Investigation on slurry erosion of different pumping materials and coatings

Abstract

Centrifugal slurry pumps are widely used for transportation of solids through pipeline in many chemical and mining industries. Erosion wear considered as one of the important parameters which plays significant role in slurry transportation system by affecting both initial cost and life of its components. In the present work, erosion wear due to solid–liquid mixture has been investigated using a slurry erosion pot tester. Three different slurries namely fly ash, bottom ash and sand were used to carry out the erosion wear experiments. Four different pumping materials namely grey cast iron, stainless steel SS304, SS316L and super duplex stainless steel SDSS2507 were used. The erosion tests have been conducted on different slurry pump materials to establish the influence of rotational speed, particle size, solid concentration, time duration and materials type. The solid concentrations of erodent materials lie in range from 30 to 60% (by weight). The experiments were performed at four different speeds namely 600, 900, 1200 and 1500 rev/min with time duration of 90, 120, 150 and 180 minutes. Erosion wear results of different pumping materials shows that the rank of material type was highest amongst rotational speed, time, solid concentration (wt. %) and average fineness number. Among from process parameters the rotational speed was the most influencing factor. Erosion wear of different coatings was found to be increased non-linearly with increase in time duration, rotational speed and solid concentration whereas decreased with increase in CF value. The erosion wear of different coatings follows power-law relationship with CF value. The increment of erosion wear was higher at initial phase of erosion wear experiments. It can be concluded that erosion wear process affects the particulate properties. The size of particle decreases after erosion wear experiments. High-velocity oxy-fuel (HVOF) process was used to spray the different coating powders on pump materials. The WC-10Co-4Cr cermet powder was modified by addition of small proportion of Y2O3, Mo2C and ZrO2 powders. The 3% addition Y2O3 and Mo2C was helpful to improve the erosion wear resistance of WC-Co-Cr cermet coating. The order of erosion wear is found as: WC-10Co-4Cr+3%Y2O3 < WC-10Co-4Cr+3%Mo2C < WC-10Co-4Cr < WC-10Co-4Cr+3%ZrO2.In present investigation, the coating powders used was Ni-20Cr2O3, Ni-20Al2O3, Al-20TiO2, Stellite-6 and Colmonoy-88. The maximum erosion was observed at 30º impact angle arrangement for Stellite-6 and Ni-20Cr2O3 coatings whereas at 60º impact angle for Colmonoy-88, Ni-20Al2O3 and Al-20TiO2 coatings. It was found that the Stellite-6 and Ni-20Cr2O3 coating shows the ductile erosion wear whereas Colmonoy-88, Ni-20Al2O3 and Al-20TiO2 showed the semi-brittle erosion wear behavior. Microscopically, eroded surface of Stellite-6 showed the microcutting, plastic deformation, ploughing actions with few craters and lips. Colmonoy-88 surface underwent to craters, ploughing, carbide/boride pullout, fractures and intact. Erosion wear mechanisms on eroded surface of Colmonoy-88 were neither purely ductile nor brittle. The Ni-20Cr2O3 coating underwent to ploughing, microcutting, craters, ploughing, craters, ductile fractures. Micro-hills were visualized present on the eroded surface of Ni-20Cr2O3 coating at different locations. Simultaneously, the loosely bonded alumina grains were appeared on the surface of Ni-20Al2O3 coating by impact of erodent particles. Furthermore, no crack or splat debonding was observed on the surface which indicates the semi-brittle behavior of the coating. Commonly observed erosion wear mechanisms on the surface of Al-20TiO2 coating were craters and ductile fractures which shows its semi-ductile behavior. Erosion wear was successfully simulated by using an artificial neural network. The percentage error between experimental and ANN predicted erosion wear was found as ±4.87, 4.27 and 4.98% for sand, fly ash and bottom ash slurry which indicates the close agreement between neural network and experimental results.