Siren Design and Evaluation in Mud Flow Loops and Wind Tunnels - Measurement While Drilling Signal Analysis, Optimization and Design - page 259

Of course, the mud siren, with its piston-like action and orifice similarities,
will have a pressure dependence that is linear with density and varies with “DV
+ EV 2 ” in a way determined experimentally. From a modeling perspective, one
hopes that a simpler relation can be found, so that “DV” or “EV 2 ” alone holds.
The dependence on the sound speed c is not obvious, and recourse to mud loop
testing may be necessary after wind tunnel test data have been evaluated. Care
should be exercised with the extrapolation procedures defined above.
9.4 Long Wind Tunnel for Telemetry Modeling
We recapitulate the ideas developed so far. Short wind tunnels provide
inexpensive, fast-turnaround tests for properties like torque, erosion, stable-open
versus stable-closed, and turbine no-load rotation rate, torque and power; they
are used to evaluate hydraulic effects associated with constant density flows.
Intermediate wind tunnels recognize the wave propagation, reflective and
reverberant aspects of siren-induced wave motions, which arise from
compressible fluid effects. They too, are inexpensive, providing simple means
to determine signal strength. When integrated with dynamic torque and
azimuthal position measurements, valuable data can be obtained that will be
useful in subsequent feedback and control loop design. We had discussed the
effects wave interference. These are both good and bad. For instance, it can be
used to enhance MWD signals without incurring the usual power and erosion
penalties. Alternatively, less-than-optimal placement of the siren in the collar,
or the use of inappropriate frequencies, can reduce the signals that ultimately
travel up the drillpipe by virtue of destructive interference.
9.4.1 Early construction approach - basic ideas.
To evaluate wave interactions, it is not possible to use our short and
intermediate wind tunnels. Instead, a “long wind tunnel” must be used, which
provides enough “leg room” for concept evaluation. But how long is long? The
sound speed in air is approximately 1,000 ft/sec. Suppose we consider a
maximum signaling frequency of 100 Hz. Then, the wavelength associated with
this motion is (1,000 ft/sec)/(100 Hz) or 10 ft. Thus, a total length of 2,000 ft
should suffice for most wave evaluation purposes. Because 10 ft still greatly
exceeds a typical diameter, the waves are long in an acoustic sense. Therefore
they will not reflect at bends, even ninety-degree bends (of course, the hydraulic
pressure drops required to move air at prescribed volume flow rates may be
large). A long wind tunnel built with steel tubing wound several times around a
building is shown in Figure 9.2b. But since reflections are almost non-existent,
as can be validated experimentally, the long wind tunnel can also be constructed
from flexible plastic tubing wound on 3-4 ft diameter reels, as in Figure 9.12a,
for an early “concept” long wind tunnel that is no longer in use. Preliminary
studies in this facility and an understanding of its limitations led to the CNPC
design described extensively in this topic.