The term “subcooled boiling” is used to describe the physical situation where the wall temperature is high enough to cause boiling to occur at the wall even though the bulk volume averaged liquid temperature is less than the saturation value. In such cases, the energy is transferred directly from the wall to the liquid. Part of this energy will cause the temperature of the liquid to increase and part will generate vapor. Interphase heat transfer will also cause the average liquid temperature to increase, however, the saturated vapor will condense. Additionally, some of the energy may be transferred directly from the wall to the vapor. These basic mechanisms are the foundations of the so called Rensselaer Polytechnic Institute (RPI) models.
In ANSYS Fluent, the wall boiling models are developed in the context of the Eulerian multiphase model. The multiphase flows are governed by the conservation equations for phase continuity (Equation 17–167), momentum (Equation 17–168), and energy (Equation 17–171). The wall boiling phenomenon is modeled by the RPI nucleate boiling model of Kurual and Podowski [245] and an extended formulation for the departed nucleate boiling regime (DNB) by Lavieville et al [258].
The wall boiling models are compatible with three different wall boundaries: isothermal wall, specified heat flux, and specified heat transfer coefficient (coupled wall boundary).
Specific submodels have been considered to account for the interfacial transfers of momentum, mass, and heat, as well as turbulence models in boiling flows, as described below.
To learn how to set up the boiling model, refer to Including the Boiling Model.
According to the basic RPI model, the total heat flux from the wall to the liquid is partitioned into three components, namely the convective heat flux, the quenching heat flux, and the evaporative heat flux:
(17–407) |
The heated wall surface is subdivided into area , which is
covered by nucleating bubbles and a portion
, which is covered by the fluid.
The convective heat flux is expressed as
(17–408) |
where is the single phase heat transfer
coefficient, and
and
are the wall and liquid
temperatures, respectively.
Where is the conductivity,
is the periodic
time, and
is the diffusivity.
The evaporative flux is given by
(17–410) |
Where is the volume of the bubble
based on the bubble departure diameter,
is the active nucleate
site density,
is
the vapor density, and
is the latent heat of evaporation, and
is the bubble departure frequency.
These equations need closure for the following parameters:
Area of Influence
Its definition is based on the departure diameter and the nucleate site density:
(17–411) |
Note that in order to avoid numerical instabilities due to unbound empirical correlations for the nucleate site density, the area of influence has to be restricted. The area of influence is limited as follows:
(17–412) |
The value of the empirical constant is usually set to 4, however it
has been found that this value is not universal and may vary between
1.8 and 5. The following relation for this constant has also been
implemented based on Del Valle and Kenning's findings [101]:
(17–413) |
and is
the subcooled Jacob number defined as:
(17–414) |
where
Frequency of Bubble Departure
Implementation of the RPI model normally uses the frequency of bubble departure as the one based on inertia controlled growth (not really applicable to subcooled boiling). [87]
(17–415) |
Nucleate Site Density
The nucleate site density is usually represented by a correlation based on the wall superheat. The general expression is of the form
(17–416) |
Here the empirical parameters from Lemmert and Chawla [264] are used, where and
. Other formulations are also available, such as
Kocamustafaogullari and Ishii [238] where
(17–417) |
Here
|
| |
|
| |
|
| |
|
|
Where is the bubble departure diameter and the density function is defined
as
(17–418) |
Bubble Departure Diameter
The default bubble departure diameter for the RPI model is based on empirical correlations [487] and is calculated in meters as
(17–419) |
while Kocamustafaogullari and Ishii [237] use
(17–420) |
with being the contact angle in degrees.
The bubble departure diameter in millimeters based on the Unal relationship [495] is calculated as
(17–421) |
(17–422) |
(17–423) |
(17–424) |
where is the flow pressure,
is the wall superheat,
is
latent heat,
is the near wall bulk velocity, and
m/s. The subscripts
,
, and
denote the solid material, liquid, and vapor phase,
respectively.
When using the basic RPI model (RPI Model), the temperature of the vapor is not calculated, but instead is fixed at the saturation temperature. To model boiling departing from the nucleate boiling regime (DNB), or to model it up to the critical heat flux and post dry-out condition, it is necessary to include the vapor temperature in the solution process. The wall heat partition is now modified as follows:
(17–425) |
Here ,
, and
are the liquid-phase
convective heat flux, quenching heat flux, and evaporation heat flux,
respectively (described in detail in RPI Model). The extra heat fluxes are
representing the convective heat flux of the vapor phase, and
representing heat flux to any other possible gas phases in a system.
These can be expressed as
(17–426) |
(17–427) |
Similar to the liquid phase ,
the convective heat transfer coefficients
and
are computed from the wall function formulations.
The function depends on the local liquid volume fraction
with similar limiting values as the liquid volume fraction. Lavieville
et al [258] proposed the following expression:
(17–428) |
Here, the critical value for the liquid fraction is
In wall boiling, the critical heat flux (CHF) condition is characterized by a sharp reduction of local heat transfer coefficients and the excursion of wall surface temperatures. It occurs when heated surfaces are no longer wetted by boiling liquid with the increase of vapor content. At critical heat flux conditions, vapor replaces the liquid and occupies the space adjacent to heated walls. The energy is therefore directly transferred from the wall to the vapor. In turn, it results in rapid reduction of the heat removal ability and sharp rise of the vapor temperature, and most importantly, the wall temperatures. In addition, wall boiling departs from the nucleating boiling regime, and the multiphase flow regime changes from a bubbly flow to a mist flow.
To model the critical heat flux conditions, the basic approach adopted in ANSYS Fluent is to extend the RPI model from the nucleate boiling regime to critical heat flux and post dry-out conditions, while considering the following:
The generalized and non-equilibrium wall heat flux partition
The flow regime transition from bubbly to mist flows
The wall heat partition is defined in the same way as Equation 17–425, with the exception of the function
definition. Here, the function depends on the local liquid/vapor volume
fraction with the same limiting values as the liquid volume fraction,
that is, between zero and one. Lavieville et al. [258] proposed the following expression:
(17–429) |
The critical value for the liquid volume fraction is , and for the vapor phase,
it is
.
There are also some other functions available to define the wall heat flux partition. When defining wall boiling regimes, Tentner et al. [483] suggested the following expression based on the vapor volume fraction:
(17–430) |
Taking the thin film boiling into consideration in the wall heat flux partition, Ioilev et al. [196] used a linear function to extend Equation 17–430 to the critical heat flux condition:
(17–431) |
Where the breakpoints have been set to and
.
In ANSYS Fluent, Equation 17–430 is chosen as the default formulation for the wall heat flux partition.
When wall boiling departs from the nucleate boiling regime and reaches the critical heat flux and post dry-out conditions, the multiphase flow regime changes from a bubbly flow to a mist flow. Consequently, the liquid phase switches from the continuous phase to the dispersed phase, while the vapor phase becomes the continuous phase from the originally dispersed phase in the bubbly flow regime. With the flow regime transition, the interfacial area, momentum transfer terms (drag, lift, turbulent dispersion, interfacial area, and so on), heat transfer and turbulence quantities will change accordingly.
To mimic the change of the flow regime and compute the interfacial transfers, the so-called flow regime maps, based on cross-section averaged flow parameters, are traditionally used in sub-channel one-dimensional thermal-hydraulic codes. In CFD solvers, the concept of flow regime maps has been expanded into a local, cell-based interfacial surface topology to evaluate the flow regime transitions from local flow parameters. The ensemble of all the computational cells with their usually simple local interfacial surface topologies can provide complex global topologies to represent the different flow regimes as the traditional sub-channel flow regime maps.
As a first step, this implementation adopts a simple local interfacial
surface topology to control the transition smoothly from a continuous
liquid bubbly flow to a continuous vapor droplet flow configuration [483], [196]. It assumes that inside
a computational cell, the local interfacial surface topology contains
multi-connected interfaces, and the flow regimes are determined by
a single local flow quantity — the vapor volume fraction :
Bubbly flow topology: the vapor phase is dispersed
in the continuous liquid in the form of bubbles. Typically
Mist flow topology: the liquid phase is dispersed
in the continuous vapor in the form of droplets. Typically
Churn flow: this is an intermediate topology between
the bubbly and mist flow topology, where
The interfacial surface topologies are used to compute the interfacial
area and interfacial transfers of momentum and heat. Introducing to represent interfacial
quantities (interfacial area, drag, lift, turbulent drift force and
heat transfer), then they are calculated using the following general
form:
(17–432) |
Here is computed using
equation Equation 17–430 or Equation 17–431, but with different lower and upper
limits of the breakpoints. Typically, the values of 0.3 and 0.7 are
used and
and
are the interfacial quantities from bubbly
flow and mist flow, respectively. They are calculated using the interfacial
sub-models presented in Interfacial Momentum Transfer and Interfacial Heat Transfer.
It may be noted that in the boiling models, the liquid is usually
defined as the first phase, and the vapor as the second phase. Once
this is defined, it remains unchanged with the flow regime transition.
When and
are calculated, however, the primary or
secondary phases are switched. For
, the liquid is treated as the
primary phase, while the vapor is the secondary phase. Contrary to
this, for
, the vapor becomes the primary phase and
the liquid is the secondary phase.
The interfacial momentum transfer may include five parts: drag, lift, wall lubrication, turbulent drift forces, and virtual mass (described in Interphase Exchange Coefficients). Various models are available for each of these effects, some of which are specifically formulated for boiling flows. Also, user-defined options are available.
The interfacial area can be calculated using either a transport equation or algebraic models as described in Interfacial Area Concentration. For boiling flows, the algebraic formulations are typically chosen. See Interfacial Area Concentration for model details.
By default, Fluent uses the following model for bubble diameter
as a function of the local subcooling, :
(17–433) |
where:
As an alternative, the bubble diameter, , can
be given by the Unal correlation [495] :
(17–434) |
To use the Unal correlation, Equation 17–434, you can use the following scheme command:
(rpsetvar 'mp/boiling/bubble-diameter-model 2)
To return to the default formulation, Equation 17–433, you can use the scheme command:
(rpsetvar 'mp/boiling/bubble-diameter-model 1)
When the flow regime transitions to mist flow, the droplet diameter can be assumed to be constant or estimated by the Kataoka-Ishii correlation [219]:
(17–435) |
| where: | |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
The interfacial drag force is calculated using the standard model described in Interphase Exchange Coefficients (and defined in the context of the interfacial area from Interfacial Area Concentration). As described in Fluid-Fluid Exchange Coefficient, ANSYS Fluent offers several options to calculate the drag force on dispersed phases. For boiling flows, the Ishii model (Ishii Model) is typically chosen, though any of the models listed in Fluid-Fluid Exchange Coefficient are available.
As described in Lift Force, ANSYS Fluent offers several options to calculate the lift forces on the dispersed phases. For boiling flows, this force is important in the nucleating boiling regime. In the RPI model, the Tomiyama model (Tomiyama Lift Force Model) is usually chosen to account for the effects of the interfacial lift force.
As described in Turbulent Dispersion Force, ANSYS Fluent offers several options to calculate the turbulent dispersion force. For boiling flows, this force is important in transporting the vapor from walls to the core fluid flow regions. In the RPI model, the Lopez de Bertodano model (Lopez de Bertodano Model) is usually chosen to account for the effects of the turbulent dispersion force.
As described in Wall Lubrication Force, ANSYS Fluent offers several options to calculate the wall lubrication force on the dispersed phases. For boiling flows, this force can be important in the nucleating boiling regime. In the RPI model, the Antal et al. model (Antal et al. Model) is usually chosen to account for the effects of the wall lubrication force.
In the wall boiling models, the virtual mass force can be modeled using the standard correlation implemented for the Eulerian multiphase model as described in Virtual Mass Force.
As the bubbles depart from the wall and move towards the subcooled region, there is heat transfer from the bubble to the liquid, that is defined as
(17–436) |
where is the volumetric heat transfer coefficient. The heat transfer coefficient can be
computed using either the Ranz-Marshall or Tomiyama models described
in The Heat Exchange Coefficient.
The interface to vapor heat transfer is calculated using the constant time scale return to saturation method [258]. It is assumed that the vapor retains the saturation temperature by rapid evaporation/condensation. The formulation is as follows:
(17–437) |
Where is the time scale set to a default
value of 0.05 and
is the isobaric heat capacity.
The evaporation mass flow is applied at the cell near the wall and it is derived from the evaporation heat flux, Equation 17–437:
(17–438) |
To model boiling flows, turbulence interaction models are usually included in the multiphase turbulence models to describe additional bubble stirring and dissipation. As described in Turbulence Models, three options are available for boiling flows: Troshko-Hassan (default), Simonin et al, and Sato.