Analysis of the Turbulent Properties of Flow in Water-jet Pump Based on Wall-modeled Large-eddy Simulations

doi:10.12382/bgxb.2024.0850

Abstract

Abstract:

Scale-resolving simulations are potential for accurate predictions of unsteady flows.Wall-modeled large-eddy simulation (WMLES)has the advantages of wide application and relatively low computational cost.However,the application of WMLES to the turbulent flow in water-jet pump is limited.The WMLES based on the analysis of turbulent length scales is applied to simulate the flow fields in a prototype pump and a modified pump,and the simulated results are compared to the experimental results.The head rise and torque coefficients predicted by the WMLES agreed well with the experimental data.The analysis of the turbulent properties of flows in the pumps demonstrate that the WMLES is capable of reproducing the horseshoe vortices,corner vortices,tip concentrating vortices and tip leakage vortices that are difficult for the conventional Reynolds-Averaged Navier-Stokes (RANS)method to reproduce.The horseshoe and corner vortices of the modified pump are much weaker than those of the prototype pump,and the limiting streamlines of the modified pump are smoother that those of the prototype pump.The lower dissipation of energy caused by the weaker vortices led to the higher efficiency of the modified pump.The results demonstrated that the WMLES based on the analysis of turbulent length scales could be effectively applied to predict the performance and turbulent properties of flow in the water-jet pump,which can provide the data support for the improvement of water-jet pump.

Key words: large-eddy simulation, wall modeling method, turbulent length scale, water-jet pump, unsteady flow

XU Hui, CHEN Zuogang, CAI Youlin. Analysis of the Turbulent Properties of Flow in Water-jet Pump Based on Wall-modeled Large-eddy Simulations[J]. Acta Armamentarii, 2024, 45(S2): 55-64.

Figures/Tables 17

Fig.1 Schametic diagram of the turbulent energy spectrum and dissipation spectrum for homogeneous isotropic turbulence

Fig.2 Workflow of the wall-modeled large-eddy simulation based on the analysis of turbulent length scales

Fig.3 Geometrical models of the prototype pump (left)and modified pump (right)

Fig.4 Computational domain of flow in the water-jet pump

Fig.5 Upstream and downstream sections of pump for pressure monitoring

Fig.6 Comparison of the turbulent length scale and the filtering length scale in the pump

Fig.7 Distribution of the meshes on the pump surfaces and longitudinal section

Table 1 Comparison of the experimental (EFD)and RANS results of the prototype pump

K_Q	K_H(EFD)	K_H(RANS)	e(EFD)/%	K_M(EFD)	K_M(RANS)	e(LES)/%	η(EFD)	η(RANS)
0.7279	2.2355	2.294	2.62	0.3047	0.3018	-0.97	0.8498	0.8806
0.6927	2.3716	2.430	2.48	0.3062	0.3038	-0.80	0.8538	0.8820
0.6741	2.4275	2.488	2.49	0.3065	0.3033	-1.06	0.8497	0.8802
0.6451	2.5077	2.580	2.87	0.3060	0.3034	-0.86	0.8412	0.8729
0.6160	2.5617	2.625	2.49	0.3032	0.3017	-0.50	0.8282	0.8531
0.5831	2.5972	2.629	1.22	0.3009	0.2989	-0.66	0.8010	0.8161

Table 2 Comparison of the experimental (EFD)and LES results of the prototype pump

K_Q	K_H(EFD)	K_H(LES)	e(EFD)/%	K_M(EFD)	K_M(LES)	e(LES)/%	η(EFD)	η(LES)
0.7279	2.2355	2.2502	0.66	0.3047	0.3091	1.43	0.8498	0.8433
0.6927	2.3716	2.3919	0.86	0.3062	0.3098	1.16	0.8538	0.8512
0.6741	2.4275	2.4410	0.56	0.3065	0.3091	0.85	0.8497	0.8472
0.6451	2.5077	2.5339	1.04	0.3060	0.3078	0.56	0.8412	0.8453
0.6160	2.5617	2.6008	1.53	0.3032	0.3051	0.60	0.8282	0.8359
0.5831	2.5972	2.6523	2.12	0.3009	0.3006	-0.10	0.8010	0.8188

Table 3 Comparison of the experimental (EFD)and LES results of the modified pump

K_Q	K_H(EFD)	K_H(LES)	e(EFD)/%	K_M(EFD)	K_M(LES)	e(LES)/%	η(EFD)	η(LES)
0.6741	2.7575	2.7743	0.61	0.3318	0.3306	-0.37	0.8954	0.9007

Fig.8 Representations of the horseshoe vortices induced by the roots of impeller leading edges of the prototype pump and modified pump by means of the q criterion

Fig.9 Representations of the horseshoe vortices induced by the roots of impellers of the prototype pump (left)and modified pump (right)by means of the q criterion

Fig.10 Representations of the concentrating vortices near the impeller tips of the prototype pump (left)and modified pump (right)by means of the q criterion

Fig.11 Comparison of the time-averaged limiting streamlines on the pressure surfaces of the prototype pump (top)and modified pump (bottom)

Fig.12 Comparison of the time-averaged limiting streamlines on the suction surfaces of the prototype pump (top)and modified pump (bottom)

Fig.13 Comparison of the corner vortices induced by the diffusers of the prototype pump (left)and modified pump (right)by means of the q criterion

Fig.14 Comparison of the time-averaged limiting streamlines on the diffuser surfaces of the prototype pump (left)and modified pump (right)

References 17

[1]	SHANG W Q, LI D, LUO K, et al. Evaluation of the spatiotemporal unsteady characteristics of the tip leakage vortex based on a direct numerical simulation database[J]. Physics of Fluids, 2022, 34(6):065131.
[2]	CHOI H, MOIN P. Grid-point requirements for large eddy simulation:Chapman’s estimates revisited[J]. Physics of Fluids, 2012, 24(1):011702.
[3]	SPALART P R. Comments on the feasibility of LES for wings,and on a hybrid RANS/LES approach[C]// Proceedings of First AFOSR International Conference on DNS/LES.Reston, US: Greyden Press, 1997.
[4]	CAVAZZINI G, PAVESI G, ARDIZZON G. Pressure instabilities in a vaned centrifugal pump[J]. Proceedings of the Institution of Mechanical Engineers,Part A:Journal of Power and Energy, 2011, 225(7):930-939.
[5]	QIN D H, PAN G, LEE S, et al. Underwater radiated noise reduction technology using sawtooth duct for pumpjet propulsor[J]. Ocean Engineering, 2019, 188:106228.
[6]	ZHANG Q, JAIMAN R K. Numerical analysis on the wake dynamics of a ducted propeller[J]. Ocean Engineering, 2019, 171:202-224. doi: 10.1016/j.oceaneng.2018.10.031
[7]	SPALART P R, DECK S, SHUR M L, et al. A new version of detached-eddy simulation,resistant to ambiguous grid densities[J]. Theoretical and Computational Fluid Dynamics, 2006, 20(3):181-195.
[8]	SHEN X, ZHAND D S, XU B, et al. Experimental and numerical investigation on the effect of tip leakage vortex induced cavitating flow on pressure fluctuation in an axial flow pump[J]. Renewable Energy, 2021, 163:1195-1209.
[9]	FUREBY C, TOXOPEUS S L, JOHANSSON M, et al. A computational study of the flow around the KVLCC2 model hull at straight ahead conditions and at drift[J]. Ocean Engineering, 2016, 118:1-16.
[10]	LIEFVENDAHL M, JOHANSSON M. Wall-modeled LES for ship hydrodynamics in model scale[J]. Journal of Ship Research. 2021, 65(1):41-54.
[11]	SMAGORINSKY J. General circulation experiments with the primitive equations:I.The basic experiment[J]. Monthly Weather Review, 1963, 91(3):99-164.
[12]	GERMANO M, PIOMELLI U, MOIN P, et al. A dynamic subgrid-scale eddy viscosity model[J]. Physics of Fluids A:Fluid Dynamics, 1991, 3(7):1760-1765.
[13]	LILLY D K. A proposed modification of the Germano subgrid-scale closure method[J]. Physics of Fluids A:Fluid Dynamics, 1992, 4(3):633-635.
[14]	吴霆, 时北极, 王士召, 等. 大涡模拟的壁模型及其应用[J]. 力学学报, 2018, 50(3):453-466. doi: 10.6052/0459-1879-18-071
	WU T, SHI B J, WANG S Z, et al. Wall-model for large-eddy simulation and its applications[J]. Chinese Journal of Theoretical and Applied Mechanics, 2018, 50(3):453-466. (in Chinese) doi: 10.6052/0459-1879-18-071
[15]	SCHUMANN U. Subgrid scale model for finite difference simulations of turbulent flows in plane channels and annuli[J]. Journal of Computational Physics, 1975, 18(4):376-404.
[16]	GRÖTZBACH G. Direct numerical and large eddy simulation of turbulent channel flows[J]. Encyclopedia of fluid mechanics, 1987, 6:1337-1391.
[17]	MATHEY F, COKLJAT D, BERTOGLIO J, et al. Specification of LES inlet boundary condition using vortex method[J]. Progress in Computational Fluid Dynamics, 2006, 6:58-67.