You are on page 1of 8

4. Longitudinal Stability and Trim.

4.1. Aircraft StaticsForce Balance


Consider the case of an aircraft in steady ight. Note that this is a statics problemthough there is motion, it is non-accelerating so that one may dene a system of coordinates for which the system is static. If the aircraft is treated as a rigid body, standard equations describe the equilibrium of the aircraft:
X X

F = 0 (4.1)

M = 0

Since both force and moment are three-component vectors, there are six of these equations. Because of the symmetry of most aircraft, the equations may be decoupled into two sets of three equationsconsider the set describing the aircraft equilibrium in the longitudinal plane. In particular,
X X X

Fx = 0 Fz = 0 (4.2)

My = 0

At this point, a set of reference axes must be dened in which to describe the vehicle equilibrium. First, the system must actually be static, so the axis system must move (with constant velocity) with the aircraft. It is customary to dene the origin of coordinates to be located at the aircraft center of gravity. There are several (at least four) reasonable ways to dene the coordinate directions. Perhaps the simplest of these is to let x point directly forward (out the nose of the aircraft) along the waterline. The y -axis extends out the right wing, and, to enforce right-handedness, z then points generally downward. These axes, termed body axes, are xed relative to the airplane; therefore, their orientation must change relative to the velocity vector with changes in the airplane angle of attack. An alternative is to use wind axes, or aerodynamic axes, which are xed relative to the velocity vector. Most references in use wind axes to develop the static equilibrium equations. There are advantages and disadvantages to each, and (clearly) the solutions obtained from each must be the same. The following gures illustrate the two axis systems:

horizon L xb V zb Body axis system W T D

V xw

horizon D
W

zw Wind axis system

Now, using the free-body diagrams above, the equations for the two systems can be derived: Body Axes: P Fx = 0 = T D cos W sin + L sin P Fz = 0 = W cos L cos D sin P My = 0 = Wind Axes: P Fx = 0 = T cos W sin D P Fz = 0 = W cos L T sin P My = 0 =

(See if you can show that the equations are equivalent.) It is immediately noticed that the equations in the wind axes seem simpler since each force balance has only three terms, whereas, for body axes, there are four terms in the xequation. Also note that the angular dependence seems to be dierent for the two cases, though the angles are related through the expression = At this point, it is dicult to tell whether it is more desirable to use or as a variable. Before discussing the moment balance, analyze the force equations in more detail. First, the unknown variables must be determined. What those are depends on the statement of the problem, but a typical situation may be that the aircraft properties are known and a certain ight path and velocity are desired. For this case, W and are known variables (along with v ), and we do not know L, D, , or T . If we choose to work in body axes, we could consider to be unknown as well. We are thus left with four (or ve) unknowns but only two (or three) equations. The additional two equations could be simply 1 2 v CL a S 2 1 2 D = v (CD0 + K2 a )S 2 L =

(4.3)

Note that a means aerodynamic angle of attack whereas will now refer to the aircraft angle of attack. If the wing is mounted on the aircraft at some angle, iw , the two are related by a = + iw . For the situation discussed thus far, these make up a complete set of equations since no further unknowns are introduced. Many of the performance calculations for an aircraft are made using the force trim equations only. Some values have been calculated using the wind-axis set of equations for the following aircraft data: 2

W (lbs) U (kts) (slugs/ft3 ) CD0 CL S (ft2 ) A R e 132,000 675 .000858 .025 5.026 1200 8.0 .75 These data represent an aircraft in the size range of an MD-80 transport. Two aircraft were considered: the rst had the wing attached to the fuselage such that the zero-lift line and the fuselage centerline were parallel. The second wing is set on the fuselage at an angle of 0.05 radians, or 2.865 degrees. 0 0.0222 -0.0222 0 0.0222 -0.0222 T (lbs) D (lbs) L (lbs) iw 9,829 9,769 130,911 0.1111 0 12,759 9,748 130,558 0.1107 0 6,897 9,786 131,202 0.1112 0 9,817 9,799 131,398 0.0614 0.05 12,743 9,786 131,188 0.0612 0.05 6,888 9,807 131,544 0.0615 0.05

Question: What is the major eect of climbing or descending on the forces and angle of attack? Explain the results. Why is the lift force smaller than the aircraft weight? A sailplane, or glider, has no thrust. For such a case, the force-balance equations in the wind axes become
X X

Fx = 0 = W sin D Fz = 0 = W cos L

(4.4)

and the expressions for L and D remain the same. Now the ight-path angle cannot be speciedit will depend on the characteristics of the airplane. It is interesting to investigate the behavior of the aircraft whose characteristics are described above in the event of a thrust loss. The gure shows the sink angle ( ) and the sink rate (V ) as functions of the forward speed. Now compare the above results to those obtained using standard performance methods. Using equations 4.4, we obtain sin = v 2 (CD0 + K2 D a) = W W/S

Strictly speaking, this is a transcendental equation since v ( = U 2 + V 2 ) depends on the ight-path angle according to tan = V /U . In addition, depends on the ight-path angle since 2L a = 2 (4.5) v CL S and L, as well as , can be considered dependent on . However, since is generally a small q angle (at least we usually hope so) we make the approximations that v = U 1 + (V /U )2 U and tan so that U 2 (CD0 + K2 a) (4.6) 2W/S 3

Figure 4.1: Sink angle and sink rate vs forward speed. In addition, L W so that the angle of attack can be approximated as 1 cos = a = 2W/S U 2 CL (4.7)

(4.8)

Using this expression for leads to an approximation for the ight-path angle given by = CD0 2 2W/S 1 U + 2W/S A R e U2 (4.9)

Solving for the minimum sink angle (maximum ) gives = 0.0728 radians (-4.17 ) at a forward speed of U = 362 kts. Because the angles are very small, the approximate method gives quite acceptable solutions. Question: How would you nd the minimum sink rate and the forward velocity at which this rate is achieved?

4.2. Aircraft TrimMoment Balance


The solutions to the longitudinal force-balance equations may be quite adequate for determining such performance values as maximum range, maximum speed, etc., but a problem

Lw

V xw

horizon

MACW
W

LH MACH

zw
Figure 4.2: Forces acting on an aircraft shown in wind axes. arises when more detail is required. For example, an aircraft may y level at many dierent speeds (within some range) at a given altitude. Since a = 2L U 2 SCL (4.10)

and since the lift, wing area, lift-curve slope and density remain (at least almost) constant, clearly the angle of attack must vary as the speed is altered. It is not possible to determine how this is accomplished from the above discussion. Consider also that the lift may be written as L = LW + LH (4.11) where the ()W refers to the wing and the ()H refers to the horizontal tail. Again, the forcebalance equations implicitly assume that all lift is generated by the wing, and they do not provide enough information to determine how the lift is shared between the two surfaces. A further equation is required. Consider the gure showing an aircraft in wind axes. In general, the thrust line will not be located along the aircraft centerline, but for now assume that it is. From the gure, we can see that
X

My = 0 = MacW + MacH + (xW cos + zW sin )LW + (xH cos + zH sin )LH

(4.12)

where the xs and z s refer to the locations of the wing and horizontal-tail aerodynamic centers. Assume that is a small angle, and that z/x is also small. This would not necessarily be the case for a T tail, but allow the assumption for now. For such a situation, the equation reduces to (in coecient form) CMcg = CMacW + S cCMacH + xW CLW + S xH CLH (4.13)

where S is the ratio of tail to wing areas, c is the ratioof the mean chords, and x is the distance to the surface aerodynamic center normalized to the wing mean chord. The lift coecients depend on the respective angles of attack. Thus, setting the expression to equal 5

zero will indicate how the lift is distributed between the wing and tail for a determined angle and total lift. The moment about the aerodynamic center is constant with angle of attack and is a property of the airfoil section. We will now review our set of trim equations (in coecient form): CL = CLW + S CLH CLW = CLW ( + iW ), CLH = CLH ( + iH ) 2CLW = A RW 0 = CMacW + S cCMacH + xW CLW + S xH CLH

(4.14)

Question: What is the and how does it arise? These, combined with our (now revised) previous set of force-balance equations, results in an alternative complete set of trim equations which can be solved to nd the lift on the wing and tail, separately, and the tail incidence (iH ) necessary to accomplish this. Using the above airplane parameters along with the following characteristics: S c xW xH A RW .1 .2 -0.05 -2.5 8 gives the following results for U = 400 kts: T (lbs) D (lbs) L (lbs) iH LW (lbs) LH (lbs) 0 9,895 9,877 131,381 0.0626 -0.03169 132,706 -1329 The values of lift, drag, thrust, and angle of attack are little changed from those obtained using just the force balace. However, the complete set of equations provides considerably more information about the conditions for trim. Note: A trimmed airplane is one for which the forces and moments balance. An aircraft must be trimmed to y in a steady condition. CLH 4.0 iW (radians) 0.05

4.3. Longitudinal Static Stability


Equation 4.13 can be solved for given aircraft parameters as a function of . Using the values chosen or computed above, the value of CMCG vs angle of attack is shown for the baseline case in the gure. Note that the moment goes to zero for an angle of about 3.5 which corresponds to that computed using the full set of equations. At any other , the moment is non-zero, indicating that the aircraft is not trimmed and will not remain in steady ight at this other angle. The other lines on this plot show the same information for the same aircraft but with the center of gravity shifted aft. Each successive shift reduces the magnitude of the slope of the curve. Finally, with xW = 0.125 and xH = 2.35, the slope has changed sign and is positive. For this case, there will never be an angle for which trim is possible since the curve will never cross the horizontal axis. Even the third curve with the smallest negative slope 6

Figure 4.3: Moment about aircraft cg; various cg locations. (xW = 0.075, xH = 2.4) will never have a valid trim point since the angle at which it will cross the zero-moment axis is well beyond the stall. Before going on, one should note that the gure is somewhat misleading. If the aircraft control settings are kept constant (i.e., iH is kept at 0.03169 radians as was required for trim at the baseline) while the center of gravity is moved, the aircraft will not trim in the steady, level condition. This is clear since the aircraft angle for CMCG = 0 is considerably higher for the second case than for the rst, and the lift on both wing and tail would be noticably altered. We must assume, then, that the trim condition has changed (the aircraft is now climbing or descending) or that there is an additional control surface, such as an elevator, which is used to maintain the desired ight condition at the zero-moment attitude. This second explanation is still unsatisfactory since the elevator itself alters the moment about the cg and, therefore, the trim angle of the aircraft. To illustrate this point, consider the second curve in the gure which shows the CM vs curve for (nominally) xW = 0.025 and xH = 2.45. Using these values to perform the complete trim analysis results in an aircraft trim angle of 3.46 , very little dierent from that required in the baseline case. The tail angle for trim, however, changes signicantly to 0.12 (.002065 radians). When designing an aircraft, it is necessary to keep in mind that the cg must be located so that the aircraft can trim at reasonable angles. Several variables, in addition to the cg location, aect the value of the trim angle, including the moments about the surface aerodynamic centers and the tail (or elevator) control setting. But this is not the only information that can be gained by studying CM vs curves. Consider a case where the baseline aircraft is ying in a trimmed condition. As constantly occurs in real ight, the aircraft is disturbed from this condition by a gust, which changes the aircraft angle. If the

angle is increased, the moment about the cg is decreased providing a nose-down pitching moment tending the aircraft back to its trim position. On the other hand, if the angle is decreased by the gust, a nose-up pitching moment results which then restores the aircraft towards its trim angle. This aircraft is said to be statically stable by virtue of its tendency, when disturbed, to return to its trim position. Question: What does the CM vs curve look like for a statically unstable aircraft? The neutral point is dened as the cg location for which the aircraft would have neutral stabilitythat is, the CM vs curve would be represented by a horizontal line. The dimensionless distance between the neutral point and the cg (made dimensionless by the wing mean chord) is known as the static margin. The static margin is positive for a statically stable aircraft. The level of stability (or instability) is determined by the slope, CM /, with negative slope corresponding to positive stability. For neutral stability 2CLW CM = 0 = xW CLW + S xH CLH 1 A R so that the neutral point is located where xW CLH = S xH CLW
! !

(4.15)

2CLW 1 A R

(4.16)

It is important to realize that stability and trim, though related, do not refer to the same thing. It may be possible to trim an unstable aircraft at a feasible aircraft angle. It also may be impossible to trim a stable one. Usually, the aircraft must satisfy both stability and trim requirements, though active control systems can articially stabilize an unstable aircraft.

You might also like