Tyrie Vella and Dr. Richard Selfridge, Electrical & Computer Engineering

Main Text

Fiber Bragg Grating (FBG) sensors are specialized optical fibers that can be used for detecting changes in strain. Under normal conditions, when a broadband spectrum of light is transmitted through the fiber, a narrow band of light is reflected by the sensor. When an FBG is embedded into a composite material and the material is placed under stress, the wavelength at which the reflected band is centered shifts in proportion to the stress, allowing the strain to be measured. By monitoring the sensor during repeated impacts on the material, structural damage can be discerned and the material could be replaced or repaired before catastrophic failure.

Existing methods of interrogating FBG sensors have two main limitations. In most systems, only the point of highest intensity is recorded. In some cases this is sufficient, but in others complex strain on the material causes multiple wavelength bands to be reflected by the sensor, and in these systems that information is lost. Other systems may capture the entire spectrum, but they do so at an infrequent rate, which is insufficient to characterize the strain on the material in the milliseconds following an impact. One such system previously developed by my mentor’s research team1 was capable of full-spectrum scanning at 525 Hz. The goal of this research project was to develop a system capable of scanning a broad spectrum at 100 kHz, nearly 200 times faster.

My primary role in this research was to develop an algorithm for calibrating the interrogation system and to create software to automate data capture and processing, while another student, Spencer Chadderdon, focused on the hardware enhancements needed for such a speed increase. A block diagram of the system is shown in Figure 1. The key component is a MEMS filter and photodiode, which converts the optical signal reflected by the sensor into an electrical signal that can be recorded using an Analog to Digital Converter (ADC). The MEMS filter is driven by a function generator, and the driving signal is also recorded.

At low frequencies calibration of the system is easily accomplished by creating a mapping of function generator voltage to reflected wavelength. The wavelength transmitted by the MEMS filter changes in response to the driving voltage. The mapping is created by using a tunable laser that produces a fixed wavelength, and noting the driving voltage that corresponds to the appearance of that wavelength in the recorded signal. Then, when data capture is performed, the mapping can be inverted to convert the recorded signal back into wavelength.

However, at high frequencies there is a significant amount of time that passes between a change in the driving voltage and the corresponding change in the wavelength. If the simple calibration method is used, then it will have a large error in one direction when the voltage increases and a similar error in the other direction when the voltage decreases. I developed a method of quantifying this time delay as part of the calibration process, so that the captured data can be interpreted accurately even when the scanning frequency is well over 100 kHz.

As our system developed, I travelled on two occasions to North Carolina State University to test it. Our collaborators, Dr. Kara Peters and her students, do research in composite materials. We connected our interrogation system to FBG sensors embedded in composite laminates they had prepared, and recorded the sensor readings while repeatedly impacting the material in a drop tower.

On the first occasion, our interrogation frequency was limited to 10 kHz by the bandwidth of the trans-impedance amplifier (TIA) that we used, which Spencer had made from scratch as a proof-of-concept. For the second set of experiments we substituted a high-quality commercial TIA, and successfully interrogated the sensor at rates up to 300 kHz.

Our results clearly show the advantages of full-spectrum interrogation at high speeds. In Figure 2, the central black line shows what simple peak-tracking methods would have recorded, while the full-spectrum image shows that a much larger strain was actually present. Figure 3 illustrates the benefits of high-speed interrogation by comparing our actual data to how it would appear using lower scanning frequencies.
We submitted our findings to the Journal of Measurement Science and Technology2 in January of this year; the paper was accepted with revisions, and we expect it to be published shortly. Also, Spencer presented our interrogation system in March at the SPIE conference for Smart Structures and Materials. A summary of our findings will also be published in the conference proceedings.

References

• Schultz S, Kunzler W, Zhu Z, Wirthlin M, Selfridge R, Propst A, Zikry M, and Peters K 2009 Full-spectrum interrogation of fiber Bragg grating sensors for dynamic measurements in composite laminates Smart Materials and Structures 18 115015
• T Vella, S Chadderdon, R Selfridge, S Schultz, S Webb, C Park, K Peters and M Zikry Full-spectrum interrogation of fiber Bragg gratings at 100 kHz for detection of impact loading Measurement Science and Technology
• In addition to my mentors, Dr. Richard Selfridge and Dr. Stephen Schultz, I would like to acknowledge my co-researcher Spencer Chadderdon; our collaborators at NCSU, Dr. Kara Peters, Sean Webb, and Chun Park; and the sponsorship of the National Science Foundation.