Cavitation CFD using STAR-CCM+ of an Axial Flow Pump with Comparison to Experimental Data

Transcription

1 Cavitation CFD using STAR-CCM+ of an Axial Flow Pump with Comparison to Experimental Data Edward M. Bennett, Ph.D. Vice President of Fluids Engineering March 17, 2014

2 The Project Mechanical Solutions, Inc. (MSI), an engineering consultancy, was approached by a major pump manufacturer to undertake a redesign of a line of axial pumps A class of axial pumps requires redesign to achieve reduced Net Positive Suction Head Required (NPSHr) The customer wishes to validate STAR-CCM+ against the existing configuration before proceeding with redesign An Internal Research and Development Effort was undertaken to examine the efficacy of the cavitation model in STAR-CCM+ The axial pump configuration was analyzed using several turbulence models and associated cavitation parameters A second complex test configuration for which exact test conditions were known was also analyzed Conclusions were made regarding STAR-CCM+ capability to resolve cavitation problems in complex pump configurations 2

3 Definitions The phenomenon of cavitation occurs when the net pressure in the fluid decreases below the vapor pressure, e.g Pa for water at 25 C The net pressure in the fluid is a function inlet pressure, which is commonly stated in terms of head, i.e. Net Positive Suction Head (NPSH) Net Positive Suction Head Available (NPSHa) is the actual fluid energy at the inlet, defined as the difference between the inlet total head and vapor pressure expressed in terms of head Net Positive Suction Head Required (NPSHr) is the NPSH point at which the pump performance drops below some acceptable level, often defined as a point at which the total dynamic head (TDH) produced by the pump drops by 3% Cavitation can thus be reduced by increasing NPSHa via inlet conditions or decreasing NPSHr via geometry modifications 3

4 Breakdown Curve The occurrence of cavitation is visually presented via the breakdown curve, where NPSHa is plotted against TDH As NPSHa is lowered, the onset of cavitation is marked by a drop in TDH the value of NPSHa at the 3% drop point defines NPSHr 4

5 5 Axial Pump Configuration

6 Axial Pump Flowpath Geometry Pipe Propeller Inlet 6

7 7 Flowpath Mesh in STAR

8 8 Mesh Details

9 Mesh Statistics Domain Vertex Count Cell Count Inlet 107,154 26,848 Propeller 2,095, ,104 Pipe 624, ,902 TOTAL 2,826,447 1,053,854 9

10 Axial Pump Model Setup Realizable k-ε turbulence model Segregated flow solver 2 nd -order convection scheme Multi-phase Volume of Fluid (VOF) model Rayleigh-Plesset cavitation model Boundary conditions: - Variable inlet total pressure via pressure reference point rpm rotating speed kg/s inlet and exit mass flow (12000 gpm) Transient timestep of 2.123e-4 s (360 per rev) iterations per step IMPORTANT: MSI did not receive any details regarding test conditions, such as temperature or experimental setup 10

11 Alternative Setups SST k-ω turbulence model SST model with doubled cavitation seed density (2e12/m 3 ) Spalart-Allmaras turbulence model Finer mesh (remeshed with all mesher sizing values halved) 11

12 12 Flowfield and Pressure Contours

13 13 Streamlines

14 Vapor Fraction Contours Inlet Total Pressure kpa Inlet Total Pressure kpa Inlet Total Pressure kpa Inlet Total Pressure kpa 14

15 Cavitation Breakdown Results Main Model 15 N ss NPSHa [ft] Inlet Total Pressure [psi] Outlet Total Pressure [psi] Total Pressure Rise [psi] TDH [ft] TDH Drop [%] % % % % % % % % % % % %

16 Cavitation Breakdown Results Alternative Models: SST Model SSTmodel double seed density Spalart-Allmaras model rke model finer mesh NPSHa [ft] TDH [ft] NPSHa [ft] TDH [ft] NPSHa [ft] TDH [ft] NPSHa [ft] TDH [ft]

17 17 Turbulence Model Results

18 18 Seed Density Results

19 19 Mesh Refinement Results

20 Axial Pump CFD Conclusions STAR-CCM+ performed well in predicting the trend of the cavitation breakdown Further mesh refinement may bring results even closer to data Turbulence model did not greatly impact the results Bubble seed density did not have a major impact MSI did not have access to the experimental rig setup and this could have additional effect on results

21 Additional Test Case An additional test case became available to MSI A complex double suction pump was made available with complete data regarding the cavitation data The data included fluid temperature, so precise representations of the liquid and vapor density could be applied in the CFD model

22 20 Double-Suction Pump Drawing

23 21 Double-Suction Pump Test Data

24 22 Double-Suction Pump Mesh

25 Mesh Statistics Domain Vertex Count Cell Count Suction 4,898,638 1,459,107 Impeller 6,945,706 2,639,844 Volute 2,929, ,935 TOTAL 14,774,181 5,010,886 23

26 Double-Suction Pump Setup SST k-ω turbulence model Segregated flow solver 2 nd -order convection scheme Multi-phase Volume of Fluid (VOF) model Rayleigh-Plesset cavitation model Boundary conditions: - Variable inlet total pressure via pressure reference point rpm rotating speed kg/s inlet and exit mass flow ( m 3 /hr) Transient timestep of 1.675e-4 s (360 per rev) 20 iterations per step 24

27 25 Velocity Flowfield

28 26 Pressure Contours

29 27 Streamlines

30 Vapor Fraction Contours Inlet Total Pressure 175 kpa Inlet Total Pressure 80 kpa Inlet Total Pressure 40 kpa Inlet Total Pressure 27 kpa 28

31 Cavitation Breakdown Results N ss NPSHa [m] Inlet Total Pressure [kpa] Outlet Total Pressure [kpa] Total Pressure Rise [kpa] TDH [m] TDH Drop [%] % % % % % % % 29

32 30 NPSH Curve

33 Conclusions STAR-CCM+ proves to be an accurate tool for cavitation analysis Turbulence model selection does not appear to have major effect on the results Bubble seed density does not appear to have major effect on the results Matching the fluid temperature and experimental setup is critical to good results 31

34 Acknowledgements MSI is acknowledged for funding this effort MSI gratefully acknowledges the Technical Support Group of CD-adapco for their continued guidance and support