PERFORMANCE AND WEAR CHARACTERISTICS OF A CENTRIFUGAL SLURRY PUMP HANDLING SOLID-LIQUID MIXTURES

Abstract

Centrifugal pumps are being extensively used for hydraulic transportation of solids over short and medium distances through pipelines where the requirements of head and discharge are moderate. The performance and erosive wear behaviour of the pump components are the most critical design and selection parameters. An improvement in performance reduces the energy expenditure while the reduction in erosion enhances the service life. The review of literature suggests that the efforts have been made to estimate the reduction in water performance of the pump for handling different types of solid particles and to find methods to mitigate it. Different correlations were proposed to estimate the pump performance handling slurry. Regarding the erosion of the pump components, different techniques were used to identify the zones of maximum localized wear. Two dimensional numerical modeling of the pump is generally performed to investigate the erosion of the components. The knowledge of the dominating parameters affecting the erosion of the pump components at different operating conditions is still not conclusive. There is a need to develop an understanding of erosive wear distribution of the components all along their length and width of the flow passages. The constants in the empirical models used to predict the erosion are generally varied with the properties of target material and erodent, and impact conditions. In view of above, the present study is aimed to fill the aforementioned gaps from the experimental and numerical investigations of the performance and wear of a centrifugal slurry pump. The performance of a 50 mm centrifugal slurry pump is evaluated experimentally with solid-liquid mixture to investigate the effect of flow rate, particle size, solid concentration, and rotational speed on the performance. The slurries of fly ash-water and sand-water are used to conduct the experiments. The accuracy of the available correlations to predict the head drop due to slurry is compared with the experimentally measured data. In search of an alternative approach to correctly predict the pump performance characteristics with solid-liquid mixture, the computational fluid dynamics (CFD) modeling of the centrifugal slurry pump model is performed using the commercial CFD code Fluent 19.0. Two modeling approaches namely Multiple Reference Frame (MRF) and Sliding mesh (SM) are used to predict the pump performance characteristics. The accuracy of predicting pump performance characteristic using the SM approach (unsteady) is found better than the MRF approach (steady). With the SM approach, a deviation below ±2.5% for complete head discharge characteristic and +5% for the complete efficiency discharge characteristic with respect the experimental data is obtained. Further, the multiphase modeling of the pump is performed viii using SM approach with two models, Mixture and Eulerian-Eulerian multiphase. The equi-size particulate sand-water slurry is used for simulation. The Eulerian-Eulerian multiphase model predicted the effect of the solids on the pump performance close to the experimental results as compared to Mixture model. The obtained accuracy with Eulerian-Eulerian model for predicting the effect of solids on head and efficiency is around ±2% and ±3%, respectively. The predicted results using Eulerian-Eulerian model confirm that the head and efficiency of the pump decrease with the increase in particle size and solid concentration. The particles of high specific gravity show less reduction in head and efficiency of the pump. The effect of solids on head and efficiency ratios of the pump is not the same. The difference in head and efficiency ratios majorly depends on the specific gravity of solids. Further, the effect of variation in particle size and concentration on the flow field in the impeller and casing has also been analyzed at best efficiency point operation. Non-homogeneous suspension of particles inside the blade channels and casing passages is examined. The particulate concentration is observed higher near the impeller back shroud, pressure side of the blades, and non-suction side of the casing as compared to other locations. Furthermore, the numerical modeling of the pump is performed with multi-size solid particulate slurry to investigate the effect of variation in particle size distribution on pump performance. The numerical modeling for multi-size solid particulate slurry predicted the head and efficiency ratio with the deviation of ±2% and ±3.5%, respectively, as compared to the experiments. The predicted pump performance with different multi-size particulate slurries shows that the drop in head and efficiency increases with the increase in weight fraction of bigger size particles in the multi-sized slurry. The non-uniformity in the particle flow field inside the impeller and casing increased with the increase in weight fraction of bigger size particles in multi-sized slurry. In the second phase of the study, erosion studies are performed in a laboratory test rig, pilot plant test setup and CFD code Fluent. A large size slurry pot tester of 270 liters capacity is used to investigate the erosion behavior of target materials namely steel 304L, grey cast iron (GCI) and high chromium white cast iron (HCWCI) in the velocity range of 9.0-18.5 m/s. The solid-liquid mixture is prepared using three different solids namely, sand, fly ash and iron ore by mixing with tap water to get 1% weight concentration. The erosion behaviour of the target materials is evaluated by varying the orientation angle from 15o-90º. The erosion rate (ER) in g/g of solids is found to increase with velocity having power index value varying between 2-3.5, which increases with increase in impact angle and depends on the target material and erodent. The ER of the material also increases with the increase in particle size with power index varying between 0.8-1.4 depending on the target material. Based on the generated ix experimental data, empirical correlations are developed to estimate the ER of all the three target materials with three different particulate slurries as a contribution of cutting and deformation wear of the equipment handling solid-liquid mixture. The correlation developed to estimate the erosion rate (ER) of steel 304L with sand-water slurry is given as:  2 C D90 ER  ER  ER sin where the cutting erosion rate (ERc) is given as: 11 2.15 0.71 C ER 1.51 10 f ( )V d     and the deformation erosion rate (ERD90) is given as: 13 2.8 0.98 D90 ER 2.36 10 V d    In the above expressions, V is the velocity, α is the particle impact angle, and d is the particle size. To investigate the erosive wear profile of the pump casing experimentally and its relationship with numerically simulated flow field, experiments are conducted in a pilot plant test rig with two equi-sized sand particulate slurries at two pump speeds and two flow rates. Wear specimens (1.5 x 1.5 x 0.01 cm3) of steel 304L are affixed at fourteen different locations along the centerline of the casing wall. The erosion of the specimens is determined based on the measurement of weight loss to study the wear pattern along the casing. Further, the flow field inside the pump is numerically simulated using Eulerian-Lagrangian modeling to correlate the particle impact condition with the measured wear. Scanning electron microscopic (SEM) images of worn out samples are also examined in each case to identify the dominant mechanism of erosion at different locations of the casing. It is observed that the wear at the volute tongue is contributed by both the cutting and deformation whereas, at all other locations, the material is removed due to cutting and ploughing. Further, the developed correlation from the pot tester data for steel 304L is used to predict the erosion of the pump casing. The predicted erosion profile of the casing centerline showed reasonably good agreement with the experiments. The numerical simulations are further performed to study the effect of particle size and flow rate on erosion rate distribution in the casing and impeller blade surface of HCWCI. On the basis of the present experimental and numerical studies on performance and wear characteristics of a centrifugal slurry pump, the following conclusions are drawn:  The effect of pump speed, flow rate, solid concentration, specific gravity, particle size, and particle size distribution on the pump performance is established. The head ratio x and efficiency ratio are found to be different which was found to depend on the specific gravity of solids.  An effective multiphase modeling approach to predict the pump performance characteristics using equi-size and multi-size particulate slurry is determined. The particle motion inside the wetted passage of the pump components is established.  The effect of impact angle, particle size and velocity on erosive wear of steel 304L, GCI and HCWCI is established. CFD based empirical correlations are developed to estimate the erosion rate of the above target materials in the velocity range of 9.0 to 18.5 m/s with sand, fly ash and iron ore particulate slurry as contribution of cutting and deformation wear.  The numerical modeling of the pump with the empirical correlations developed from pot tester data showed reasonably accurate prediction of casing erosion profile. The effect of particle size and flow rate on the erosion rate distribution along the length and width of the casing and impeller blade passages are established.