Numerical Design and Development of High Efficiency Centrifugal Dredge Pumps
Type:
Presented during:
CEDA Dredging Days 2021
Authors:
S. Sapkota, M. Winkelman, and E.V. Duursen
Abstract
Centrifugal dredge pumps, used in the dredging industry, are designed to continuously transport and handle mixtures of water and solids (also called slurry). The flow in the impeller, as we know, is three-dimensional and highly turbulent. Dredge pumps used by Damen Dredging Equipment (DDE) are not designed from the condition "no-shock" or, small incidence at the impeller blade's inlet, unlike ordinary centrifugal pumps. The pump design is based on the spherical passage to accommodate larger particle sizes that may hamper the pump components. Operators are more concerned about the continuity of the operation or uptime, which depends on the extent of pipeline functionality.
An original approach by DDE resulted in the development of DDEturbo, DDE's in-house impeller design software based on Matlab environment, to predict the spherical ball passage from the blade input parameters. What sets this software apart from the commercial Impeller design software available in the market is that the designer can obtain the spherical ball passage directly in the design phase, unlike the traditional way of modeling the design and measuring the ball passage. The paper presents the approach of hydraulic blade design using DDETurbo, modeling, and numerical investigation with ANSYS CFX.
The numerical study is based on the results of the steady Reynolds Averaged Navier Stokes (RANS) model of water flow with the Shear Stress Transport turbulence model for closure. The basic hydraulic pump design process incorporates setting inputs (spherical ball passage, blade angles, related diameters), generating model, and Computational Fluid Dynamics (CFD) analysis. Additional steps include Fluid-Structure Interaction (FSI), Unsteady RANS simulation, Modal Analysis of Pump-shaft system, verification, and validation of the results.
CFD analysis is used in the numerical prediction of the performance (head generated, pump efficiency, power consumed and Net Positive Suction Head required (NPSHr)) of the pump in the form of performance curves.
Prediction of the suction properties of the impeller is based on the Cavitation study. Multiphase flow analysis with water vapor was used to predict the NPSHr. Lowering the outlet pressure of the impeller to mimic the cavitation phenomena is the basic idea behind the analysis. Numerical results depicting the performance characteristics are presented. Analysis of the velocity and pressure field along with the suction properties depicting the effect of cavitation are explored.
An investigation on the performance capabilities of ANSYS CFX and NUMECA FINE Turbo for impeller Blade design is discussed as a separate study. The investigation revealed that Numeca Fine Turbo can handle pump performance in a similar capacity to ANSYS CFX. The upgrade in the number of processor cores utilized and software change from CFX to Numeca is predicted to save DDE approximately 14% of the previous design time.
The paper expects to inculcate the readers and CFD engineers with the theory and practical aspects of impeller blade design while shining light on the processor cores scalability and comparison between two CFD softwares for impeller design. The results of this study helped to develop highly efficient (>85%) pumps with stable operation regimes and better suction properties in a faster way.
Keywords: CFD, Centrifugal Pumps, Cavitation, Ansys, Numeca.