Siren Design and Evaluation in Mud Flow Loops and Wind Tunnels - Measurement While Drilling Signal Analysis, Optimization and Design - page 233

9.1.5 Frequently asked questions.
Here we address common questions and issues that concern potential uses
and misuses of wind tunnel modeling.
x To which fluids does wind tunnel modeling apply? The petroleum industry
correctly emphasizes the non-Newtonian nature of drilling muds and
cements versus Newtonian fluids like air and water; their rheological
properties are, of course, important when studying losses in drillpipes and
annuli. However, for high Reynolds number turbine and siren torque
analysis, and for erosion analysis in turbines, sirens and surface strainers,
rheology plays a secondary role and air is suitable as the modeling fluid.
For example, in Chapter 8, actual turbine torque data collected in mud flow
loops and wind tunnels fall on identical performance curves when properly
normalized. In addition, direct post-mortem evaluations conducted by this
author have demonstrated almost identical metal erosion patterns (for the
same test hardware, e.g., turbines, sirens, flow strainers, proppant transport
tools, stabilizers) whether the flowing mud is water, oil or emulsion based.
Properties like pressure drop and skin friction, affected by flow parallel to
the oncoming mud, are rheology-dependent, and cannot be properly
modeled with air - recourse to computation or mud experiment is
necessary. Again, we emphasize that turbine torque and power in air
(arising from transverse flow) accurately simulates that for mud flows at
any speed. This is also true of flows past sirens. Note that it may not
always be possible to model foam flows because of their highly
compressible, two-phase, non-Newtonian nature, and that any conclusions
must be determined on a case-by-case basis.
x Can MWD signal strength be studied using static ' p measurements in a
short wind tunnel? Several companies have characterized MWD signal
strength by measuring the pressure drop (or, “'p”) across stationary pulsers
and sirens, from upstream to downstream, in mud flow loops, assuming that
these 'p's are representative of signal strength. Such datasets are then
incorporated in telemetry planning software used at the rigsite. This
procedure is completely incorrect because such 'p measurements bear no
relation to the acoustic signals that travel uphole. Why? The flow past a
stationary cement block, for example, is associated with a high pressure
drop, but this does not mean it is useful for high-data-rate sound
transmission. Of course, the block may turn slowly, causing 'p to change
slowly, but this represents a slow change to hydrostatic pressure. This slow
change is detectable uphole, but it is not useful for high speed telemetry.
True sound transmissions can only be effected by rapid movement, e.g., the
action of a piston in a pipe, the vibrations of a tuning fork in a room, the
propagation of sound in acoustic mufflers and, of course, the rapid opening