29
pattern will be a sine wave with a time period corresponding to the rotational
speed. If problems are encountered, such as extremely high runout magnitude or
a hot bearing, further analysis should be performed. If two proximity probes are
mounted 90 degrees apart at each elevation, their output can be hooked up to a
two-channel oscilloscope to plot the actual runout orbit pattern. This can be
helpful in detecting a rub in the bearing. A spectrum analyzer is also useful when
troubleshooting a vibration problem. The spectrum analyzer breaks down the
signal from the proximity probes into its component frequencies. The runout
readings taken should also be compared to the readings taken before the unit was
shut down for the overhaul.
speed. If problems are encountered, such as extremely high runout magnitude or
a hot bearing, further analysis should be performed. If two proximity probes are
mounted 90 degrees apart at each elevation, their output can be hooked up to a
two-channel oscilloscope to plot the actual runout orbit pattern. This can be
helpful in detecting a rub in the bearing. A spectrum analyzer is also useful when
troubleshooting a vibration problem. The spectrum analyzer breaks down the
signal from the proximity probes into its component frequencies. The runout
readings taken should also be compared to the readings taken before the unit was
shut down for the overhaul.
Load rejection tests.
If a hydraulic transient study was performed to determine
the correct wicket gate timing, either because the unit was uprated or the original
data was lost, load rejection tests should be performed to verify the findings of
the study. The study should provide predicted spiral case pressure versus time
graphs and unit overspeed versus time graphs for the forebay and tailrace
elevations at the time of the tests. The wicket gate timing should have been set
to the value recommended by the hydraulic transient study before the unit was
watered up.
data was lost, load rejection tests should be performed to verify the findings of
the study. The study should provide predicted spiral case pressure versus time
graphs and unit overspeed versus time graphs for the forebay and tailrace
elevations at the time of the tests. The wicket gate timing should have been set
to the value recommended by the hydraulic transient study before the unit was
watered up.
Load rejection tests should be performed at lower loads before performing a full
load rejection. Immediately following a test at a lower load, the test data should
be studied to make sure that a rejection at the next higher load can be performed
without exceeding design limits. If, from the test data, it appears that the
pressure could possibly exceed the design pressure at full load, the wicket gate
timing should be slowed down before the full load rejection test is performed. If
pressure is significantly lower than the hydraulic transient study and the
overspeed higher, the wicket gate timing can be made faster.
load rejection. Immediately following a test at a lower load, the test data should
be studied to make sure that a rejection at the next higher load can be performed
without exceeding design limits. If, from the test data, it appears that the
pressure could possibly exceed the design pressure at full load, the wicket gate
timing should be slowed down before the full load rejection test is performed. If
pressure is significantly lower than the hydraulic transient study and the
overspeed higher, the wicket gate timing can be made faster.
If the unit being tested is equipped with a pressure regulator or an automatic
bypass valve, several load rejections should be done from not more than 15-
percent gate to verify the wicket gate timing. While the wicket gate timing still
should be set with the unit unwatered, some regulator designs are such that the
timing can only be accurately tested during an actual load rejection.
bypass valve, several load rejections should be done from not more than 15-
percent gate to verify the wicket gate timing. While the wicket gate timing still
should be set with the unit unwatered, some regulator designs are such that the
timing can only be accurately tested during an actual load rejection.
Immediately prior to the test, the following data should be recorded:
· Gate position
· Load
· Forebay elevation
· Tailrace elevation
· Spiral case pressure
· Load
· Forebay elevation
· Tailrace elevation
· Spiral case pressure