Fluid Dynamics Equations 179
4.4.2. Geometric boundary conditions
4.4.2.1.
Solid walls
The usual solid walls are relatively easy to treat, since the fluid touches the wall:
the velocities of the solid wall and the fluid are equal where they are in contact
. The
reason for this fact is related to the roughness of the walls on the molecular level,
and to the thermal excitation by which mean momentum is transferred from one
medium to another (it is incidentally the same as the interpretation of viscosity and
contact action). This interpretation is such that the physico-chemical nature of the
wall does not influence the adherence condition.
This adherence condition of the fluid at the wall is always very well satisfied in
ordinary conditions where the mean free path of the molecules is small compared
with the roughness. The same is not true in the study of rarefied gases, where we
must take account of the properties of the wall and introduce a slipping coefficient.
In certain cases, the walls are permeable, in other words they let some matter
pass through them. This is the case when we suck or blow through a porous
medium. In these conditions, the difference between the tangential fluid velocities
and the wall are zero at the wall. The normal fluid velocity with respect to the wall
depends on the fluid injection process. An analogous condition is encountered in the
presence of
phase changes at the wall
: evaporation, fusion, and other heterogenous
chemical reactions which consume or produce fluid.
The boundary conditions for thermal and diffusion problems were discussed in
sections 2.3.2 and 2.4.6, which the reader can refer to.
4.4.2.2.
Flow entry and exit zones
In addition to the walls which guide the flow, we generally need to specify the
conditions at the entrance or exit of the domain studied. Real flows are always
generated by machines (solid surface in movement) or by differences in conditions
between upstream and downstream reservoirs. In practice, we know how to impose
wall velocities, injection or extraction flow rates, unsteady forces on a wall (by
means of electromagnetic devices). However, it should be noted that we do not
know how to impose a given pressure or velocity distribution throughout a fluid.
In fact, the entrance of a flow into a domain is a rather particular zone, as we
have an
initial condition
which is largely analogous to that of the temporal variable
(it is in fact an initial condition in Lagrangian variables). In practice, such a zone is
found at the exit of another fluid domain and we need to specify a velocity or
pressure distribution which is compatible with the equations of motion. This is
straightforward when we can take a uniform flow or zone at rest with a constant