Effects of Fiber Volume on Modal Response of Through-Thickness Angle Interlock Textile Composites ()
1. Introduction
When considering how wavy layers interact with one another in a 3D textile composite, and how such layers might alter overall laminate mechanical properties, one can expect difficulties in studying the dynamic behavior of such textile composites. Waved layers will create interlaminar stresses and strains, which may vary the geometry of the unit cell of the laminate, increasing the damping and causing variations in the dynamic response. To obtain significant data, instead of fabricating and analyzing a costly series of samples, finite element models can be used to predict the behavior under different scenarios. The required number of degrees of freedom for models to accurately capture the possible geometry variations, fiber distortion and contact among adjacent fiber tows is unknown, and it will vary with assumptions for the tow geometry. As such, the analytical approach is also expensive in terms of required modeling time. Simpler analytical models will enable less expensive studies on dynamic behavior. In order to determine how detailed a model should be, experimental data and analytical data are compared.
Panels for experimental characterization have been supplied by Albany International Techniweave, Inc. The fiber architecture is a through thickness angle-interlock weave in which the fibers traverse completely through the thickness of the fiber perform. The through-the-thickness fiber angles were measured as approximately 40˚ in the central portion of the composite cross-section, and approximately 63˚ on the edges (Figure 1).
The three IM7/PR520 panels, identified as panel02, panel03 and panel04, had identical as-woven geometry but, due to different compaction and cure pressures, exhibit different fiber volumes and different thickness (Table 1).
Due to the most prominent assumption in micromechanics of composite materials, which formulates strain compatibility at the fiber-matrix interface, each stiffness parameter is directly related to the fiber volume [1] and
Figure 1. Section of panel02, through-thickness angle interlock.
Table 1. IM7/PR520 predicted panel material properties [2].
so it is possible to determine the equivalent orthotropic properties.
Because of the known idealized as-woven geometry of the samples, it has been possible to focus the attention on understanding if the increased fiber (tow) distortion at higher cure pressure affects the dynamic response of the panels.
The study is divided in two different steps. Initially, based on a modal analysis, the accuracy of the analytical model is determined. As a second step, the main problem is addressed and the influence of the curing pressure (or fiber volume) is studied.
2. Experimental Analysis
2.1. Experiment Setup
The structural modes of the composite panels were found experimentally using a Scanning Laser Vibrometer (Figure 2), which gathers vibration information by detecting the laser Doppler shift and using it to measure the velocity of the object’s surface. The system consists of a scanning head and a control/processing computer (which controls the scanning head and evaluates the data retrieved). The measurement data is digitally recorded in the workstation where software controls the data acquisition and offers user-friendly functions to evaluate the measurement data.
The panel was suspended on two nylon strings to simulate unconstrained conditions (Figure 3). Under the panel was a noise source consisting of an insulated speaker box with a noise outlet tube. This allowed the generated sound pressure to be concentrated on a small area of the panel.
2.2. Procedure
The procedure for the initial step of the experiment consisted of identifying frequencies that produce a large response and then scanning the panel at those frequencies to individuate the modal shapes.
To identify the frequencies, a single point on the panel was scanned while a periodic chirp was generated. The periodic chirp was setup to sweep the frequencies from zero to four thousand Hertz. This range has been chosen to allow the determination of at least six modal frequencies, considered enough to qualify the dynamic response of the model. From this scan, the software was able to produce a digital Fourier transform (DFT). From this DFT the frequencies producing the largest responses were noted (Figure 4).
One scan of the entire panel was then made for each of the resonant frequencies noted. From this scan a collection of the first six vibration modes of the panel were identified (Figure 5). The above procedure was conducted for all three composite panels.
Figure 4. Digital Fourier transform for panel02.