## Exploitable results

The fuselage is a major source of drag for helicopters in high-speed forward flight, where it can represent 50% of the total drag of the helicopter. Drag reduction is therefore a key objective for industry in the development process of new rotorcraft. Because of limitations in current prediction tools to accurately estimate fuselage drag, fuselage design is presently based on numerous tests of small wind tunnel models. Nevertheless, these wind tunnel tests are obtained at low Reynolds number and the experimental results cannot directly be used to estimate the helicopter performance. These failures were the main motivations of the Brite-EuRaM HELIFUSE project.
An extensive wind tunnel test was completed in the ONERA F1 pressurised wind tunnel on a modular 1/4th scale model of the DGV fuselage. These tests were completed both to investigate the influence of Reynolds number on the flow-field, loads and moments applied to the fuselage, and to obtain a detailed database for code validation. The range of Reynolds numbers investigated in HELIFUSE covers the full domain of interest from small scale to scale one high-speed flight conditions. The test results showed that the Reynolds number highly affects the fuselage drag for the low values of this parameter, even when very limited strong viscous effects such as flow separation are present, but this effect is obviously non-linear with respect to the incidence and sideslip angles. However, fuselage pressure is very little affected by Reynolds number, because almost no flow separation is present on the DGV fuselage for the test domain investigated.
Prediction computations were conducted by the Partners in parallel with the wind tunnel tests in order to assess the capability of Navier-Stokes methods to predict the flow-field and loads acting on the fuselage. In general, the Navier-Stokes methods used were able to accurately predict fuselage pressure, but large differences were found between the predicted values of skin-friction, leading to large discrepancies in the estimated drag value. Additional computations were then defined and done in order to understand these differences (effect of grid, effect of solver, effect of integration, comparison structured/unstructured methodologies, effect of strut). It allowed to significantly improve the comparison between the computations using multi-block structured methods. A second phase of the work consisted in improving the Navier-Stokes methods used in the project. Essentially two points were considered, namely the introduction of additional turbulence models and their testing in order to select the best one suited for a given solver, and the implementation of preconditioning techniques to improve the accuracy and efficiency of compressible solvers at low-speed conditions. The final phase of the computational activities consisted in using the improved methods for completing validation computations for five test cases using common grids.
The HELIFUSE project has improved both the experimental knowledge and simulation and the numerical prediction aspects of helicopter fuselage aerodynamics. Improvements have been obtained in the prediction of the effects of Reynolds number, angle of attack and geometry variations on fuselage flow-field and loads. Although the comparison with experiment has only slightly improved from the initial prediction exercise, there is now significant insight into the requirements of each methodology and of each code in order to be able to capture the physics. This is particularly so for the structured methods for which strengths and limitations have been clearly identified.