Comparison of flight test data with a computational fluid dynamics model of a Scottish Aviation Bulldog aircraft

Abstract

The following paper presents detailed aerodynamic data of a Scottish Aviation Bulldog light aircraft. The data is taken from the pre-stall region of the aircraft flight envelope through two flight test methods and from a geometrically accurate computational fluid dynamics (CFD) model of the full scale aircraft, which was meshed in Ansys ICEM CFD and solved in Ansys Fluent. The fidelity of the CFD model was achieved by development of a CATIA solid model with surfaces matching a spatial point cloud of the aircraft taken using a 3D laser scanner. Following a CFD verification process, a 3·4m hybrid mesh with a Spalart-Allmaras (SA) turbulence model was found to give the best overall lift and drag characteristics. Further detailed comparisons with the glide flight test data showed the CFD drag polar to have 63% lower zero lift drag, although this discrepancy was related to the simplification of the original CATIA surface model, which excluded the undercarriage, aerials and other protuberance drags. Inclusion of estimates of these sources of drag resulted in a match in zero lift drag to within 15% and a maximum lift to drag of 10:1 which was within 11% of the glide flight test result. The remaining drag discrepancy is attributed to other effects including trim drag and the surface finish of the actual aircraft.

Aerodynamic performance of X-plane with two different types of inboard stores

“…Figure 4 shows an example of convergence analysis at angle of attack of 15 o for the three considered configurations. The solutions have been normalized by the number of elements of 1.0×10 6 . It can be seen that the solutions begin to reach its convergence at 1.6×10 6 elements and beyond this point, the subsequent solutions differ by no more than 0.01%.…”

“…There are many other studies that have been performed to verify the CFD simulation results with the experimental data. These include, among others, the use of 3D RANS CFD software to simu-late the canard for SAAB passenger aircraft [3] and blended wing body (BWB) model [4], ANSYS CFX CFD software for different wing designs of a remote control aircraft [5], CATIA and ANSYS FLUENT for Scottish Aviation Bulldog light aircraft [6], general Zonal Euler Solver (ZEUS) to simulate the supersonic missile of SM-2 Block IVA [7] and the use of CFD++ software to study the drag force on a different geometry of nacelle [8]. All these results have been verified with the wind tunnel testing and they are found to be in good agreement.…”

Undercarriage drag prediction for a fixed undercarriage light aircraft

“…In Fluent v16.2, flow direction vectors and corresponding parameters were specified using the standard direction cosine matrix for a (1, 2, 3) Euler angle sequence to consider a full range of angles of attack from 3.4 • to 11.4 • so that direct comparisons to existing flight test data [5] could be made. All simulations were initiated with the Fluent set-up as summarised in Table 1.…”

“…All simulations were initiated with the Fluent set-up as summarised in Table 1. Reference values were taken directly from the aircraft manual, and flight conditions were chosen to simulate the flight test conditions [5]. Convergence of the solutions was typically obtained within 5000-8000 iterations in all cases, the criteria of which is based on the root mean square (RMS) variance of the lift and drag monitors.…”

“…The increased detail of the model, see Fig. 1, as compared to the previous CATIA half model [5], is expected to improve the correlation between CFD results and flight test data. It is worth noting that in the prior publication regarding the same aircraft [5], the authors mainly attributed the discrepancies in CFD results with respect to flight test data to the absence of the undercarriage.…”