3. FAN TEST - RESULTS AND DISCUSSIONS
3.1 Discussion
on the equipment, the fans, and the rotor
Fans direction effects on performance
To study the effect of fan direction, a single fan was adjusted to blow
at only one of the halves of the rotor. Blowing the drag assisting half
caused the rotor to accelerate fast to the maximum speed while blowing
the counter wind half caused the rotor to rotate very slowly on the wrong
reverse direction. By assisting the rotor to a certain rpm first, the
rotor under the unfavourable wind condition still can accelerate to the
maximum speed. This respond shows that the uniformity of the wind from
the fans can affects the self-starting capability and performance of the
rotor significantly.
Quality of the wind from multiple fans
The wind from fans comes in intermittent gust form with slight direction
changes from time to time. The wind speed displayed by handheld anemometer
fluctuates in ranges of roughly ±0.25m/s (±6% for average
speed of 4.5 to 5.5m/s) if sudden increase or decrease not included. Sometimes
the fluctuation range was much higher that a much subjective reading had
to be decided. However, the resulted averaged wind speed values are logical
as it increased with the fan speed as expected. Table 1 shows one of the
wind speed readings taken in front of the rotor.
| |
at 1.5R to left |
left column |
centre column |
right column |
at 1.5R to right |
| upper row |
6.7 |
5.3-5.8 |
4.9-5.2 |
4.5-5.1 |
5.9 |
| centre row |
5.7 |
5.9-6.5 |
4.9-5.3 |
5.5-6.0 |
4.3 |
| lower row |
6 |
5.6-5.9 |
5.7-5.9 |
5.3-5.5 |
6 |
Table 1: Wind speeds at 9 points
in front of the stationary rotor for ‘high’ fans speed
(NOTE: These anemometer readings were not yet calibrated to wind tunnel
pitot tube in Section 3.1 of Report II)
Observation on the band brake
Sometimes during the test, sudden increases in the spring scale reading
occurred due to sudden friction increase between the cotton rope and Al.
pulley surface. Cleaning up the pulley and cotton rope surface seems to
solve this problem.
The friction in spring scale when it slides can be reduced by aligning
the scale with the cotton rope direction. Small vibration in the spring
scale was sometime observed which caused the reading to fluctuates around
±5grams (±5% for spring scale and weights differences from
100g to 180g). Interestingly, the spring scale vibration peaked suddenly
every time the rotor reached the stalling condition.
Rotor structure vibration and airfoil bending
The vibration of the whole rotor is most severe around 300 to 400rpm,
which is the range for most of the peak power values. Vibration during
startup and maximum speed is much smaller. At maximum speed, the pipe
and shaft swirled slightly around the rotation axis. This is probably
caused by clearance at bearing shaft joint and pipe frame joint instead
of the deflection of structure. Some of the nut and bolt joints without
spring washer were loosen during the testing time might be caused by vibration.
The deflection of the solid wood airfoils was clearly seen at maximum
rpm of around 550rpm. Fortunately, none of these broke during the testing
time.
3.2 Result for additional testings of
tachometer and stroboscope
Tachometer sensitivity
The tachometer refreshes it’s reading display every 1.1 seconds
or more depends on the object’s rpm. Table 2 shows that the tachometer
refreshes faster as the rpm increases. Selecting a small time interval
(near to 1 second) for rpm reading will cause high percentage of error
while selecting a long time interval will flatten out the peak value information.
At 200rpm, if rpm reading is taken for every 10s interval, the resulting
error percentage for rotor torque or power will be +13% or -18%.
| Rpm |
t (s) |
| 120 |
19 |
| 200 |
15 |
| 500 |
11 |
Table 2: Tachometer refresh time for total of
10 readings
Differences between stroboscope and tachometer
reading
The tachometer was not available during the band brake testing, so the
stroboscope was used. Differences were found between readings from both
tools. The lab technician hinted that the tachometer was more accurate
so the stroboscope reading will be ‘calibrated’ to the tachometer
scale. Table 3 shows that readings obtained from the stroboscope were
about 0.77 times of the tachometer readings.
| Stroboscope 3x rpm |
Tachometer rpm |
Tacho/strobo |
| 2030 |
522 |
0.77 |
| 1960 |
511 |
0.78 |
| 1790 |
458 |
0.77 |
Table 3: Differences between stroboscope
and tachometer readings
|
