Parallax correction for precise near-wall flow investigations using particle imaging

Abstract

For the basic understanding of turbulence generation in wall-bounded flows, precise measurements of the mean velocity profile and the mean velocity fluctuations very close to the wall are essential. Therefore, three techniques are established for high-resolution velocity profile measurements close to solid surfaces: (1) the nanoprobe sensor developed at Princeton University, which is a miniaturization of a classical hot-wire probe [Exp. Fluids 51, 1521 (2011)]; (2) the laser Doppler velocimetry (LDV) profile sensor, which allows measurement of the location of the particles inside the probe volume using a superposition of two fringe systems [Exp. Fluids 40, 473 (2006)]; and (3) the combination of particle image velocimetry and tracking techniques (PIV/PTV), which identify the location and velocity of submicrometer particles within the flow with digital imaging techniques [Exp. Fluids 52, 1641 (2006)]. The last technique is usually considered less accurate and precise than the other two. However, in addition to the measurement precision, the effect of the probe size, the position error, and errors due to vibrations of the model, test facility, or measurement equipment have to be considered. Taking these into account, the overall accuracy of the PTV technique can be superior, as all these effects can be compensated for. However, for very accurate PTV measurements close to walls, it is necessary to compensate the perspective error, which occurs for particles not located on the optical axis. In this paper, we outline a detailed analysis for this bias error and procedures for its compensation. To demonstrate the capability of the approach, we measured a turbulent boundary layer at Re(δ)=0.4×10(6) and applied the proposed methods.