The aim of this study is to measure secondary flow in a semifilled pipe. Firstly, the experimen- tal setup for stereo particle image velocimetry (stereo-piv) measurements with a laminar pipe flow at Reb = 1000 is validated. The theoretical basis for this is the Hagen-Poiseuille law. The mean streamwise velocity distribution shows promising results, but some irregularities can be found for the mean crossflow. These disturbances should not be seen in a completely deve- loped laminar flow. The conclusion is that the inlet length of the pipe to the measuring point is too short for the flow to develop fully. In addition, a non-negligible registration error can be re- cognised. This is also shown within the Reynolds stresses, especially in azimuthal direction. Here, the magnitude found is clearly above the expected values. The following measurement of the turbulent semifilled pipe flow at Reτ = 504 shows similar results. The data is compared with J.Brosda and M.Manhart [1] (Reτ = 460). As in the measurement of the laminar pipe flow the mean streamwise velocity component indicates promising results. The Reynolds stresses in streamwise direction correlate with the comparison data. This correlation decreases for in plane directions. The velocity field shows significant movements, nonetheless no secon- dary flow can be found at this point. As in the laminar measurement, the flow has not fully developed yet. Therefore, a longer pipe is recommended for future experiments.     «

The aim of this study is to measure secondary flow in a semifilled pipe. Firstly, the experimen- tal setup for stereo particle image velocimetry (stereo-piv) measurements with a laminar pipe flow at Reb = 1000 is validated. The theoretical basis for this is the Hagen-Poiseuille law. The mean streamwise velocity distribution shows promising results, but some irregularities can be found for the mean crossflow. These disturbances should not be seen in a completely deve- loped laminar flow. The conc...     »