#Fall 2012 End of Semester
Gustavo and Scott went to PCSI to do a data collection. We were aiming to do the first collection with all three blades at the same time. Unfortunately, there is a problem with one of the I2C channels are we were able to collect two out of three. Scott will investigate this problem.
One key issue is the mounting of the BeagleBone and sensors. There is a cap on the turbine blades in the middle of the hub. The BeagleBone can fit in the in this space. Stronger mounting that will resist vibration will be a next step.
We attached the sensors similarly to how the accelerometers were attached to the original turbine. The sensors are taped to the blade and the wires themselves are also taped to stop them from moving as the blade spins. A next step for this is mounting the sensors with a more permanent method to minimize any movement relative to the blade. We also need to mount them at a specific distance for measurements consistent with the original turbine setup.
We collected 10,000 samples per sensor with the wind tunnel running. We observed that the speed of the front turbine was fluctuating during the collection. The cause of this is unknown because the other turbine was rotating more consistently.
The data is below:
Here we can see the fluctuation in the speed of rotation. The sensor is configured for +/-2g range for maximum sensitivity, so the lines go off the graph at the beginning when it was spinning fastest.
In order to validate the data, the next step will be to put out accelerometer on the same blade as the original accelerometers and compare the two.
PCSI also has a calibrator for low frequency acceleration that we will try next semester as well.