Modeling and numerical simulations of dendritic crystal growth
Abstract
A simple phase field model for one component melt growth is presented, which includes anisotropy in a certain form. The formation of various dendritic patterns can be shown by a series of numerical simulations of this model. Qualitative relations between the shapes of crystals and some physical parameters are discussed. Also it is shown that noises give a crucial influence on the side branch structure of dendrites in some situations.
References (22)
- L.N. Brush et al.
J. Crystal Growth
(1989) - S. Osher et al.
J. Comput. Phys.
(1988) - H. Honjo et al.
J. Crystal Growth
(1982) - S.-C. Huang et al.
Acta Metal.
(1981) - D.A. Kessler et al.
Phys. Rev. A
(1985) - E. Ben-Jacob et al.
Phys. Rev. A
(1984) - Y. Saito et al.
Phys. Rev. A
(1988) - G. Fix
- J.B. Collins et al.
Phys. Rev. B
(1985) - G. Caginalp
Arch. Rat. Mech. Anal.
(1986)
SIAM J. Appl. Math.
Cited by (1290)
Unconditional error analysis of linearized BDF2 mixed virtual element method for semilinear parabolic problems on polygonal meshes
2024, Journal of Computational and Applied MathematicsIn this paper, we construct, analyze, and numerically validate a class of -mixed virtual element method for the semilinear parabolic problem in mixed form, in which the parabolic problem is reformulated in terms of the velocity and the pressure of the time-dependent Darcy flow. The Newton linearized method for the nonlinear term is designed to cooperate with the second-order backward differentiation formula of the temporal discretization scheme. This allows each time step to only require the solution of a small and well-structured linear system rather than the solution of a nonlinear system. The linearization improves computational efficiency without decreasing convergence rates. Moreover, the “Fortin” operator and its approximation properties are applied to derive an optimal error estimates for the virtual element solution of the velocity and the pressure. Finally, its remarkable performance is illustrated by several numerical examples that also validate the theoretical rates of convergence.
Phase-field simulation of magnesium-metal anode interface growth in different electrolytes during electrodeposition
2024, Journal of Energy StorageVarious types of Mg electrolytes have emerged with the deepening and development of research on Rechargeable Magnesium-Metal Batteries. In order to assess the potential of these electrolytes to inhibit the growth of Mg dendrites, we apply a Mg-electrodeposition phase-field model with surface energy anisotropy to simulate the interface growth processes of Mg-metal anodes in different electrolytes during electrodeposition. The results show that the conductivity and Mg-ion diffusivity of the Mg electrolyte are key factors affecting the interface growth, with the diffusivity having a particularly significant effect. In the designed 1.0 M MgX/THF electrolyte with an applied overpotential of −0.30 V, by profoundly analyzing the fluctuation of the interface contour, we find that when the diffusivity reaches 1.4 × 10−10 m2/s, the Mg-metal anode interface can achieve stable growth.
Phase-field modeling of zinc dendrites growth in aqueous zinc batteries
2024, International Journal of Heat and Mass TransferThe growth of zinc dendrites poses a major challenge in developing stable aqueous zinc batteries. It is crucial to understand the mechanism underlying dendrite growth and the influences of various factors on the process of dendrite growth. In this study, a phase field model is constructed to simulate the growth of zinc dendrite and investigate the temporal and spatial distributions of ions and electric potential in both static and flowing electrolytes. The influences of various critical factors, including overpotential, anisotropic strength and electrolyte flow rate, on the morphology evolution of electrodeposit are investigated. Results reveal that the surface morphology evolution is intimately dependent on the competition between ion transport and the electrochemical reaction, and it can be modulated by various parameters. In particular, flowing electrolyte not only enhances the convective ion transport but also suppresses the dendrite bifurcation on the surface, resulting in a uniform distribution of the concentration gradient and a compact deposit. However, nonuniformity induced by the flowing electrolyte from the inlet to outlet may impede the stable cycling of zinc anodes in the flow batteries. Therefore, further design and optimization of the flow field are desired for zinc-based flow batteries. This work provides insights into the growth process of zinc dendrites and the influences of several critical parameters, which could inform the design of stable aqueous zinc batteries.
Modeling solid air dendrite growth solidification with thermosolutal diffusion using non-isothermal quantitative phase field method
2024, International Journal of Thermal SciencesExposure of air to liquid hydrogen may result in the formation of external oxygen-enriched solid air particles, posing significant safety hazards to the liquid hydrogen system. This study involves the development of a non-isothermal growth model for solid air (air of solid state) dendrites, which is driven by the coupling of thermosolutal diffusion, and is implemented using the quantitative phase field method. The primary aim is to investigate the morphological evolution, growth rate, and solute distribution in solid air dendrites under various thermal conditions. The findings indicate that non-isothermal simulations, considering the latent heat of solidification, offer a more realistic representation of solid air dendrite growth compared to isothermal simulations. Specifically, the non-isothermal simulations show a significant reduction in dendrite tip growth rate by 34.14 % and a peak oxygen concentration decrease of 14.6 % compared to isothermal simulations. When the initial subcooling is set at 4 K, the solid-liquid mushy zone exhibits a maximum oxygen concentration of 0.403. Boundary heat transfer drastically alters the temperature distribution in the computational domain, thereby affecting the growth pattern of solid air dendrites. Boundary heat transfer significantly alters the lengths of dendritic arms near that boundary. Specifically, heating causes a reduction in arm lengths by 30.74 %, while cooling induces an increase of 143.29 % compared to the case without any boundary heat flux. Under heating conditions, the diffusion capacity of the oxygen solute increases with temperature. However, simultaneously, the growth of dendrite arms is inhibited, causing less oxygen solute to be released into the liquid phase. As a result, there is a significant decrease in oxygen concentration in the calculated domain. This effect becomes more pronounced as the boundary heat flux increases, with a 33.26 % reduction in dendrite arm length and a 10.64 % reduction in oxygen concentration for a heating with 162 kW/m2. The opposite effect is observed during cooling.
Modeling of rapid solidification in Laser Powder Bed Fusion processes
2024, Computational Materials ScienceThis paper addresses a numerical framework for modeling site-specific microstructure evolution in Laser Powder Bed Fusion (LPBF), where rapid solidification significantly influences the microstructure and properties of manufactured components. The fine microstructure and rapid solidification can enhance mechanical properties in final parts, yet pose challenges in controlling microstructure and addressing potential defects. In this study, an in-house developed free surface thermo-fluidic Computational Fluid Dynamics (CFD) model is coupled with a binary alloy phase-field model characterized by rapid solidification kinetics. The CFD model considers complex fluid flows related to Marangoni and recoil pressure, laser-material interactions, and rapid melting and solidification of the melt pool. A convergence study for the phase-field model is conducted, comparing simulated partition coefficients with the Continuous Growth Model (CGM) theory. The study investigates the individual effects of temperature gradients and solidification rates on simulated microstructure. Results indicate that, for the same order of magnitude change, the solidification rate has a more pronounced effect on microstructure than temperature gradients. The paper concludes with the simulation of site-specific microstructure evolution within the melt pool, comparing the results with experiments conducted at varied scanning speeds. The predicted microstructure aligns well with experimental measurements, suggesting the potential utility of the developed phase-field model in quantitative microstructural modeling. Overall, this work contributes to understanding microstructure evolution under rapid solidification conditions, providing a fundamental basis for optimizing the LPBF process.
An efficient numerical method for the anisotropic phase field dendritic crystal growth model
2024, Communications in Nonlinear Science and Numerical SimulationIn this paper, we propose and analyze an efficient numerical method for the anisotropic phase field dendritic crystal growth model, which is challenging because we are facing the nonlinear coupling and anisotropic coefficient in the model. The proposed method is a two-step scheme. In the first step, an intermediate solution is computed by using BDF schemes of order up to three for both the phase-field and heat equations. In the second step the intermediate solution is stabilized by multiplying an auxiliary variable. The key of the second step is to stabilize the overall scheme while maintaining the convergence order of the stabilized solution. In order to overcome the difficulty caused by the gradient-dependent anisotropic coefficient and the nonlinear terms, some stabilization terms are added to the BDF schemes in the first step. The second step makes use of a generalized auxiliary variable approach with relaxation. The Fourier spectral method is applied for the spatial discretization. Our analysis shows that the proposed scheme is unconditionally stable and has accuracy in time up to third order. We also provide a sophisticated implementation showing that the computational complexity of our schemes is equivalent to solving two linear equations and some algebraic equations. To the best of our knowledge, this is the cheapest unconditionally stable schemes reported in the literature. Some numerical examples are given to verify the efficiency of the proposed method.