CFD–wind-tunnel correlation of a scale model

Paper · Master's thesis, Ch. 4 & 6

This paper closes the loop on the front-wing work: comparing the CFD predictions against physical measurement, and dissecting the flow field of the optimal designs to explain where their performance comes from.

Scale model

A 32.4% scale model of the wing was divided into printable sections, 3D printed and assembled, then tested in a closed-circuit wind tunnel reaching 50 m/s. At that scale and speed the model reproduces real speeds close to 60 km/h — an F1 slow corner — so full Reynolds matching is not achieved and the correlation is bounded accordingly.

The boundary layer on the model is deliberately tripped near the leading edge so that model, full-scale wing and simulation all share a turbulent boundary layer, which is what makes the comparison meaningful despite the Reynolds mismatch.

3D-printed scale model assembled in the tunnel test section/assets/img/papers/tunnel-model.png

Fig. 1 — 32.4% scale printed model installed for testing.

Solution quality

Residual histories for the selected optima are reported first, establishing that the compared solutions are themselves converged.

Residuals plot for the median optimum/assets/img/papers/residuals-median-optimum.png

Fig. 2 — Residuals plot for median optimum (Z-mode).

Flow field analysis

For the selected optima the paper examines the chordwise pressure coefficient at mid-span, wall y+ on both surfaces to confirm the near-wall treatment is being applied in its valid range, and the surface pressure coefficient viewed from above and below.

Chordwise pressure coefficient at mid-span for both optima/assets/img/papers/cp-midspan.png

Fig. 3 — Chordwise C_p at mid-span (y = 0.3 m) for both optima.

Wall y+ on the pressure and suction surfaces/assets/img/papers/wall-yplus.png

Fig. 4 — Wall y+ on the top (pressure) and bottom (suction) surfaces.

Surface pressure coefficient from above and below/assets/img/papers/surface-cp.png

Fig. 5 — Surface C_p viewed from the pressure and suction sides.

Wake structure

On a plane behind the wing (x = −0.4725 m), the streamwise vorticity, total-pressure loss and turbulent kinetic energy are examined together. Taken as a set these explain how the geometry produces its downforce and, just as importantly, where its losses are being generated.

Streamwise vorticity, total-pressure loss and TKE on the wake plane/assets/img/papers/wake-plane.png

Fig. 6 — ω_x, total-pressure loss and k on the plane x = −0.4725 m behind the wing.

Nominal versus optimized

The paper closes with the comparison that motivated the whole campaign: the optimized geometries against the nominal starting point, in both aerodynamic modes.

← See the project