Numerical assessment of wind tunnel blockage effects for bluff bodies

The blockage effect due to the presence of wind tunnel walls represents a big obstacle to the accuracy of the acquired data during testing. The flow constriction around the model depends primarly on wind tunnel dimensions and shape of the tested model. In this thesis, a numerical study of the correlation between a change in vehicle length and wind tunnel blockage effect is performed on a simple car model known as Ahmed body. To do that, two case scenarios are simulated: an open case, used as reference for the comparison, and a wind tunnel case, to simulate the effects of blockage. A comparison between RANS, DDES and SAS is perfomed to choose the simulation technique. Furthermore, the accuracy of different blockage correction methods available in literature is analyzed. RANS exhibits a good agreement with DDES and SAS results with the latter that is proven to better reproduce the main flow feature of the Ahmed body with 25° of slant angle. In particular, it is able to predict the separation bubble over the slant but not the exact separation point while, in RANS and DDES, no separation bubble is exhibited. However, RANS is chosen to investigate wind tunnel blockage because it represents a trade-off between computational cost and quality of the simulation. It has been observed an enhanced blockage effect when the length of the test vehicle is increased keeping the same width-to-heigth ratio. In particular the increment in drag coefficient, net of the skin friction contribution, is larger than the lift one which is almost independent on the model length. Lastly, the comparison of wind tunnel correction methods has shown that the accuracy is highly dependent on the model that is studied with most of the methods that fail to predict the correct change in drag coefficient when the vehicle length is increased.