The flow around a model of the campus Gløshaugen of the Norwegian University of Science and Technology in Trondheim is investigated experimentally with cobra probes and pressure taps at selected locations. The model is in a ratio of 1:320 and includes a total of 18 building complexes with a high degree of detail, spread over an area of more than 1000 m^2 in reality. This can be considered a typical representation of an urban environment. An atmospheric boundary layer is emulated in the wind tunnel by the use of triangular spires and a horizontal bar upstream of the model. The wind tunnel has a width of 2.70 m and a height of1.80 m along a 11.15 m long test section. The model has a maximum elevation ofthe wind tunnel floor of 0.2 m, resulting in an overall blockage below 5%. Wind from four different directions is simulated with different orientations of the model in the wind tunnel. The inflow Reynolds number Re with regard to the highest building height and turbulence intensity u′∞/U∞ are held constant at Re = 3.3·10^4 and u′∞/U∞ = 12.4% at building height. Reynolds number independency is shown in the operating range of the facilities. From previous CFD simulations promising locations regarding wind energy exploitation are selected. Velocity measurements are conducted at various heights at these locations to obtain velocity profiles and all over the urban model at a fixed height to create a comprehensive flow field. The cobra probes measure at f = 650 Hz and capture the velocity in all three directions within a 45 degree acceptance cone. These measurements are complemented by measurements of the static pressure at fixed pressure taps on three buildings and on the ground.

At the windward edge of a flat roof building the flow generally separates and a zone of highly turbulent, partially recirculating flow results close to the roof. This region is to be avoided when placing a windturbine due to its low wind speeds and high turbulence. Previous studies show that around a bluff body outside of the recirculation zone an acceleration of the flow can be observed. The occurrence of this effect is investigated in this experimental study. It is demonstrated that obstacles upstream of a potential site mitigate flow acceleration above a building. Wakes of multiple buildings interact, resulting in large areas of low wind speeds, increased turbulence intensities and deflections of the main wind direction in the flow approaching subsequent buildings. In strongly obstructed flows this can lead to the acceleration being reduced to a negligible level. For selected buildings also the effect of different positions on a roof is examined. It is shown that if a flow acceleration is visible it is stronger closer to the windward edge, whereas further downstream the effect washes out. The impact of these effects on available wind power is illustrated.