White Paper
FINAL REPORT - Part 2
Prepared by:
Dr. Thomas J. Gieseke
NUWCDIVNPT - Code 8233
March 29, 1999
A combined experimental and numerical approach was taken for this study. Westfall Manufacturing funded the experimental portion (described in this document) while the computational portion was funded using NUWC internal funds (a summary document has not been released). All testing was conducted in the NUWC Transient Flow-Loop Facility (figure 3) at the Naval Undersea Warfare Center Division Newport, RI, building 1246. This facility is a 12,000-gallon recirculating flow-loop capable of providing steady state and transient flows at pressures up to 50 psi and freestream velocities up to 50 ft/s in a 6 inch diameter pipe. The flow velocity is controlled by varying the speed of the 300 hp drive pump or by throttling a large ball valve.
|
|
. Figure 3. Transient Flow-Loop Facility
In its normal configuration, the flow-loop has a rectangular test with internal dimensions 9 x 18 x 93 inches followed by 60 feet of 16 inch diameter pipe and a return elbow. The mixer test section was inserted in the lower leg of the flow-loop downstream of the primary test section (noted in Figure 3) by replacing a section of the 16inch diameter pipe which passed through the control room shown in Figure 3. The mixer test section consisted of a reducer, a length of 6 inch PVC pipe, the mixer, a section of Acrylic pipe, an acrylic box surrounding the pipe, and an expansion section. The tank was filled with water to create a single nominally constant index of volume through which laser measurements inside of the pipe could easily be taken. Figures 4 and 5 are drawings provided by Westfall Manufacturing of the test set-up. Figures 6 through 9 show images of the test arrangement. Westfall Manufacturing provided all the necessary integration hardware for testing while NUWC provided all instrumentation.
Figure 4. Drawing of the test set-up
Figure 5. Drawing of the acrylic box for mixer tests
Figure 6. Image of the mixer test section, traverse system, and LDV system
Figure 7. Image of the Westfall Mixer in the flow-loop.
Figure 8. Image of the NUWC laser system in operation
Figure 9. Alternate view of the mixer test set-up
Two measurement systems were used to monitor the mixer operation. A Validyne differential pressure transducer with a 20psid range was used to monitor the pressure loss across the device and a Laser Doppler Velocimetry System was used to monitor the flow-field down stream of the mixer.
Pressure taps were included in the mixer assembly to monitor the pressure loss across the device. An upstream tap was place 1 diameter upstream of the mixer and a downstream tap was place ½ diameter downstream of the mixer. One quarter inch nylon pressure lines were plumbed between the pressure taps and the Validyne pressure transducer. The voltage output of the transducer was monitored with a Texas Science Instruments integrating voltmeter. One hundred second integration times were used to determine the mean transducer output.
The pressure transducer was calibrated using the following procedure. The pressure line to the downstream side of the mixer was disconnected from the facility, exposing it to air pressure. The facility flow valve downstream of the test section was closed and a reference pressure gauge was installed in the facility at the location of the mixer. The facility pump was operated from rest up to a speed that generated pressures exceeding the anticipated maximum dynamic head during mixer operation. At each operating speed, the transducer output and the reference pressure output were recorded. Figure 10 shows the comparison of the reference gage output and the transducer voltage output. The calibration curve-fit is shown in the figure.
Figure 10. Pressure calibration results
The flow field measurements were taken using a single-component TSI Laser Doppler Velocimetry system. The test arrangement was configured to facilitate measurements in the acrylic pipe. An acrylic box was placed around the pipe and the volume between the pipe and the box was filled with water. The effect was to create a single mass with constant index of refraction, thus permitting light rays to travel undeflected from the outer window surface through the acrylic box, the water between the walls and the pipe, the pipe, and the water in the pipe without significantly deflecting. This configuration enabled relatively unrestricted motion of the LDV measurement volume within the pipe section.
Velocity profile measurements were taken at nine stations downstream of the mixer. At each station, mean and standard deviation streamwise and vertical velocity measurements were conducted. These measurements were taken on a grid spanning one quarter of the pipe. The measurements proceeded as follows:
1) The LDV probe was placed at a prescribed station downstream of the static mixer using the automatic traverse table.
2) The LDV measurement volume (point of beam intersection) was positioned in the core of the mixer wake, at a prescribed vertical position with respect to the centerline of the pipe.
3) The measurement volume was progressively traversed horizontally closer and closer to the pipe walls until the data rate (frequency at which particles pass through the measurement volume) dropped excessively.
4) The automatic traverse system was then used to cause the measurement volume to travel in small increments back across the pipe toward the pipe centerline.
5) At each point during the traverse, approximately 32,000 data points or 200 seconds of data (whichever was reached first) were taken and recorded to disk for processing.
6) Steps 4 and 5 were repeated until an entire horizontal profile was acquired.
7) Steps 2 through 6 were repeated until an adequate number of horizontal profiles were taken.
8) Steps 1 through 7 were repeated until all streamwise stations were sampled.
9) The probe head was rotated to measure the vertical component of velocity and steps 1 through 8 were repeated for a subset (six) of the streamwise stations.
Once the measurements were completed, the unprocessed LDV data was converted to velocities and statistics were determined from those values. The time history data was not recorded due to computer space limitations, consequently only the velocity statistics are available for review.