Reference Manual

6−9
Figure 6-7. Cavitrol III Trim
W2479
Figure 6-8. NotchFlo DST Trim for
Fisher Globe Style Valves
W8538
However, if cavitation damage can be eliminated
using the special trims discussed, it becomes
practical to use the path treatment method to
further reduce the local noise caused by the
cavitating liquid. This may be accomplished
through the use of heavy-walled pipe and
acoustical or thermal insulation.
Much technology now exists for predicting and
controlling noise in the industrial environment.
Prediction techniques accurately alert the designer
to the need for noise control. When it is a problem,
a variety of solutions are available ranging from
simple path insulation to sophisticated control
valves which eliminate noise at the source.
Two-Phase Noise
As the properties of the fluids vary, the noise
generation, propagation, and pipe excitation
processes area are all affected. Acoustical wave
speed and the density of the fluid are key
considerations. In an all gas or all liquid
application, these are reasonably predictable at
any point from the inlet of the downstream piping.
However, for a multiphase fluid, either
one-component or two-component, there can be
tremendous variations in these important
parameters. In fact, at the vena contracta where
the velocities are greatest, the phases may
separate and form annular flow, with the gas and
the liquid phases having different velocities. This
possibility makes the noise generation process
nearly impossible to model.
Between the vena contracta and the downstream
piping, the phases may be re-oriented to a
homogenous mixture. Propagation of a pressure
wave in this region would be again nearly
impossible to model, as even if it is perfectly
homogenous, the void fraction would be constantly
changing with pressure.
Wave speed and density are also important in
determining the efficiency with which a sound field
is coupled to the pipe wall to cause vibrations and
subsequent external noise radiation.
Emerson engineers have conducted field studies
on applications where flashing noise was present
in an attempt to quantify the problem, if indeed
there was one. After an extensive search there
were not any applications which were considered
noise problems, nor have any surfaced since.
Based upon this experience, two conclusions were
made:
1. 1. A technically appropriate two-phase noise
prediction method does not exist
2. Two-phase, or pure flashing, applications do
not create noise problems.