Intense pressure pulsation, resulted from the flow structure shedding from the blade trailing edge and its interaction with the volute tongue and the casing, is detrimental to the stable operation of centrifugal pumps. In the present study, unsteady pressure pulsation signals at different positions of the volute casing are extracted using high response pressure transducers at flow rate of 0–1.55ΦN. Emphasis is laid upon the influence of measuring position and operating condition on pressure pulsation characteristics, and components at the blade passing frequency fBPF and root-mean-square (RMS) values in 0–20.66fn frequency band are mainly analyzed. Results clearly show that the predominant components in pressure spectra always locate at fBPF. The varying trends versus flow rate of components at fBPF differ significantly for different points, and it is considered to be associated with the corresponding flow structures at particular positions of the volute casing. At the near-tongue region, high pressure amplitudes occur at the position of θ = 36 deg, namely the point at the after tongue region. For different measuring points, angular distributions of amplitudes at fBPF and RMS values in 0–20.66fn frequency band are not consistent and affected significantly by the pump operating conditions.

Cavitating flow developing in the blade channels is detrimental to the stable operation of centrifugal pumps, so it is essential to detect cavitation and avoid the unexpected results. The present paper concentrates on cavitation induced vibration characteristics, and special attention is laid on vibration energy in low frequency band, 10-500 Hz. The correlation between cavitating evolution and the corresponding vibration energy in 10-500 Hz frequency band is discussed through visualization analysis. Results show that the varying trend of vibration energy in low frequency band is unique compared with the high frequency band. With cavitation number decreasing, vibration energy reaches a local maximum at a cavitation number much larger than the 3% head drop point; after that it decreases. The varying trend is closely associated with the corresponding cavitation status. With cavitation number decreasing, cavitation could be divided into four stages. The decreasing of vibration energy, in particular cavitation number range, is caused by the partial compressible cavitation structure. From cavitation induced vibration characteristics, vibration energy rises much earlier than the usual 3% head drop criterion, and it indicates that cavitation could be detected in advance and effectively by means of cavitation induced vibration characteristics.

The effect of shark skin on the boundary-layer separation process under dynamic conditions (maneuvers) has been studied experimentally. We use a foil covered with biomimetic shark skin to explore how this type of surface impacts boundary-layer dynamics in both steady and accelerating conditions. The effect of denticles is assessed via particle image velocimetry in the wake. It is shown that dynamic conditions and small-scale disturbances can mitigate boundary-layer separation through instantaneous modification of the local pressure-gradient distribution. For instance, the region of favourable pressure gradient can be extended by accelerating the foil. The acceleration results in a thinner separated shear layer on the foil surface when compared to the steady reference case. This remarkable difference indicates that local roughness (introduced through for instance biomimetic shark skin) may trigger an interaction with relatively large-scale structures in the boundary layer for effective boundary-layer control during unsteady propulsion and maneuvering.

Three-dimensional (3D) flow reconstruction over a 6:1 prolate spheroid using scanning stereoscopic particle image velocimetry has been conducted in a towing-tank facility. Forces, moments, surface pressure and reconstructed 3D vortical-flow structures are acquired in order to characterize the separated flow within the subcritical regime at three Reynolds numbers (Re = 0.5 × 106, 1.0 × 106 and 1.5 × 106), and at four incidence angles (α = 5◦, 10◦, 15◦ and 20◦). Key flow features are evaluated through the reconstructed 3D flowfield data. In particular, the interaction of the crossflow vortices and the surface-pressure distribution are discussed. It is shown that the crossflow vortices consist of helical vortex tubes as the dominant coherent structures, which are more distinct at large α. For α ≥ 15◦, four high-vorticity regions were observed. A pair of upper vortex tubes, originate on the windward side of the model, while a lower vortex tube is formed on the strong curvature region on the leeward side of the model. The vortex dynamics are further characterized through stretching and tilting terms via the vorticity-transport equation. Larger incidence angles α lead to a stronger alignment of the crossflow vortices with the mean flow direction, which is reflected in the tilting terms. A weak Re dependency of the loads and flow structures within the range 1.0 × 106 ≤ Re ≤ 1.5 × 106 is reported, while a significantly different result was captured for separation at Re = 0.5 × 106. An increase in separation size at Re = 0.5 × 106 at 20◦ results in an additional flow structure; a pair of secondary vortices is formed, leading to a change in the pressure distribution on the model surface when compared to higher Re.

No abstract