Lattice Boltzmann Modeling Of The Evolution Of Condensate Liquid Droplets And Solidification Gas Pores

The cognition of gas-liquid-solid interfacial wetting phenomena is necessary for deeply understanding the wettability of superhydrophobic nanostructured surface and the mechanism of solidification gas pore formation.Superhydrophobic nanostructured surface is an important type of bioinspired materials and widely applied in industry and daily life.On the other hand,microporosity is one of major defects in castings.The morphology and distribution of microporosity in castings influence the mechanical properties of metal products significantly.The lattice Boltzmann method(LBM)has some special advantages in the simulations of the wettability of nanostructured surfaces and the formation of solidification microporosity.The cellular automaton-finite difference method(CA-FDM)is an important approach for the numerical modeling of solidification microstructures.In this thesis,the phenomena of droplet condensation and evaporation on nanostructured rough surfaces are simulated using the multiphase LBM models;a coupled CA-FDM-LBM model is developed for the simulation of dendrite growth and gas pore formation during solidification of alloys;a CA model is applied for the simulations of temperature gradient zone melting(TGZM)and dendrite coarsening in the musy zones of alloys without gas pores.Through the numerical modeling of these problems,some key information and physical mechanisms that are difficult to be detected in experiments can be revealed.Several types of two and three dimensional(2D and 3D)multiphase LBM models are employed to simulate the processes of droplet condensation and evaporation on nanostructured rough surfaces.Validations of the multiphase LBM models are performed through the tests of Laplace'law and contact angle theories.The 3D multiphase LBM model is implemented with the GPU-CUDA accelerated algorithm.The computation efficiency of the GPU-CUDA accelerated model is around 300 times higher than that of the CPU-based model when the computational domain is composed of 256~3 computational nodes.Three typical preferential nucleation modes of condensate droplets are observed through the LBM simulations of droplet condensation on the various nanostructured rough surfaces.The droplets nucleated at the top of posts(top nucleation)or in the upside interpost space of nanostructures(side nucleation)will generate the Cassie state,while the ones nucleated at the bottom of the nanostructures(bottom nucleation)produce the Wenzel state.The simulated wetting states of condensate droplets on the nanostructures are identical with the experimental observations.The relationship between the geometrical parameters of the nanostructures and the preferential nucleation modes of condensate droplets is established from the simulations.It is demonstrated that the nanostructures with taller posts and a high ratio of post height to interpost space are beneficial to produce the top and side nucleation modes.In addition,the simulation results show that the nucleation mode and final wetting state of the condensate droplets can change by setting heterogeneities of geometrical parameters and local intrinsic wettability of the nanostructures.The behavior of droplet evaporation is simulated.It is found that the shapes and contact angles almost remain stable for the evaporating droplets on the smooth surfaces,but display obvious oscillations for those presenting the Wenzel and Cassie states on the nanostructured rough surfaces.The Cassie-to-Wenzel wetting state transition takes place in the late stage of the evaporation for the droplets that initially display the Cassie state on the nanostructured surfaces.The simulations of droplet condensation and evaporation provide guidance for the design of nanostructured rough surfaces with desirable superhydrophobic properties.The coupled CA-FDM-LBM model is applied to simulate the processes of equiaxed/columnar dendrite growth and gas pore formation during solidification of an Al-4 wt.%Cu alloy.The CA-FDM-LBM model is capable to reasonably describe the evolution of hydrogen concentration,natural nucleation and the subsequent growth,coalescence,movement of the gas pores,as well as the interaction between the gas pores and the dendrites,in the processes of columnar/equiaxed dendrite growth.The simulated morphology of the microporosity compares well with the experimental micrograph.The simulations exhibit the incubation period before the gas pores emerge.In the incubation stage,the hydrogen concentration in liquid gradually increases.The gas pore nuclei caused by natural nucleation preferentially appear at the roots of the dendrite arms when the hydrogen concentration in liquid exceeds the supersaturation for porosity nucleation.With the increase of the cooling rate and the decrease of the initial hydrogen concentration in the melt,the temperature of porosity nucleation,the porosity fraction and the mean porosity size decrease.The evolution of columnar dendrite growth and gas pore formation is simulated in 2D and 3D to reveal the similarities and differences.The simulations of dendrite growth and gas pore formation using the CA-FDM-LBM coupled model developed in this thesis offer a useful tool for optimizing the casting process for aluminum alloys.A CA model including the solidification and melting mechanisms is applied to study the migrations of liquid pools and secondary dendrite arms due to TGZM in the musy zones of alloys without gas pores.The influences of pulling velocity and temperature gradient on TGZM dynamics are investigated.The simulations show that the time-averaged velocities of secondary arm migration increase with increasing temperature gradient.The phenomenon of TGZM in the mushy zone for an Al-7 wt.%Cu alloy with initially equiaxed microstructures is simulated.The simulations exhibit liquid inclusions migrating through the grains and along grain boundaries towards the high temperature direction.During holding in a temperature gradient,the mean concentration in the mushy zone decreases towards the high temperature direction.The simulation results are compared with the relevant experiments and analytical predictions.The phenomena of dendrite coarsening for a SCN-2 wt.%ACE alloy during isothermal holding are simulated using the CA model.Some typical coarsening mechanisms are characterized by comparing the local equilibrium and actual compositions at the solid-liquid interfaces.The effect of melting on dendrite coarsening is investigated in the CA simulations.The CA simulations are able to quantify the evolution of interfacial curvature,local equilibrium composition,and local actual composition in the process of microstructural evolution,providing scientific guidance for understanding the complex interactions among local temperature,solidification/melting,interface shape variation and solute diffusion.