3. SPIN DESIGN TECHNIQUES

3.1 INITIAL DESIGN PROCESSES

The extent to which spinning characteristics are considered in the initial design phase will depend largely on the intended role of the aircraft. For example, a two-seat glider will be required to have an unrestricted spin clearance for training purposes. In this case the designer would obviously start by looking at the spin characteristics of existing designs with an attempt to relate these characteristics to particular aspects of the design. There are several empirical criteria which are useful aids in this process.

One of these was developed by Kerr (2) and is a very simple method requiring only knowledge of the shape of the aircraft and the relative inertia distribution. The aerodynamic rolling moments in a spin of 45^o incidence are claculated using the simple expressions given. As it is a spin recovery criterion the moments are calculated for the anti-spin rudder. The wing contribution to the rolling moment was derived from NACA rotary balance test data using a rectangular wing. The contributions due to the body, fin and rudder are given as a fucntion of the first moment of profile area about the centre of gravity combined with a weighting factor. The relative merits of different fuselage shapes and fin/tailpane combinations are reflected in the assigned value of these weighting factors shown in Table 1.

TABLE 1

Note that an adverse tailplane and fin intraction can produce a propelling moment in the spin hence the negative weighting factors. The extent of this adverse interaction is dependent on elevator position and so this should be included as a variable in the analysis. It was noted before that the wing contribution was derived from isolated wing test data so an obvious omission in this method is the interference between the wing and other components of the aircraft.

The total aerodynamic anti-spin rolling moment is plotted against the ratio of pitch to roll moment of inertia on a graph upon which empirical boundaries of spin recovery characteristics have been drawn. The designer should perform these calculations for a number of existing configurations similar to his new design to refine the location of these boundaries for his own application.

Another empirical method was developed by Bowman (3) as a guide for the design of tails to ensure good spin recovery. Figure 7 shows the results of a series of spin tunnel tests where a number of design changes to the tail of a representative light aircraft were made. The original design, tail number 1, has a low value of the tail damping power factor with correspondingly poor spin recovery characteristics. At the other extreme, tail number 5 has the highest tail damping power factor and gave the quickest recoveries.

Bowman also commented on the influence of wing position on spin characteristics. The higher dihedral effect of the high wing plue the absence of any adverse interaction with the tail because of the higher wing wake leads to some improvement when compared with low wing aircraft.

Both of these empirical methods can also be used as a guide in the flight development of gliders to improve their spinning characteristics as they cover the two common modifications - tailplane strakes and ventral fine. Both os these provide increased damping in the spin causing the aircraft to spin more steeply and so enable an easier recovery.

Much of the design data base for spinning is derived from light aircraft or military fighters and in its application to glider design the following two points must be remembered:

(i) a glider has a much higher ratio of roll inertia to pitch inertia which increases the importance of the elevator as a recovery control and also increases the possibility of the spin being oscillatory.

(ii) as with military fighters the nose of the glider fuselage can influence the spin and so there is additional scope for design improvements. A flat elliptical cross-section which might result from a side-by-side cockpit can provide a propelling moment in a spin. Fitting horizontal strakes can be very beneficial as they

(a) significantly increase the spin damping by retarding the flow around the nose

(b) provide a nose-up pitching moment which tends to reduce the rate of rotation

So, in the design of conventional gliders there is a large amount of data to assist the designer in the development of satisfactory spin and recovery characteristics. However, if a novel configuration is contemplated or if other design requirements lead to a configuration which is likely to have poor spinning characteristics then some model spin tests should be conducted.

3.2 Rotary Balance Testing

The obvious disadvantages of these tests are, firstly, like most model test there are the problems and unknowns of scaling and, secondly, little information on recovery characteristics is produced. However the method has one big advantage in that basic aerodynamic data on the complete configuration and several permutations of partial configurations indicate where any potential problem areas exist.

For example, Figure 8 shows some typical results from a rotary balance test series and even without the computer analysis we can make a number os useful observations. Firstly, the complete model is autorotative with a stable spin mode indicated at Ub/2V = 0.4. We can also see that without the tailplane the damping is high leading the conclusion that there is adverse interference between the tailplane and fin.

A complete test series would include a number of different control deflections so from scan of the results we would get a good indications of the control effectiveness for recovery.

With a complete matrix of rotary balance data - a range of angle of attack, spin rate and sideslip angle - as well as an estimate of the moments of inertia it is possible to solve the equtions of motion for a steady spin.

The solution for the pitching moment equation is shown in Figure 9 and is a plot of angle of attack against spin rate for which the aerodynamic moment equals the inertial moment. It clearly indicates the well-known relationship between fast, flat spins and slow, steep spins.

Figure 10 shows the solution of the rolling moment equation.

The final curve in this series, Figure 11 is the solution of the yawing moment equation indicated by each intersection of the aerodynamic and inertia terms. Only stable solutions, as shown here, represent a spin mode. This particular example has a steady spin with an angle of attack of about 40^o, a spin rate Ub/2V of 0.2 and a sideslip angle of -4^o. From a balance of forces with drag equal to weight we can calculate the vertical velocity and also determine the rate of turn in the spin.

The absence of an intersection of the two curves in Figure 11 indicates that a spin mode does not exist. A shallow intersection or series of intersections is perhaps indicative of an osciallatory spin.

One of the important results in these tests is the effect of controls. For example, in a trainer we would be looking for a strong intersection of the curves with pro-spin controls and a wide separation with anti-spin controls.

With such testing techniqyes it is possible to find out a great deal about the mechanisms of the spinning behavious of any design in a very short time and at a resonably low cost.

3.3 Dynamic Model Test

The concept is very simple - a dynamically scaled model made to spin in a vertical current of air in a spin tunnel will provide the complete steady state spin characteristics. With provision for activation of the controls then the recovery characteristics are also readily determined. Test runs occupy little time and so a variety of control positions and loading conditions can be assessed very quickly. Minor design changes can also be easily accommodated and so the dynamic model tests represent a very powerful aid in a development programme.

As with the rotary balance tests there are problems associated with scale effects. Another problem, associated with the separated airflow, is the possibility of "cliff-edge" effects where a series of changes may have no effect up to a certain point beyond which even a slight change may have drastic consequences. These two problems in particular require experienced interpretation of the test results.

The use of these two model test techniques on the one project is most effective but is likely to be beyond the budget of a small aircraft company. Rather than neglect model spin tests completely it is worth considering the use of a miniature spin tunnel as built by Robelen (4). The tunnel has only a 0.6 m diameter working section so the models are very small and this has raised serious concerns about the validity of the results. The tunnel has been used for one test series of modifications to the Victa Aircruiser and the results appeared quite satisfactory. As the technique is very inexpensive there is merit in further testing to develop comparisons with other spin tunnels as well as full-scale results so that its limitations can be defined.

1 2 4 CONTINUE

Back to Spinning