Page 235 - Introduction to Computational Fluid Dynamics
P. 235

P1: IBE
                           CB908/Date
                                        0 521 85326 5
            0521853265c07




                            7 Phase Change                                         May 25, 2005  11:14










                            7.1 Introduction

                            There is hardly a product that, during its manufacture, does not undergo a process of
                            melting and solidification. Engineering processes such as casting, welding, surface
                            hardening or alloying, and crystallisation involve phase change. The processes of
                            freezing and thawing are of interest in processing of foods. Phase-change materials
                            (PCMs) are used in energy storage devices that enable storage and retrieval of
                            energy at nearly constant temperature.
                               The phenomenon of melting or solidification is brought about by a process
                            of latent heat (λ) transfer at the interface between solid and liquid phases. For a
                            pure substance, throughout this process, the temperature T m (melting point) of the
                            interface remains constant whereas in the liquid and solid phases, the temperatures
                            vary with time. Both λ and T m are properties of a pure substance. Within each of
                            the single phases, heat transfer is essentially governed by a process of unsteady heat
                            conduction, although, under certain circumstances, convection may also be present
                            in the liquid phase under the action of body (buoyancy, for example) or surface
                            (surface tension) forces.
                               There are two approaches to solving phase-change problems:

                            1. the variable domain formulation and
                            2. the fixed domain (or fixed-grid) formulation.
                            In the first approach, which has several variants, two energy equations are solved
                            in the solid and the liquid phases with temperatures T s and T l , respectively, as
                            dependent variables. In addition to the initial (i.e., at t = 0) and the domain bound-
                            ary conditions, the following interface conditions are also invoked to match the
                            temperatures of the two phases:
                                                          T s = T l = T m ,                     (7.1)

                                                      ∂T s      ∂T l
                                                    k s      − k l     = ρλV i ,                (7.2)
                                                       ∂n       ∂n
                                                           i       i
                     214
   230   231   232   233   234   235   236   237   238   239   240