Actuator Short Step Life Tests
MEMORANDUM from R. Lacasse December 29, 1994
Introduction
At the October 1994 GBT Advisory Committee Meeting, it was
pointed out that actuator life tests to date had consisted of running
the actuator from one end of its travel range to the other in one
continuous motion. This method of testing yields one start/stop cycle
per 1.7 inches of travel. In actual operation, many more start/stop
cycles per inch are expected. Motor wear, which is expected to limit
actuator life in most cases, is a function of both the number of rotor
revolutions and start/stop cycles. It was recommended that NRAO set up
an experiment to measure the effect of start/stop cycles on motor
life. This memo reports on that experiment.
Experimental Set-up
The 4-panel, 9-actuator mock up near the Interferometer Control
Building was used as the test bed for this experiment. A sketch is
shown in Figure
1. The actuators in this test-bed are driven electrically in
exactly the same way that they will be on the telescope, over a
similar length of cable (about 300 ft.). The actuators are subject to
the Green Bank outdoor environment. Panels and wind provide loading.
In this test-bed it is not possible to move the actuators over large
angles, as they would be on the telescope; however this was not seen
as a serious limitation.
The top three actuators (labeled 1, 2, and 3 on Figure 1) were
used as controls. They traveled 1.7" (near full stroke) in one
continuous motion while the bottom three actuators (7, 8, and 9)
traveled the same distance, but in small steps. The top three
actuators are referred to below as the stroked actuators, the bottom
three actuators as the stepped actuators and the 1.7" of motion is
referred to as a stroke. The stroked actuators had 1 start/stop cycle
per stroke. The stepped actuators were turned on for 0.8 sec and off
for 0.8 sec resulting in steps of approximately 12 mils or 300
microns. It was verified that the stepped actuator motors experienced
a negligible temperature rise (< I degree C). The calibration of
the LVDT position transducers was checked with dial gauges; they were
found to be within +/-1% of one another. Thus, during the course of
the experiment, all actuators were moved 7488 strokes or approximately
12,730", and accumulated approximately 238 hours of running time. The
stroked actuators experienced 7488 start stop cycles while the stepped
actuators experienced about 1,000,000 each (actuator 7: 1,057,836;
actuator 8: 1,073,07; actuator 9: 1,099,888). The selection of
1,000,000 steps as a goal for the experiment was based on a variety of
estimates of operational requirements. A memo by Lockman will give
further details on this. In any case the ratio of start/stop cycles
for the stepped versus the stroked actuators is large enough (133)
that the dependence of brush wear on start/stop cycles should be
elicited.
Brush Wear Results
Data on brush wear is shown in Table 1. The data is also depicted
graphically in Figures 2 through 7. At best, there appears to be only
a slight dependence of brush wear on the number of start/stop
cycles. Average brush wear for 7488 start/stop cycles was 6.42 mils;
average brush wear for 1,000,000 start/stop cycles was 7.58 mils.
Brush wear of over 100 mils is required to make the motor fail.
Commutator Wear Results
Data on commutator wear is shown in Table 2. The average stroked
actuator wore 0.67 mils. The average stepped actuator wore 3.3 mils.
There seems to be a significant result here. However, it should not
seriously impact motor life. Commutator bars are 40 mils thick and
thus it would take 80 mils of wear (measured as a diameter) to go
entirely through them. We have had motors work well with 15 to 25
mils of wear.
Other Results
One result of the experiment that was not anticipated was the
accumulation of grease in the motor. Two of the three stepped
actuators had to be cleaned during the course of the experiment as
grease was interrupting contact of the brushes with the commutator.
One of the three stroked motors was found to have some grease on its
windings after the experiment. This potentially serious problem is
being discussed with the actuator manufacturer and will be the subject
of a future memo.
Acknowledgements
Many thanks to Dwayne Schiebel who helped with the set-up,
measurements, and sanity checks throughout the experiment.
Figure 1 Sketch of 9-actuator test set
Table 1 Brush wear data
Figure 2, 3 Brush Length Distribution
Figure 4, 5 Brush Length Distribution
Figure 6, 7 Brush Wear Distribution
Table 2 Commutator wear data