Testing : Wind tunnel test intro | tunnel equipment | tunnel results, [pg2], [pg3] | tunnel conclude |
Multiple fans test intro | fan equipment | fan results, [pg2], [pg3], [pg4] | fan conclude

3. FAN TEST - RESULTS AND DISCUSSIONS
-> continued pg.2

3.3 Rotor rpm, torque, power, Ct, and Cp curves

Observation on the rotor unloaded acceleration
Rpm vs. time graph is plotted on Graph 1 for one of the collected result, which is shown in Table 4. The graph shows the way the rotor accelerated from stationary to maximum speed. First, the rpm rises fast but for a brief period only. Then, the rpm increases at slower pace for a long period. Next, the rpm starts to increase fast again for the second time and this time the rise is the fastest. Finally, it slows down quickly when reaching the maximum speed. At the maximum speed, the rpm continuously fluctuates around this speed. The acceleration pattern described above was observed in most of the testing conditions with exception when the rotor fails to self-start.

Graph 1: Rpm vs. time for one data set from testing under 3 fans with maximum fan speed

Table 4: Collected rpm data with time interval of 5s between each reading.
(Shaded areas indicate fast rpm increase)

The rotor angular acceleration, torque, and power at certain rpm under no load can be approximated to equations below. As the data was collected at 5s interval while the values were calculated at 10s interval (to reduce error), two curves can be plotted from a single set of rpm data.

Graph 2: Torque vs. rpm for 3 fans with maximum speed

Graph 3: Power vs. rpm for 3 fans with maximum speed

Graph 2 shows that the rotor produced highest torque during the acceleration period just before reaching maximum speed. The starting torque is the second highest torque and is almost half or a third of the peak torque. Graph 3 shows that the maximum power occurred at or slightly after the peak torque. Both graphs shows accuracy problem as the number data points on the curve around the peak is too low to show the exact peak.

The rotor performance is usually plotted in term of power coefficient vs. tip speed ratio or the torque coefficient vs. tsr. The tip speed ratio is defined as below. The normalized torque and power curves retain its’ original dimensional shape as shown in Graph 4 and 5.

Graph 4: Torque coefficient vs. tip speed ratio for 3 fans with maximum speed

Graph 5: Power coefficient vs. tip speed ratio for 3 fans with maximum speed

3.4 Errors in testings

The errors in the calculation of Cp and Ct originated from the equipment limitation, bad procedures, and approximation in the calculation process itself. Some of these errors can be measured or estimated so an estimation of error percentage in Cp and Ct can be made as below.

Calculation 1: Percentage of error in Ct and Cp

The error percentages above does not include:
- discretization error in angular acceleration calculation caused by low number of points defining the steepest slope in rpm vs. time curve. Smaller time interval cannot be used because of the slow refresh rate on the tachometer display.
- effect of wind direction, sudden increase or decrease of wind speed, and intermittent gust form of wind from fans
- positioning anemometer in wrong direction causing higher or lower wind speed
- rough calculation round up errors such as in calculating inertia, wind density, swept area, and wind speeds

Only 2 out of the 22 sets of collected data was filtered out because of extreme values or shape. Most of the accepted data produced performance curves with decent shapes and values.

Last updated at November 6, 2002
Comments are welcomed