COG or COM understanding? Flight dynamics

[Julio Jerez]

like I said before you may have teh defulat build in coefocend of drag teh each newton body has.
you need to make sure it is set to zero for both linera and angula velocity.

you can detrmine if thsi is true, but hacking you simulation to produce zero drag force just when you go to gligh mode,
and mesaure the chnage in linear speed.

when you are in a glide mode with zero drag, then enregy conservation soudl hold.
besaically the total energy before and fate a time stape soudl be almost identical.
if you gain altitude you sodul lose speed, if you loose altitude you gane speed if you turn and gane angula velocity you sodul lose linear velocity and so on.

[andygib]

Hi Julio, from what i can see the only default drag coefficients leadwerks impliments is linear and angular "damping", i have these set to both set to 0.

When i am flying and i set thrust to 0, the nose dives and it starts to decend and loose altitude but the speed stays the same and is low, yet lift is perpendicular to airflow, so when plane is diving there is low lift holding it back, so must be drag.

here is a another short video, the jerking is from the video capture.
,
http://www.youtube.com/watch?v=mRpK1f6h ... e=youtu.be

Thank you
Andy

[Julio Jerez]

my guess is that if the plane nose down the moment you let go of teh trust, then maybe the trust force line of action do no pass trugh teh center of mass and it is generation some troque that is conter balaced but the aerodynamics turque

you either make the thust lien of action pass though teh plane origin or you nee to know what the truqoe is and when you glide you nee to change the angle of attack and teh control surace to zero out that torque.

[andygib]

Well, for trust im simply adding it to the planebody in local space.
Z is body front

Code: Select all

thrust = Vec3(0,0,thrustValue)
F = AddVec(finalForce,thrust) //finalForce i have transformed to plane local coords.

Do i need to add the thrust at a specific point?

Thanks
Andy

[Julio Jerez]

no, that should do it, unless the game engine is considering that the center of the visual mesh and the center of mass o fteh body are different and it is calculatinf the torque genertaed by that force.
you need to find out out iof that is the case

[andygib]

I will ,do but im sure it is right that if i set thrust to 0, the nose of the plane will go down. This is correct isnt it? Put problem is speed doesnt increase in the dive.

Andy

[Julio Jerez]

the only reason that teh plane will nose down is if the truth force was producing some turque.

at forze that act on the center of mass can no possible rotate a body.

[andygib]

Hmm its a shame ive got this far and cant seem to crack this next bit.

If all forces and torques are correct should the "glide/fall" speed automatically work or do i need to do more? When an engine is shut off, you can push the nose down and gain speed, but this isnt happeneing.

It is like there is no momentum, if i flight level and then turn of thrust, it slows down VERY fast?

Thanks
Andy

[Julio Jerez]

if i flight level and then turn of thrust, it slows down VERY fast?

The only reason for that is that there must be a damping value somewhere, belive me that is the reason.
try setting Gravity, and all force to zero when the vehicle have some linear speed, see if it slows down, if it does then there is some damping value being applied.

[andygib]

Ok, ive done that it doesnt slow down. It keeps going at a constant speeed?

Thanks
Andy

[andygib]

Ok, i have found some new drag coeficients for the cessna 172 (that is what im basing this on) and they seem to work much better.

Andy

[Julio Jerez]

do not try to fix teh Bug by adding new feature, a new drag value will not fix anythong if ther is a bug, you must find teh Bug first, then you add that new funtioality.

Ok, ive done that it doesnt slow down. It keeps going at a constant speeed?Andy

Ok so if this is the case you must have one of the lift forces generation a force that act to slow the plane. keep doing the same.
set the engine, gravity and rear wing to zero, and only leave teh lifg force generate bu teh left and righ wings. see if the result are as you expect, and keep doing tha until you fing teh Bug

[andygib]

O, ill try this now.

One question, should all rotations of the plane be around the center of mass?

Edit: I have gone though setting all airfoils and when any airfoil i leave ON slows it down a bit due to drag, are you saying i should set ALL drag to 0 for the test?

I think its the drag that is the problem, if i have big coeficents, plane flies to slow, if i have low coeficients then my plane flies, but when i turn it will keep turning, there is no angular drag.

Andy

[Julio Jerez]

yes the body will alway rotate around the center of mass.
passically you apply a netforcw, and a net torque

net force act on teh center of mass and cannot rotate the body.
net torque also acct o fteh centr of mass but can no move teh body it can only rotet teh body.

therfore as you add you force compunde build up, you cna inspecte net fore and net torque to find out whin of thsoe is generated wrong.
if the vehicle slow down it must be because some of the non zero force that are appliyed have a compound tha t act to reduce the linear velocity.

now you you nee to do is enable gravity, and one by one apply the airfoils.
I would start by the main wings, because the have simetry, see if the plane glide, then keep addin teh oteh elements, real alerons, rudet only you detrmine whic one is unstable and fix it.

[andygib]

Its very strange, all airfoils slow plane down when i set gravity to 0 becasue each airfoil is calculating drag. This is what i belive is causeing the glide issue, there is to much drag?

I will try to show a video to show. The Plane flies really good with thrust, when i put thrust to 0, the plane slows very fast and just drops to ground, no "momentum" or "glide".

Is there a simple equation i can use for the coeficient table lookups?

Also do i calculate AOA for each airfoil (rudders and tail wing?) or just AOA for airplane?

This is what im currently using:

Code: Select all

--Planes coeficienct lift tables (based on a cessna 172)
plane.airfoil.cl1[1] = {0,0.18,0.3,0.5,0.7,0.82,1.0,1.2,1.5,1.2,1.0} --left wing
plane.airfoil.cl1[2] = {0,0.18,0.3,0.5,0.7,0.82,1.0,1.2,1.5,1.2,1.0} --right wing
plane.airfoil.cl1[3] = {0,0.18,0.3,0.5,0.7,0.82,1.0,1.2,1.5,1.2,1.0} --tail wing
plane.airfoil.cl1[4] = {0,0.18,0.3,0.5,0.7,0.82,1.0,1.2,1.5,1.2,1.0} --vertical wing
plane.airfoil.cl1[6] = {0,0,0,0,0,0,0,0,0,0,0} --fuselarge

--Planes coeficienct drag tables
plane.airfoil.cd1[1] = {0.2,0.2,0.3,0.4,0.6,0.7,1,1.2,1.4,1.7,1.9} --left wing
plane.airfoil.cd1[2] = {0.2,0.2,0.3,0.4,0.6,0.7,1,1.2,1.4,1.7,1.9} --right wing
plane.airfoil.cd1[3] = {0.2,0.2,0.3,0.4,0.6,0.7,1,1.2,1.4,1.7,1.9} --tail wing
plane.airfoil.cd1[4] = {0.2,0.2,0.3,0.4,0.6,0.7,1,1.2,1.4,1.7,1.9} --vertical wing
plane.airfoil.cd1[6] = {0.2,0.2,0.3,0.4,0.6,0.7,1,1.2,1.4,1.7,1.9} --fuselarge

EDIT: Ok i have done some extensive testing today and here is my conclusion: If i have the drag coeficients normal then the plane flies great, but at 0 thrust the plane stop in the air and falls to ground. If i lower the drag coeficients then the plane glide, but when i turn the plane is uncontrollable (like flying in space).

Andy