The breakup of drop covered with vapor film is numerically simulated. The moving particle semi-implicit method is used to solve the 2-dimensional unsteady Navier-Stokes equations for drop, vapor and ambient fluid. The...The breakup of drop covered with vapor film is numerically simulated. The moving particle semi-implicit method is used to solve the 2-dimensional unsteady Navier-Stokes equations for drop, vapor and ambient fluid. The results show that vapor film suppresses the drop breakup and hence the critical Weber number increases with the increasing thickness of vapor film. The breakup process can be divided into two stages. The drop deformation and breakup mainly occur in the later stage. Three breakup mechanisms are unveiled, which are almost the same as that of drop breakup without vapor film except for the stronger Rayleigh-Taylor instability for drop with vapor film. Our simulation results are comparable with the previous experiments.展开更多
In this paper,the two-dimensional Rayleigh-Taylor(RT) instability is directly simulated using the moving particle semiimplicit(MPS) method,which is based on the fully Lagrangian description.The objectives of this pape...In this paper,the two-dimensional Rayleigh-Taylor(RT) instability is directly simulated using the moving particle semiimplicit(MPS) method,which is based on the fully Lagrangian description.The objectives of this paper are to investigate preliminarily the effect of viscosity and finite size domain on the evolution of the RT instability.The simulation results demonstrate that(1) the mushroom-like vortex motions are formed in late time due to fluid viscosity,which give rise to the secondary shear flow instability,(2) the finite thickness of the fluid layer limits the development of the RT instability.The above results are consistent with the experiments and theoretical analyses.Meanwhile,the linear growth rate of the RT instability obtained from the numerical simulation is also in agreement with theoretical analyses.And the nonlinear threshold from the simulation result is comparable with the theoretical estimate.Two stages of the nonlinear evolution of the RT instability are revealed in the numerical simulation,nonlinear saturation and turbulent mixing.展开更多
Moving particle semi-implicit （MPS） method is a fully Lagrangian particle method which can easily solve problems with violent free surface. Although it has demonstrated its advantage in ocean engineering application...Moving particle semi-implicit （MPS） method is a fully Lagrangian particle method which can easily solve problems with violent free surface. Although it has demonstrated its advantage in ocean engineering applications, it still has some defects to be improved. In this paper, MPS method is extended to the large eddy simulation （LES） by coupling with a sub-particle-scale （SPS） turbulence model. The SPS turbulence model turns into the Reynolds stress terms in the filtered momentum equation, and the Smagorinsky model is introduced to describe the Reynolds stress terms. Although MPS method has the advantage in the simulation of the free surface flow, a lot of non-free surface particles are treated as free surface particles in the original MPS model. In this paper, we use a new free surface tracing method and the key point is ＂neighbor particle＂. In this new method, the zone around each particle is divided into eight parts, and the particle will be treated as a free surface particle as long as there are no ＂neighbor particles＂ in any two parts of the zone. As the number density parameter judging method has a high efficiency for the free surface particles tracing, we combine it with the neighbor detected method. First, we select out the particles which may be mistreated with high probabilities by using the number density parameter judging method. And then we deal with these particles with the neighbor detected method. By doing this, the new mixed free surface tracing method can reduce the mistreatment problem efficiently. The serious pressure fluctuation is an obvious defect in MPS method, and therefore an area-time average technique is used in this paper to remove the pressure fluctuation with a quite good result. With these improvements, the modified MPS-LES method is applied to simulate liquid sloshing problems with large deforming free surface. Results show that the modified MPS-LES method can simulate the large deforming free surface easily. It can not only capture the large impact pressure accur展开更多
A numerical model has been developed to study sloshing of turbulent flow in a tank with elastic baffles. The Moving-Particle Semi-implicit method(MPS) is a kind of meshless Lagrangian calculation method. The large edd...A numerical model has been developed to study sloshing of turbulent flow in a tank with elastic baffles. The Moving-Particle Semi-implicit method(MPS) is a kind of meshless Lagrangian calculation method. The large eddy simulation(LES) approach is employed to model the turbulence by using the Smagorinsky Sub-Particle Scale(SPS)closure model. This paper uses MPS-FSI method with LES to simulate the interaction between free surface flow and a thin elastic baffle in sloshing. Then, the numerical model is validated, and the numerical solution has good agreement with experimental data for sloshing in a tank with elastic baffles. Furthermore, under external excitations,the MPS is applied to viscous laminar flow and turbulent flow, with both the deformation of elastic baffles and the wave height of the free surface are compared with each other. Besides, the impact pressure with/without baffles and wave height of free surface are investigated and discussed in detail. Finally, preliminary simulations are carried out in the damage problem of elastic baffles, taking the advantage of the MPS-FSI method in computations of the fluid–structure interaction with large deformation.展开更多
The micro- and macro-time scales in two-phase turbulent channel flows are investigated using the direct nu- merical simulation and the Lagrangian particle trajectory methods for the fluid- and the particle-phases, res...The micro- and macro-time scales in two-phase turbulent channel flows are investigated using the direct nu- merical simulation and the Lagrangian particle trajectory methods for the fluid- and the particle-phases, respectively. Lagrangian and Eulerian time scales of both phases are cal- culated using velocity correlation functions. Due to flow anisotropy, micro-time scales are not the same with the theo- retical estimations in large Reynolds number （isotropic） tur- bulence. Lagrangian macro-time scales of particle-phase and of fluid-phase seen by particles are both dependent on particle Stokes number. The fluid-phase Lagrangian inte- gral time scales increase with distance from the wall, longer than those time scales seen by particles. The Eulerian inte- gral macro-time scales increase in near-wall regions but de- crease in out-layer regions. The moving Eulerian time scales are also investigated and compared with Lagrangian integral time scales, and in good agreement with previous measure- ments and numerical predictions. For the fluid particles the micro Eulerian time scales are longer than the Lagrangian ones in the near wall regions, while away from the walls the micro Lagrangian time scales are longer. The Lagrangian integral time scales are longer than the Eulerian ones. The results are useful for further understanding two-phase flow physics and especially for constructing accurate prediction models of inertial particle dispersion.展开更多
基金supported by the National Natural Science Foundation of China(Grant Nos.50325620 and 10372050).
文摘The breakup of drop covered with vapor film is numerically simulated. The moving particle semi-implicit method is used to solve the 2-dimensional unsteady Navier-Stokes equations for drop, vapor and ambient fluid. The results show that vapor film suppresses the drop breakup and hence the critical Weber number increases with the increasing thickness of vapor film. The breakup process can be divided into two stages. The drop deformation and breakup mainly occur in the later stage. Three breakup mechanisms are unveiled, which are almost the same as that of drop breakup without vapor film except for the stronger Rayleigh-Taylor instability for drop with vapor film. Our simulation results are comparable with the previous experiments.
基金supported by the National Natural Science Foundation of China (Grant No. 50476008)
文摘In this paper,the two-dimensional Rayleigh-Taylor(RT) instability is directly simulated using the moving particle semiimplicit(MPS) method,which is based on the fully Lagrangian description.The objectives of this paper are to investigate preliminarily the effect of viscosity and finite size domain on the evolution of the RT instability.The simulation results demonstrate that(1) the mushroom-like vortex motions are formed in late time due to fluid viscosity,which give rise to the secondary shear flow instability,(2) the finite thickness of the fluid layer limits the development of the RT instability.The above results are consistent with the experiments and theoretical analyses.Meanwhile,the linear growth rate of the RT instability obtained from the numerical simulation is also in agreement with theoretical analyses.And the nonlinear threshold from the simulation result is comparable with the theoretical estimate.Two stages of the nonlinear evolution of the RT instability are revealed in the numerical simulation,nonlinear saturation and turbulent mixing.
基金supported by the National Natural Science Foundation of China(Grant No.50979059)
文摘Moving particle semi-implicit （MPS） method is a fully Lagrangian particle method which can easily solve problems with violent free surface. Although it has demonstrated its advantage in ocean engineering applications, it still has some defects to be improved. In this paper, MPS method is extended to the large eddy simulation （LES） by coupling with a sub-particle-scale （SPS） turbulence model. The SPS turbulence model turns into the Reynolds stress terms in the filtered momentum equation, and the Smagorinsky model is introduced to describe the Reynolds stress terms. Although MPS method has the advantage in the simulation of the free surface flow, a lot of non-free surface particles are treated as free surface particles in the original MPS model. In this paper, we use a new free surface tracing method and the key point is ＂neighbor particle＂. In this new method, the zone around each particle is divided into eight parts, and the particle will be treated as a free surface particle as long as there are no ＂neighbor particles＂ in any two parts of the zone. As the number density parameter judging method has a high efficiency for the free surface particles tracing, we combine it with the neighbor detected method. First, we select out the particles which may be mistreated with high probabilities by using the number density parameter judging method. And then we deal with these particles with the neighbor detected method. By doing this, the new mixed free surface tracing method can reduce the mistreatment problem efficiently. The serious pressure fluctuation is an obvious defect in MPS method, and therefore an area-time average technique is used in this paper to remove the pressure fluctuation with a quite good result. With these improvements, the modified MPS-LES method is applied to simulate liquid sloshing problems with large deforming free surface. Results show that the modified MPS-LES method can simulate the large deforming free surface easily. It can not only capture the large impact pressure accur
基金financially supported by the National Natural Science Foundation of China(Grant Nos.51479116 and 11272213)
文摘A numerical model has been developed to study sloshing of turbulent flow in a tank with elastic baffles. The Moving-Particle Semi-implicit method(MPS) is a kind of meshless Lagrangian calculation method. The large eddy simulation(LES) approach is employed to model the turbulence by using the Smagorinsky Sub-Particle Scale(SPS)closure model. This paper uses MPS-FSI method with LES to simulate the interaction between free surface flow and a thin elastic baffle in sloshing. Then, the numerical model is validated, and the numerical solution has good agreement with experimental data for sloshing in a tank with elastic baffles. Furthermore, under external excitations,the MPS is applied to viscous laminar flow and turbulent flow, with both the deformation of elastic baffles and the wave height of the free surface are compared with each other. Besides, the impact pressure with/without baffles and wave height of free surface are investigated and discussed in detail. Finally, preliminary simulations are carried out in the damage problem of elastic baffles, taking the advantage of the MPS-FSI method in computations of the fluid–structure interaction with large deformation.
基金supported by the National Natural Science Foundation of China (11132005 and 50706021)
文摘The micro- and macro-time scales in two-phase turbulent channel flows are investigated using the direct nu- merical simulation and the Lagrangian particle trajectory methods for the fluid- and the particle-phases, respectively. Lagrangian and Eulerian time scales of both phases are cal- culated using velocity correlation functions. Due to flow anisotropy, micro-time scales are not the same with the theo- retical estimations in large Reynolds number （isotropic） tur- bulence. Lagrangian macro-time scales of particle-phase and of fluid-phase seen by particles are both dependent on particle Stokes number. The fluid-phase Lagrangian inte- gral time scales increase with distance from the wall, longer than those time scales seen by particles. The Eulerian inte- gral macro-time scales increase in near-wall regions but de- crease in out-layer regions. The moving Eulerian time scales are also investigated and compared with Lagrangian integral time scales, and in good agreement with previous measure- ments and numerical predictions. For the fluid particles the micro Eulerian time scales are longer than the Lagrangian ones in the near wall regions, while away from the walls the micro Lagrangian time scales are longer. The Lagrangian integral time scales are longer than the Eulerian ones. The results are useful for further understanding two-phase flow physics and especially for constructing accurate prediction models of inertial particle dispersion.
