In the design of slurry transport equipment used in the mining and dredging industries, the effects of solid particle velocity and concentration on hydraulic performance and wear need to be considered. Two ultrasonic techniques have been used to investigate slurry flows through a centrifugal pump casing: a local particle concentration measurement technique (Furlan et al., 2012) and a pulsed ultrasonic Doppler Velocimetry (PUDV) technique (Hanjiang, 2003, Garman, 2015).
Local particle velocities and concentrations have been obtained in a flow of soda lime glass particles (diameter of 195 μm) and water through the casing of a centrifugal slurry pump operating close to the best efficiency point using the two ultrasound techniques. For the concentration measurements, the acoustic properties of slurry flows such as sonic velocity, backscatter, and attenuation are correlated to the volume fraction of solid particles. The algorithm utilizes measurements obtained from homogeneous vertical pipe flow fields as calibration data in order to obtain experimental concentration profiles in the non-homogenous flow regimes which are encountered in the pump casing.
The PUDV technique correlates the Doppler shift in frequency associated with the movement of particles towards or away from the transducer. A two measurement (angle) technique is applied within the pump casing in order to account for the components of particle velocity which are orthogonal to the casing side wall. The techniques are utilized to obtain concentration and velocity profiles within the pump casing for overall average loop particle concentrations ranging from 7–11 % by volume.
The experimental results are compared with the concentration and velocity fields that are predicted by in-house finite element computational fluid dynamics (CFD) codes (Pagalthivarthi and Visintainer, 2009) which are used to predict wear in centrifugal slurry pump wet end components. Reasonable agreement is observed for both the concentration and velocity fields. Specifically, measurements indicate that there is a reduction of in-situ concentration and hence a corresponding radial acceleration of the particles with respect to the fluid occurring within the impeller which has also been predicted by computational predictions of flow through the impeller (Pagalthivarthi et al., 2013). Additionally, the prediction of the existence of secondary flow patterns by the casing computational code has been supported with the velocity measurements.
On the use of acoustic particle velocity fields in adjoint‐based inversion