28 R.A. Kettle et al. that the building’s structural integrity has become compromised it could alert the occupants of the building to evacuate, or conversely it could assure the occupants that the building is still stable and they should remain under cover inside the building. 3.2 EMI Method and Current State of Real-Time SHM The electromechanical coupling properties of piezoelectric materials (PZTs) are the key to the SHM impedance method. The direct and converse coupling allows the piezoelectric materials to act as both a sensor and an actuator. By attaching a PZT patch to a structure, the electrical impedance of the PZT becomes a function of the mechanical impedance of the host structure in the area where the PZT is located [1]. Assuming that the properties of the PZT remain constant, any changes in the electrical impedance of the PZT can be ascribed to changes in the mechanical impedance of the structure. These changes in mechanical impedance are typically classified as “damage” to the structure. To measure the electrical impedance of a PZT the SHM community has commonly used a HP4194A or other commercially available impedance analyzers [1]. Unfortunately these impedance analyzers have several drawbacks such as being large, heavy, expensive, and slow to make measurements and because of this there has been significant effort in the SHM community to develop better alternatives. There have been several suggested replacements including using a spectrum analyzer [2], a digital signal processor (DSP) [3], and a standard data acquisition device [4, 5]. It has been shown though that both the spectrum analyzer and digital signal processor hold a few disadvantages compared to the conventional data acquisition device. The spectrum analyzer method is essentially just an approximation based off of a resistor value which tends to make it inaccurate at high frequencies and the DSP method is restricted by the small memory space available on the chip, which restricts the range and frequency resolution [5]. The data acquisition device method is advantageous because it allows for a large range of hardware options, including the use of data acquisition devices with exceptionally high sampling rates. Several recent papers have investigated real-time SHM being performed on, or developed for, buildings [6–8]. Data sampling rates range from 1 to 200 Hz in these works with one stating that the real-time data was only available once every 10 min [8]. These data sampling rates and data availability times are inadequate for real-time microsecond state detection. There was also the development of a real-time system with multiple sensors and temperature compensation with a sampling rate of 1.25 MS/s with a stated a completion time of 3 s [9]. This is certainly an improvement, but it is still several magnitudes away from microsecond monitoring. There have, however, been previous efforts made to push real-time SHM into the microsecond timescale. One technique that was explored was the use of an output only modal shape method which was used to detect simulated damage on a finite element model of an aluminum plate [10]. This method did successfully detect damage in the finite element model, but lacked any sort of time study for either the simulation or a real world application. Another technique investigated was wave-propagation, which is a method similar to EMI but measures mechanical waves that have propagated through the structure. Pairs of PZT transducers were used to detect damage by monitoring for changes in impact-induced waves as they travel through the structure [11]. An “instantaneous baseline” is created by the first PZT affected by the excitation and damage is detected by comparing this baseline measurement to the measurement of the next PZT in the path of the wave. This technique showed some promising results and was able to detect damage in the form of small cuts on a plate. While this work does make mention of techniques that can be used to reduce data collection time, it also unfortunately lacked any sort of time study and so it is unclear precisely how long the method takes to complete. 3.3 Adaptation of SHM Impedance Method and Equipment A key difference in this work compared to conventional SHM is the use of higher frequencies in the excitation signal. Conventionally, most SHM uses excitation signals in the realm of tens or hundreds of kilohertz [1]. In an effort to reduce measurement time, the real-time approach investigated in this work will utilize excitation signals up to 50 megahertz. Higher excitation signals have also been shown to have a more local influence on a structure [12], which can be advantageous if there are only particular locations on a structure that are of interest. Another key differentiating factors between this research and what is typically done in the SHM community is this work is concerned with the evaluation of system state, such as changes in boundary conditions and interfaces, rather than looking exclusively for structural damage. It should be noted that the detection in changes of boundary conditions is not something
RkJQdWJsaXNoZXIy MTMzNzEzMQ==