|
|
|
[Contents] | [Index] |
Cooperation with: M. Dreyer (Universität Bremen, ZARM), M. Adamov, M. Lozano Avilés (Technische Universität Berlin)
Supported by: BMBF-DLR: ``Treibstoffverhalten in Tanks von Raumtransportsystemen -- Comportement des Ergols dans les Réservoirs''
Description: The prediction of the dynamic behavior of liquids with free capillary surfaces in partly filled containers is of importance for the construction of space vehicles using liquid propulsion. In the case of cryogenic propellants, this behavior is influenced to a large extent by thermal effects, such as evaporation. As a first step for simulations, a model for two-phase flow with evaporation has been developed.
To model the situation mathematically, balance equations for mass, momentum, and energy in the bulk of the phases and on the phase boundary, respectively, were established, assuming gas and liquid to be incompressible Newtonian fluids and using the Boussinesq approximation to model buoyancy. In the bulk of both phases this leads to the Navier-Stokes equations with convection-diffusion equations for heat and vapor:
[![]() ![]() |
= | 0, | (5) |
[![]() ![]() |
= | - (![]() ![]() ![]() ![]() ![]() |
(6) |
[![]() ![]() ![]() ![]() ![]() ![]() |
= | j![]() |
(7) |
![]() ![]() ![]() ![]() ![]() |
= | j. | (8) |
Further conditions are needed to determine the evaporation rate j.
Thus, we add the assumption that the temperature is
continuous and is always equal to the saturation temperature
given by the partial pressure of vapor
(
,
) on the phase boundary. Note that in a
precise physical sense this assumption is contradictory to a
non-vanishing evaporation rate, since it is the statement of
equilibrium. However, the difference
-
is
negligible in many practical cases. This assumption yields a Dirichlet
boundary condition for the temperature in both phases that is
sufficient to solve the heat transport equation
(3). Equation (7) may then be
used to calculate j. This approach results in a fully coupled
two-phase flow problem.
In general, viscosities in gas are much smaller than in liquid and
thus
Tgn Tln. Using
this assumption, one may decouple the flow problems from both phases by
neglecting the shear stresses from the gaseous phase on the boundary.
The flow from the gaseous phase does not have any direct influence on
the shape of the free surface then. The flow problems in the two phases are only
weakly coupled by the temperature and the mass flux.
References:
|
|
|
[Contents] | [Index] |