Validation of Particle Velocities in Solid-Liquid Multiphase Flow Using PIV and CFD

Abstract The entrained solid particles in multiphase flows present significant challenges for fluid handling systems, particularly due to the erosion of internal surfaces. This erosion affects the life, reliability, and safety of pipelines and other flow systems. A key factor influencing erosion is the particle impact velocity, which determines the extent of material degradation over time. However, accurately predicting these velocities remains a challenge due to the complex spatial and temporal distribution of particles within the flow. Despite extensive research, a universally applicable and highly reliable model for predicting impact velocities in multiphase flows has yet to be established. Many existing erosion prediction models rely on computational fluid dynamics (CFD) simulations. However, these models often include assumptions that fail to accurately capture the intricacies of multiphase flows, especially in geometrically complex regions such as pipe bends. The need for improved predictive models is critical in industries where erosion related failures lead to increased maintenance costs and potential safety hazards. The current study focuses on examining particle velocities in a liquid-solid multiphase flow through a 90-degree pipe bend using a commercially available CFD software package that incorporates a discrete phase model. The accuracy of the CFD predictions was validated through experimental testing using particle image velocimetry (PIV), which provided a direct comparison of simulated and experimental velocity data under similar flow conditions. The PIV experimental setup consisted of a transparent glass elbow through which a recirculating water-solid mixture was pumped. To ensure high image clarity, glass beads with a diameter of 10 microns were used at a concentration of 1%, alongside fluorescent tracer particles that enhanced visualization during imaging. To assess the performance of the CFD model across varying flow conditions, both simulations and experiments were conducted at three different velocities: 0.37 m/s, 0.56 m/s, and 0.75 m/s, with a maximum volumetric flow rate of 6 gallons per minute (GPM). These velocities were selected based on the limitations of the experimental setup and to provide a representative range of flow conditions. The comparison of the CFD and PIV results demonstrated strong agreement, confirming the reliability of the CFD based approach for predicting particle velocities. The result of this study demonstrates the potential of CFD simulations to accurately model particle impact velocities in multiphase flows. Improved prediction of impact velocities can contribute to the development of more robust erosion prediction models, ultimately enhancing the design and maintenance of fluid transport systems. By refining computational techniques and incorporating experimental validation, future research can further improve the accuracy of erosion assessments and mitigate the risks associated with particle-induced wear.