Development of the Fiber Optic Wide Aperture Array: From Initial Development to Production

Introduction (1986-1990): In the late 1970s, NRL began the development of fiber optic sensors for Navy surveillance applications.1 The initial thrust was fiber optic acoustic sensors for towed arrays. The basic approach used a passive fiber interferometer to measure acoustically induced fiber strains (using techniques of fiber strain amplification) while the interrogating laser and processing electronics were located remotely from the fiber sensor array. While the technology for remote interrogation and multiplexing of fiber sensors was proceeding for towed arrays, the use of fiber optic sensors for hull-mounted arrays was being considered. By the late 1980s, NRL became actively involved in developing both fiber optic hydrophone designs, multiplexing approaches, and laser sources for the hull-mounted hydrophone systems. Initial work concentrated on both low-frequency (12 × 12 in.) and midfrequency hydrophones (less than 5 × 5 in.). By the late 1980s, however, it became clear that the most likely transition path was the development of a midfrequency range, area-averaging hydrophone to form a lightweight version of the Navy's Wide Aperture Array (WAA) hydrophone system for fast attack submarines (the Lightweight Wide Aperture Array, LWAA). The system provided major weight savings and allowed for easier and less costly technology-refresh by moving all the electronics and active optics inboard. Figure 1 shows the system concept.

FIGURE 1
System concept.

Technical challenges included designing an area-averaging fiber hydrophone with a flat response, more than an order of magnitude higher in frequency than existing designs. This was achieved by incorporating two, fiber-wrapped, air-backed metal mandrels whose acoustic responsivity was optimized for the operational hydrostatic environment. A second challenge was to provide a low noise interrogation/multiplexing approach to accommodate reliability requirements, minimize the number of fibers needed to interrogate the array, and allow sufficient bandwidth for high frequency, wide dynamic range signals. This was accomplished by using a lithium niobate phase modulator to produce the phase carrier and a modified FM-phase generated carrier interrogation/multiplexing approach. Finally, to meet the stringent noise requirements of the system, NRL identified a neodynium YAG solid state laser operating in the 1.3-µm band. Unlike the semiconductor laser being used at the time for fiber sensor interrogation, which missed the noise floor by ∼25 dB, the fast roll-off with frequency of the FM noise of this Nd:YAG source allowed the noise floor to be met with usable margin.

The Prototype System (1990-1993): After NRL had demonstrated each of these basic building blocks of the system (and performed preliminary flow noise testing of the two-mandrel fiber hydrophone2) a contract was awarded to Litton Guidance and Control Systems (now Northrop Grumman) to build a 49-channel prototype array for flow noise testing at Lake Pend Oreille, Idaho, and acoustic/vibration testing at Seneca Lake, New York. This effort was funded by an Advanced Technology Demonstration (ATD) program sponsored by the Naval Sea Systems Command. During this period, NRL (and the Naval Undersea Warfare Center (NUWC), which had responsibility for the mechanics of the array) worked closely with the contractor to ensure that this subarray would meet all the Navy's stringent specifications. This required moving to a five-mandrel design. This often required NRL to perform a number of important tests (in 1991 and 1992) to demonstrate specific technology approaches to the team that would have to be followed if the ATD were to be successful. One particular area where NRL played a major role was in defining critical performance parameters of the demodulator. Figure 2 shows an example of NRL improvements to the performance of the demodulator during flow noise testing. These tests also allowed NRL to demonstrate improvements to the baseline system, some of which would be used in the preproduction and production systems. In 1993, the prototype ATD array performed all the required tests successfully, which paved the way to the next stage of development.

FIGURE 2
Improved noise performance with NRL processing under "pop-up" conditions. Upper figure: conventional fiber optic processing, lower figure: NRL processing, contamination reduced by up to 20 dB.

Advanced Development (1994-1999): After the successful completion of ATD, Litton was tasked with designing the system with production in mind. One of the major changes was going to a two-mandrel, air-backed polycarbonate mandrel hydrophone. Fortunately, NRL had already demonstrated this basic design three years previously and we were able to verify the flow noise performance without incurring another costly flow noise test. NRL also played a major role in keeping the program on track after the contractor lost key personnel during the telecommunications boom. On numerous occasions, only NRL had the detailed system design knowledge to overcome numerous technical challenges that had to be overcome to allow the system to meet performance specifications in all acoustic/pressure environments.

Production System (2000 and On): Much of the production phase involved packaging the required components and required modifications of the hydrophone to ease manufacturability. During this phase, NRL's participation was much reduced; however, we still played a significant role in qualifying hydrophones built with optimized construction techniques. We also continued optimizing the various algorithms required for low noise demodulation of interferometric signals. In August 2003, the lead ship of the Navy's latest class of attack submarine, the USS Virginia, was christened (Fig. 3). The sonar suite of this vessel includes the Fiber Optic Wide Aperture Array, the first fiber optic surveillance grade acoustic sensor system on an operational platform. The current system still uses the basic hydrophone concept, interrogation/multiplexing, and laser proposed by NRL in the late 1980s.

FIGURE 3
USS Virginia, hull array patches circled.

Acknowledgments: The authors acknowledge the help and support of Gary Cogdell of NRL and Aileen Sansone, previously of NRL, and Roger Maple, previously at NUWC. The authors also acknowledge the work of the Northrop Grumman team who made this system a reality.

[Sponsored by NAVSEA]