Failure Analysis and Repair of a Radar Antenna

Figure 1.  Total Radiation Aperture Control Antenna used on E-2C Hawkeye surveillance aircraft.

For many Americans, Desert Storm provided the first close look at the accuracy and firepower of our modern military. While less publicized, a vast intelligence network was also in place during the conflict to identify targets and orchestrate aircraft and troop movements. "C4I" is the abbreviated name given to this network for "Command, Control, Communication, Computers, and Intelligence". C4I is at the heart of modern warfare, and is considerably dependent on airborne control and surveillance systems.  Over land, the AWACS (Airborne Warning And Control System) provides military commanders with sufficient information and control to support a tactical theater of about 500 aircraft. Complementing the AWACS is the E-2C Hawkeye Radar System designed exclusively for deployment from aircraft carriers. This aircraft/rotodome system flew thousands of missions in Desert Storm and has proven superior to the AWACS for detecting low-radar-cross-section targets over water.  A principal component of the E-2C System is its TRAC-A (Total Radiation Aperture Control Antenna), first installed in 1983. The size and weight of this antenna make it ideal for use on folded wing aircraft as it measures only 24 ft in diameter and 2.5 ft in height (see Figure 1). It is comprised of an array of coaxial elements arranged in a series of 10 compartments, or bays. Each element is kept pressurized between 38-45 psia. Pressurization is an essential requirement during high power operation, loss of which could lead to rapid electrical breakdown within the antenna. In 1992 the E-2C Group II was introduced to the fleet. Group II aircraft are equipped with higher horsepower engines and a lightweight version of the TRAC-A. Of these, ten occurrences of unexplained cracking have been detected, some after only 292 hours of operation. In all cases, the cracks occurred at similar locations in the antenna element near the rotodome's periphery, in an area known as Bay 4. If allowed to continue, the Bay 4 cracking problem could ultimately lead to loss of pressurization and possibly catastrophic antenna failure.

Laboratory tests determined the crack growth was related to high cycle fatigue. Cause of the crack initiation remains under investigation but extensive modeling and modal testing identified the detrimental vibration to be primarily torsional in nature. A detailed model of the Bay 4 element was constructed, verified, and refined using test data. The refined model was then used to develop a fix aimed at suppressing the detrimental response.

The challenge was to develop a corrective action providing sufficient torsional restraint, without increasing the low frequency fundamental lateral and vertical bending modes excited by vibration generated by the aircraft's four-blade propellers. In all, twelve such repair concepts were evaluated. 

A design selection based on pairs of struts installed along the antenna beam appeared to be the most feasible remedy. An evaluation plan was formulated which called for the proposed repair to be implemented on one side of a rotodome that would then be used for in-flight testing. Data would be collected from both sides of the assembly and a comparison of forces and cracking in the repaired side versus the unrepaired side would demonstrate either the success or failure of the proposed correction. 

The test plan called for both sides of the rotodome to be instrumented with accelerometers and strain gauges. Since the antenna rotates, a specially modified slip-ring assembly was fitted to transport signals from the antenna to the recording instrumentation. A flight protocol was worked out to assure that a sufficient number of aircraft maneuvers and antenna angles would be recorded for later analysis, and the in-flight and recording computers were to be synchronized for precise comparison of data to aircraft movements. Also, all cabin and crew activities would be recorded on video and audio equipment. 

In all, 32 channels of data would be filtered, amplified, and recorded using a signal conditioner and data recording system developed by R.C. Electronics Inc. The recorder would capture data, include time/event markers, and accommodate notations during data acquisition. Also, the same instrumentation used in laboratory preflight qualification testing was to be used in the actual flight test. This necessitated a more rigorous level of vibration testing than might otherwise have been required to verify that all subcomponents were well short of their fatigue limits. Meeting these rigorous demands proved to be somewhat more challenging than initially anticipated. 

Figure 2. These time histories show 50 sec of torsional responses in the repaired versus unrepaired sides of the TRAC-A during a maximum rate flat turn at 30,000 ft. Periodic twisting events are distinctly seen here in the unrepaired element (red trace), while the repaired element (blue trace) is comparatively low.   Figure 3. A portion of data from the unrepaired element shown in Figure 2 is displayed here expanded in time (red trace). Its spectrum (green trace) indicates responses at 74, 148 and 222 Hz; the fundamental and harmonic frequencies generated by the aircraft's twin four-blade propellers.   Figure 4.  Acceleration data taken from the repaired side of the TRAC-A shows significantly reduced responses at those frequencies where high energy twisting motion had been seen in Figure 3.

While most of the instrumentation held together well, some further measures had to be taken to improve the system's vibration resistance. Ultimately these all proved acceptable except for the hard disk drive. Though it did not exhibit catastrophic failure, it would not read or write during vibration testing. Efforts to rigidly mount the drive proved unsuccessful as did all attempts to mechanically isolate it.  With some reluctance, it was agreed that a hard drive could not physically be installed in the enclosure and be expected to function reliably. After much hand wringing, a local sporting goods store supplied a partial solution in the form of a small fabric pouch designed for hikers and backpackers. Once enclosed in the pouch, the hard drive would be strapped to the system operator and thus take advantage of the natural damping provided by a human body. The drive was then tethered via a flat cable to the system CPU through a front panel quick disconnect enabling the hard drive, collected data, and operator to separate from the instrumentation during an emergency evacuation. While the design functioned as intended it failed a flight safety inspection in consideration of a scenario in which the quick disconnect might jam and delay the operator's escape. Countering this objection, a breakaway Velcro mount was added allowing an immediate separation if necessary, thus leaving the drive and data behind. With this solution in place, flight testing proceeded, proved successful, and satisfied all test plan objectives.   Acceleration data of excellent quality was collected from eight test events along with time, heading, and antenna azimuth. These data verified that high energy torsional modes did in fact exist. The repaired side showed an 80% reduction in acceleration responses with only minor increases in vertical responses during landing. The maximum absolute acceleration recorded was 28 g during a maximum rate flat left turn at 30,000 ft, with some low frequency excitation generated from the airframe itself. Variation relative to antenna azimuth was shown to be very pronounced, with no correlation found between the rotodome speed and acceleration. Flight data showed the proposed fix to be highly effective high energy regimes (see particularly in high altitude, Figures 2-4). This project presented many unique challenges which were overcome by hard work, creative thinking, and the joint cooperation of several talented individuals. The data recording system proved reliable throughout the test. Although wearing its only moving part like an article of clothing may seem a bit unorthodox, as the saying goes, "whatever works, works." In this case, a somewhat unorthodox approach worked very well and got the job done.

Acknowledgments. It is not possible to name each of the many individuals who contributed to this study through their combined and cooperative efforts. The author does wish to recognize the organizations who provided those people, and the resources, that have made this program a success. These include: Naval Aviation Depot, North Island, San Diego, CA; Naval Air Warfare Center, Patuxent River, MD; R.C. Electronics Inc.; Veda Inc.; Northrop Grumman; and Randtron Antenna Systems, L-3 Communications. 

Jones, T., "Rotodome Vibration Survey Findings and Recommendations Report," Randtron Antenna Systems, November 1996.

Back to Application Notes Index

Home

6464 Hollister Avenue
Santa Barbara, CA 93117
Phone: (805) 685-7770
FAX: (805) 685-5853
info@rcelectronics.com

Products    Contact Us    Application Notes     Join E-mail List     Representatives     Event Calendar    Home