7
relating to the metal-mould interface, phase changes and the temperature dependence of
all thermophysical data [21], [22].
2.6.1.1 Equation of State
To account for the change of phase from liquid to solid in a binary alloy, each
computational cell contains a fraction of solid (fs ) and a fraction of liquid (fl ), where
the sum of fractions must equal unity. In theory, the fraction of solid should be
calculated from the nucleation and growth of metal dendrites. In practice, this
simulation is too complex. Instead a number of approximations may be used [23]:
Linear: This equation is based on the assumption that solidification occurs at a
constant rate across the freezing range. Although this assumption does not model any
real system, it is used when the physical properties of the metal are unknown.
T TS
fl ..2.1
Tl TS
Lever Rule: This equation is based on the assumption of equilibrium freezing.
CO ml k (T Tm )
fl
(1 k )(T Tm ) .2.2
Scheil equation: This equation is based on the assumption that there is no diffusion in
the solid state, and thus, the composition of the solid that forms in the initial stages of
freezing remains unchanged.
T Tm 1/( ko 1)
fl ( ) 2.3
ml CO
where Co is the composition, ko the partition ratio, ml the liquidus slope, Tm the pure
metal melting point, Ts the solidus, and Tl the liquidus temperature. Since each rule
depends on a different physical assumption about the metal, the computer model should
be written in such a manner that it allows the user to select the appropriate rule
whenever necessary. [24]
8
2.6.1.2 Equation of Latent Heat
To account for the latent heat lost during solidification, the latent heat (L) must
be included, in some manner, into the energy equation. Typically, this is done as a
function of the fraction solid [25], [26], [27]:
Source term: The latent heat is added to the energy equation as a source term. This
method has a tendency for the solution to oscillate across the phase change. This
behaviour is called ringing. The source term is represented as:
f S f S T
S= L = L ..2.4
t T t
Effective specific heat: The latent heat is added to the energy equation by using an
effective specific heat. It has been found that this method tends to be inefficient.
f l
CPK (T ) CP (T ) L .2.5
t
Sensible enthalpy: The latent heat is added to the energy equation by assigning a nodal
latent heat value to each computational cell according to the temperature of a cell. Upon
changing phase, the nodal latent heat of the cell is adjusted to account for latent heat
evolution. [28].
T
h= C
0
P (T )T Lf l 2.6
As each method is able to account for the loss of latent heat, it is more important to
consider the computational efficiency, than physical meaning, when choosing between
them.
2.6.1.3 Equation of Energy
The basic equation for heat conduction within ingot and mold, considering an
internal source of heat is presented below [29]:
T
[ k (T ).T ]+ Q = .C P .(T ).( ) .2.7
t
9
where;
k: thermal conductivity
t: time
Q : heat flux
CP : specific heat
T : temperature
For three dimensional heat conduction, this equation may be written:
2T 2T 2T T
k[( 2 ) ( 2 ) ( 2 )] Q .C P .( ) ..2.8
x y z t
2.6.2 Boundary Conditions
There are three such boundaries, each requiring special attention:
2.6.2.1 Wall Boundaries
The interface that occurs between the mould wall and the molten metal is called the
wall boundary. The energy boundary condition at this interface is defined by the heat
transfer coefficient between the molten metal and the mould material. At the very
beginning of solidification, when molten metal near the liquidus temperature first
comes in contact with the mold surface, heat transfer is controlled at the mold-metal
interface. The main parameter that models this phenomenon is the interfacial heat
transfer coefficient (HTC). The HTC can be expressed as follows.
qm / c hm / c (Tc Tm ) 2.9
where q is the heat flux per unit area, hm / c is the mold/casting interfacial coefficient,
T is the temperature at the casting side of the interface and Tm is the temperature at the
mold side of the interface. This mode of heat transport is commonly used when talking
about interface and convection. But once the metal begins to solidify at the
mold/casting interface it is no longer a fluid. Once the metal arrives at the liquidus
temperature, solidification will commence via a nucleation and growth process. Note
that throughout this process, the temperature profiles in the mold and metal never match
10
at the interface. This can be physically explained by discussing the small air gap, or
discontinuity, that exists at the mold metal interface throughout the entire solidification
process. This air gap exists due to metal contraction upon solidification and the gap size
depends strongly upon the casting geometry. If there was no air gap, there would be an
intimate contact between mold and metal and the HTC would be a higher number than
if there was a large air gap. In reality, an air gap exists to some degree, and
consequently, quite a low HTC exists compared to a casting-chill contact. The question
is, whether the HTC is low enough that it controls heat transfer, or the thermal
conductivity of the mold is the controlling factor. One can observe and quantify this
"competition for control" through the non-dimensional Biot (Bi) number.
The Biot number is given as
hl
Bi= .2.10
k
where h is the interfacial heat transfer coefficient, k is the thermal conductivity of the
mold, and l is the perpendicular distance from the mold metal interface [30], [31].
2.6.2.2 Surface Boundaries
The interface between the liquid metal and air is called the surface boundary. The
energy boundary condition at this interface is defined by a heat transfer coefficient
between the metal and the air. This is usually calculated from the sum of the radiation
and convective heat losses [32], [33].
2.6.2.3 External Boundaries
This boundary corresponds to a typical interface between the mould and the
surrounding air. The heat transfer coefficient is determined from the sum of the
convection and radiation losses. Fluid flow across this boundary is not possible [33].
11