The onset of local damage in structures, such as delamination, cracking, and fastener loosening, can often be difficult to detect and has long-term implications on the performance of the composite structure. These structures are often exposed to a variety of conditions, including impact, shock loading, and extreme changes in temperature. Because of both manufacturing requirements and design specifications, large and complex sections often need to be joined together to form the final structures. It is important to understand the failure behavior of these joints under a variety of static and dynamic loading conditions.

A typical cured composite panel with Embedded Strain Sensors prior to cutting into individual specimens.
Structural Health Monitoring (SHM) seeks to provide ongoing monitoring of a structure’s integrity, minimizing the need for programmed inspections, and allowing maintenance to be need-driven, rather than usage-driven. Current SHM approaches often use strain gages, accelerometers, and, more recently, piezoelectric sensors.

Methods have been devised to use the sensors in a network to “triangulate” readings/ locations of interest. This is especially true for piezoelectric sensors, which provide an actuation, as well as a sensing, function. One approach to wide-area damage detection is to harness the ability of certain classes of materials to provide a self-diagnosing function.

The unique structure of carbon nanotubes (CNTs) brings the material some outstanding properties that make it possible to apply CNTs to applications in various areas of materials, including the development of a new generation of sensors. CNTs are a promising material for detecting chemicals and biochemicals due to several intriguing properties, including their ability to mediate fast electron-transfer kinetics for a wide range of electroactive species, and large length-to-diameter aspect ratios that provide high surface area.

In this study, the unique capability of CNT-based sensors for sensing local composite damage was used. Through careful design of the specimen, it will be possible to detect the onset and progression of damage in the fiber. This work demonstrates a novel, multi-modal, nanomaterial- based sensor technology that can provide wide-area detection of damage.

CNT growth on glass fibers (fuzzy glass fiber) was carried out using a traditional chemical vapor deposition process. Two sets of fuzzy glass fiber were prepared by changing only the growth time. CNT fuzzy glass with short resident time exhibits less-dense CNT growth than long resident time.

Fuzzy fiber strain sensors were characterized as fibers, tows, monocomposites (embedded in epoxy matrix material), and composite specimens. Coupon geometries were designed to evaluate specific characteristics of the sensors of interest. Following successful evaluation of individual fibers, single tows were embedded in an epoxy matrix monocomposite and subsequently incorporated into carbon composite specimens designed to produce specific geometryinduced responses. These included straight-sided, stress-concentration, and Poisson’s-effect specimens. All composite specimens were fabricated as both unidirectional and orthotropic fiber orientations.

Experiments indicated that fuzzy fiber with high-density CNTs produced the best strain response, so sensors were developed accordingly for use in the composite specimens. The process of embedding sensors in a carbon composite panel required electrical isolation for the sensor. Instrumentation leads were bonded to the fiber tow with conductive epoxy and extended beyond the isolation layer. All composite specimens were fabricated through a similar process. Six composite panels, 12 × 12" , were fabricated with IM7/977-2 prepreg unidirectional carbon fiber tape. Three panels each were prepared with unidirectional or orthotropic layups. Sensors were embedded at specified levels in the layup depending on the type of specimen response to be tested. The figure shows a typical cured panel with embedded sensor elements prior to cutting into individual specimens.

A series of tests progressed from an initial sensor feasibility study to incorporation in segments of actual composite panels. The sensors were shown to produce consistent response, low noise, high strain capability, and were repeatable for elastic strain in the same specimen (no hysteresis or offset).

At the beginning of tensile testing, substantial thermal response was ob - served, causing unacceptable drift in the strain signal. Faster loading rates or insulating the specimen from the air and allowing it to thermally stabilize before testing overcame the thermal issues. However, a separate investigation was performed to determine the magnitude of error in strain measurement that could be expected due to thermal errors.

This work highlights the feasibility of incorporating carbon nanomaterials into structural composites as sensors. The fuzzy fiber sensors exhibit similar sensitivity to conventional strain gages, and are more easily integrated into composite structures as the sensor itself is a composite. The fuzzy fiber strain gages can be used to sense strain within composite structures, and can be readily integrated into the structural laminate to provide sensing over large sections and in locations not accessible to conventional strain gaging techniques.

This work was done by K. Lafdi, J. Sebastian, N. Schehl, M. Bouchard, M. Boehle, and L. Linge of the University of Dayton for the Air Force Research Laboratory. AFRL-0209

This Brief includes a Technical Support Package (TSP).
Fuzzy Fiber Sensors for Structural Composite Health Monitoring

(reference AFRL-0209) is currently available for download from the TSP library.

Don't have an account? Sign up here.