|
| "You Can't Beat a Man Who Won't Quit" | |
|
Chris’s enquiry came in like many. He’d used Pico Technology’s data loggers in the past, so wanted to discuss their suitability for his latest project. However, as if his obvious passion for this project wasn’t enough, as soon as he mentioned Ivan’s quest to build a Richard Pearse plane, I was hooked. He went on to explain how Ivan was building a Pearse reproduction and what they needed to test before lift-off. It sounded like something straight out of an adventure novel - revisiting the pioneering days of powered flight, proving (or disproving) long held and debated theories, specialists, engineers and enthusiasts teaming up and contributing to this historically significant project. How could anyone not want to be part of that? On the grand scale of things, Metermaster’s involvement in the project would be small but it was still to provide essential pre-flight information. In short, with ‘the Pearse’ attached to a trailer being towed down the Whenuapai runway, Chris would need to measure lift generated by its wings and lateral forces created by tail plane adjustments. The lift information was to come from load cells mounted between the trailer and each of three wheels on the undercarriage and was to be collected every second (along with wind speed and data from the lateral load cell) stored on a computer in the towing vehicle and graphed - dynamically - and of course, to be faithful to the original design calculations, all figures were to be in imperial units. No problem. Well.. |
|
|
The load cells were always going to be the most challenging. Three 500kg load cells with 2mV/V outputs, one 1000kg load cell with a 3mV/V output and a 12V power supply meant
that we’d be measuring 24mV and 36mV full scale for the 500 and 1000kg ranges respectively. Fortunately,
Pico Technology’s ADC24 has 24 bit resolution that can be applied over selectable input voltage ranges.
It was configured in differential mode to remove the common reference and thus allow all four load cells to be monitored simultaneously.
Each signal would be scaled to eliminate the inevitable load cell zero offset and of course, with factors to display the values in pounds (lbs).
Not to make it all seem too easy though, the load cell outputs presented another problem! The scaling would rely on a stable power supply and although the hardware could be powered from the towing vehicle, this would introduce a risk of noise (interference). So Chris opted to strap a good quality aviation battery to the trailer. Good or not though, battery discharge would need to be taken into account. (A 0.5V drop in battery voltage, would result in an 88 pound change in scaled values.) That alone makes comparing results of different runs difficult enough but to add to this, there was still the question of what ‘nominal’ voltage to use in the scaling. Fortunately, PicoLog provides a solution. |
 |
| W = (24 x 734.874206)/PSU.. Yep, that’ll work! | |
 |
In addition to the measured values, ‘Calculated’ values were also created.
These combined the power supply voltage and each of the load cell values in ratiometric calculations,
so negating the effect of power supply changes. This also allowed a nominal 12V to be used in the calculations, regardless of the actual battery voltage.
However, as is often the case in these projects, when one problem is solved, (yet) another is revealed. (At this point, I was really starting to get a feel for what
Ivan must have been going through.)
All four load cells were wheatstone bridges, so logging their outputs and the excitation signal (the battery) together, meant we effectively connected ground to the centre of each bridge. To provide the isolation that would be required to eliminate this problem, we connected the battery voltage to a PicoLog 1012 logger and in the PicoLog software, ticked the ‘multiple loggers’ box. This function allowed us to view the collected data as if it was all from a single logger. Both the ADC24 and the 1012 were connected to and powered by the same host PC through a small USB hub and a 10m cable back to the towing vehicle. The wind speed signal was also connected to the 1012. That came from a 3 cup yachting anemometer with a display unit and an analogue output, 0~300mV = 0~60 knots. All in all it came together nicely but I can’t wait to see the (real) runway results. Hopefully they’ll give Ivan and his team the confidence to take the next exciting step. Skywards?? |
| Mark and Chris with one of the wheel load cells. |
| System overview with test graphs |
|
