Digital Wind Tunnel

Digital Wind Tunnel is an executive that runs Cmarc repeatedly, varying angle of attack and elevator or rudder deflections to obtain longitudinal and lateral static and dynamic stability derivatives, neutral point, and elevator deflection required to trim.

Successful use of DWT requires a well-formed geometry model without problems of wake misbehavior, leakage, panel disproportion, and so on. Regions of the model surface corresponding to control surfaces must have been defined as tilt panels. If nonlinear analyses are to be used, the model must be free of large areas of separated flow and discontinuous streamlines.

The first step in running DWT is to fill in an opening form with some basic geometry information about the model. Some items can be omitted; for example, moments of inertia are not required if the dynamic stability evaluation is omitted.

Next comes a second form specific to the type of analysis to be run. Five categories of analysis are provided.

Pitch and Yaw Stability

The basic procedure is to define ranges of angle of attack or sideslip and elevator or rudder deflection. For linear analyses, only two angles and deflections are required; nonlinear analyses call more repetitions. The design point - equilibrium for a given attitude and CG location - is extrapolated from the results.

Roll rate

Roll rate is calculated using a combination of direct Cmarc analysis and an empirically-based adjustment.

Sensitivity

The Sensitivity analysis displays the degree to which six stability coefficients are influenced by five variables, including time stepping parameters.

Dynamics

The dynamics analysis estimates dynamic stability characteristics using model moments of inertia, and characterized the predicted behavior.

Validation

Many methods have historically been used to obtain stability derivatives, including flight test, wind tunnel investigation, and various numerical procedures, many of which are epitomized by the so-called DATCOM. Results from Digital Wind Tunnel fall within the normal bounds of variation among methods.