64,119,467Roskam's gross weight (833,000 lbs) numbers are probably pretty good.Extrapolated back gives the empty. The roll is very close to yours. I'musing an empty weight of 386,850lbs, using yours would give me an emptywgt roll moi within 8% of yours which looks very close to me. How wereyour moi calculated? These look like very "reasonable" numbers to me. Thedifference may very well be due to using the CV 880 moments for the 747.This one is used as both fit the profile of 4 engine jet liners and one wouldexpect small differences. So again, how were your MOI determined?For this example, the plane I used is the FS9 POSKY 747-200 and POSKY lists theMOI as: empty weight 386,850 empty: roll 18,155,743.0 pitch 17,600,873.0 yaw 19,151,035.0Quite a difference no?By the way, disregard my FS numbers for the DC9. FS9 doesn't calculate aempty wgt moi. If none is specified in the aircraft.cfg file it uses whatever isfound in the .air file which is not documented anywhere on how it was obtained.Otherwise it sets it to zero.Jim