Thanks Milton ..Bob

Barry,My preference is always to use real numbers. I have found that when set up correctly, these should give the expected response. If not, then there is another incorrect influence on the response.Recheck all your area sq. footages to ensure they are entered correctly. Then take a look at section 1101 and check lift, roll and pitch numbers for reasonableness (against well flying models of similar characteristics). Also remember to check the dihedral effect in this section.If the dihedral should be 6.25. then you should make that number work for you, and not by forcing it.Have fun,Milton

> Then take a look at section 1101 Not sure what you mean by "section 1101" Thanks Barry

In the air file, there are categories of influences to properly define your aircraft and its dynamics.Using freeware available, you can edit the detailed parameters to better control FD's. Many of these are not available in the aircraft.cfg file. If you are new to this, then I would suggest heading over to Abacus and downloading some of the excellent reading to help you better understand the world of flight dynamics. This is of course, introductory level stuff.This will certainly take a lot of time and study and can become an obsession in itself. ;-)A better option may be to shoot your flight model with all the technical data off to one of the experts (I am not one).Milton

>>The tendency to return to level flight from a bank sounds >>more like a supercharged dihedral effect. >>>>Check the air file section 1204 and make sure wing dihedral >>is less that say ... 4.00 depending on your aircraft design. >>>>Milton >>Thanks Milton >>The dihedral for this aircraft was 6.25 -- I had already >lowered it to 4 before you posted and that had improved >things a lot. I will try it a bit lower and see how it works. >>Barry Did you increase the value of roll_stability?? Example:[flight_tuning]roll_stability = 1.400 // 1.200 // 1.000After experiments with MOI and dihedral .. systematically increase roll_stability in steps of 0.2 on up to 1.80 until you get it damped. If the roll `overshoots'.. then it likely needs more damped.. so increase the roll_stability scalar. generally speaking: * 0.0 is no damping.. continuous oscillation * 1.0 critical damping.. will make one `cycle' * >1.0 overdamping.. will damp in less than one `cycle'Gregory Abbey - Edwinnengineering