In this paper, the S-CLSVOF (simple coupled level-set and volume of fluid) and BM (bifurcation model) models are used for the accurate prediction of the flow in an open pump intake with a vertical pipe. The continuous wavelet transform, which is suitable for the vortex detection is applied to the pressure signals near both air-entrained and subsurface vortices. Low-frequency with long duration for air-entrained vortex due to the vortex wandering and broadband with short duration for floor vortices (dominant in subsurface vortices) due to the generation of extreme strong vortex are observed. The vortex motion mechanism is revealed by the analysis of the transport equation of the vertical vorticity's enstrophy. Different from previous results in which the tilting effect controls the vortex motion, the stretching effect is found to be dominating when it is large enough. When going through the bell mouth, the air-entrained vortex's vorticity changes the sign and strengthens the vortex with the same sign. On the plane near the bell mouth, three vortex patterns including co-rotating pair, merging and counter-rotating pair are observed. The onset criterion of the vortex merging at a/b = 0.29~0.32 is found to be applicable to the present case. The counter-rotating pair is found to be more stable. Air-entrained vortex serves as an amplification of the strong vortices generated from subsurface.

Stall and jet-wake flow are two typical forms of unstable flow phenomena in centrifugal pumps, significantly affecting their stability. This paper investigates the interaction mechanisms between stall and jet-wake flow in a centrifugal impeller under different working conditions. The unsteady numerical study utilizes the partially averaged Navier–Stokes (PANS) model with a new dynamic fk expression derived from the rotation-corrected energy spectrum. The results reveal four stages in the flow field evolution of the centrifugal pump under different working conditions. In stage I, no stall vortices are present, and the jet-wake flow occurs. The velocity distribution at the impeller outlet depends on the pressure difference distribution between the pressure and suction sides within the flow passage. As the flow rate decreases, the pressure difference between the two sides increases, intensifying the jet-wake phenomenon. Under part-load conditions (stages II, III, and IV), the presence of stall vortices becomes the main factor affecting the jet-wake flow. These stall vortices influence the local and downstream flow fields, thus altering the distribution of the jet-wake. When the stall vortex is on the pressure side, it reduces the velocity near the pressure side, weakening the intensity of the jet-wake. On the other hand, when the stall vortex is on the suction side, it reduces the velocity near the suction side, enhancing the intensity of the jet-wake.

This study aims to research the prediction performance of a bifurcated adaptive DDES (BADDES) model in different turbulent flows with rotation and curvature

This study aims to investigate the critical state of the saddle-shaped unstable region of the axial-flow pump and propose a suitable criterion for identifying this state. The bifurcation SST $k–\omega$ model considered the rotation effect is used in the present work and verified in the numerical calculation of a water jet pump. Then, it is used to simulate the critical state of the axial-flow pump. Results show that the leading-edge separation vortex generates at 0.6Q_d, while the head declines only at 0.55Q_d. Therefore, using the inflection point of the head-flow curve as the critical state criterion is unsuitable. In addition, the fixed monitoring point is unsuitable for identifying the critical state due to the insensitivity to the amplitude, main frequency, and periodicity changes at the critical state. Finally, to identify the critical state, it is essential to arrange a monitoring point at the leading edge of the blade suction near the shroud, which should rotate with the impeller. The critical state criterion is that the main frequency position of the pressure fluctuation signal is offset at the monitoring point, and the amplitude is increased by 10 times.

Purpose: The purpose of this paper is to obtain a better understanding of the rotor–stator interaction in the vaneless region of a centrifugal pump. Design/methodology/approach: A third-order sub-grid scale (SGS) model containing the rotation rate tensor named the dynamic cubic non-linear model (DCNM) is used for simulating the flow field in a centrifugal pump with a vaned diffuser. The pressure coefficient and velocity distributions are compared with the experimental data. Focusing on the vaneless region, the pressure pulsation, Reynolds stress pulsation and Reynolds stress transport equation are analyzed. Findings: The comparison of the calculation results with the experimental data indicates that the DCNM can accurately capture the distributions of pressure and velocity in the vaneless region. Based on the instantaneous pressure signals, the pressure pulsation is analyzed to show that in the vaneless region, the dominant frequency near the impeller is twice the blade passing frequency, whereas it is equal to the blade passing frequency near the diffuser. Further exploration of the Reynolds stress pulsation shows the correlation between the two variables. Additionally, the extreme low frequency of Reynolds stress near the diffuser is found to be related to the rotation instability. To explore the turbulence characteristics in the vaneless region, the Reynolds stress transportation equation is studied. In the vaneless region, the rotation term of the Reynolds stress transport equation is negligible compared to the production term, although the rotation instability is obvious near the diffuser. The production of the Reynolds stress plays the role of redistributing the energy from the uu component to the vv component, except for the region near the impeller outlet. Originality/value: The third-order SGS model DCNM has proved to be promising in simulating the rotor–stator interaction. The analysis of the rotation instability and the Reynolds stress transport equation shed light on the further understanding of the rotor–stator interaction.