32 R.A. Kettle et al. Fig. 3.4 Information flow of the real-time data acquisition system which is data storage which will have relatively low jitter and so is better suited to interact with the FPGA. Since real-time OSs don’t allow for typical user inputs, such as using a mouse or keyboard, user input can instead be entered into the laptop. This information can then be sent to the real-time OS on the controller via the ethernet and from there through the PXIe chassis to the FPGA. This information flow can be seen in Fig. 3.4. Although balancing the relationship between the three systems can be a bit of a challenge at times, this setup offers a tremendous amount of flexibility and options. 3.6 Non-Real-Time Data Acquisition Initial results have been collected from the experimental setup using a simplified non-real-time data acquisition system and using a single frequency in the kilohertz range to excite the PZT. This set of data, although limited in scope, is meant to serve as a proof of concept for detecting collisions at high speed using the EMI method. An NI PXI-6133 S Series multifunction DAQ card was used to generate the excitation signal and acquire the response using a sampling frequency of 2.5 MS/s. Figure 3.5 shows a zoomed in section of data from the extensometer and the voltage response signal from the PZT, spanning from immediately before to immediately after the impact. The measurement circuit used was taken from the literature [8] and is shown in Fig. 3.6 where x(t) is the excitation signal, Rs is a precision resistor used to limit the current, y(t) is the response signal, ZPZT is the impedance of the PZT, and Zin is the input impedance of the data acquisition device. Because everything else in the measurement circuit is constant, the change in the voltage response of the measurement circuit is due to the change in impedance of the PZT caused by the impact. This proves that state detection through EMI in highly dynamic environments is possible. 3.7 Conclusions This paper presented a plan to adapt currently used methods and technologies employed by the SHM community for use in real-time state detection on the microsecond time scale for use in highly dynamic systems. A timing feasibility study was then conducted which outlined the steps required to make a measurement using the electromechanical impedance method. It was shown that real-time state detection could be done on the microsecond time scale. Then a modular experimental setup was presented that is intended to replicate a highly dynamic system through the creation of a collision and was designed for the express purpose of being highly customizable and having potential use even outside the goals of this research. Finally, a set of non-real-time data was presented which proves that state detection through EMI in highly dynamic environments is possible. 3.8 Future Work In order to obtain experimental results, a program for real-time data acquisition needs to be created. Of primary concern is the volume of data that has to be handled when monitoring on the microsecond scale for even just a few seconds. Once the data acquisition program is completed and the results can start being collected and evaluated, further strides in the research can be made. This includes the possibility to develop better state detection algorithms, investigating the use of different FFT architectures that may be more efficient, and discovering empirically how many averages are necessary for a reliable impedance measurement.
RkJQdWJsaXNoZXIy MTMzNzEzMQ==