1. Introduction
Composite materials play an important role in aerospace, rail, automotive, and energy applications because of their high specific strength, specific modulus, processability, and chemical stability [
1,
2,
3,
4,
5,
6]. Composite stiffeners, such as C, L, π and I beams, are one of the most common and essential structural components in modern aircrafts, rail cars, automobiles and wind blades. Stiffeners often bear complex loadings, both in-plane and out-plane, making them critical parts in primary structure assemblies, such as the spars and ribs in aircraft wings, tails, and fuselages. Currently most of such composite stiffeners are made of unidirectional or 2D woven fabrics, or the combination of both. The main drawbacks of these stiffeners are their delamination at low loadings and low transverse strengths, limiting their uses where shear and transverse loads, such as in bending, are substantial [
7]. This disadvantage is especially prominent for composite structures having curved shapes, such as ribs, angle bracket, stiffeners, and wind blades [
7]. When these structures are subjected to tensile, compressive bending in the plane of curvature, the interlaminar shear and radial stress develop in the through-thickness direction, namely the Z-direction, especially in the regions where the high bending moments are present, thereby resulting in premature delamination failure.
By comparison, three dimensional (3D) woven fabric composites with fibers in the Z-direction render high through-thickness strengths, damage resistance, delamination resistance, and impact resistance [
8,
9,
10,
11,
12,
13]. These improvements are attributed to the presence of continuous Z-direction fibers, and the 3D composite structure is considered to be the smoking gun for high transverse strengths. Unfortunately, these 3D woven fabrics, such as 3D orthogonal woven (3DOW) fabrics, are very difficult and expensive to produce, particularly when fabric thickness is higher than ten millimeters is needed, making them nearly impossible to be economically manufactured and extensively used in common structures such as stiffeners.
Unidirectional fibers, either in forms of fiber tape or non-crimp fabrics, together with biaxial woven fabrics, make most of today’s composite stiffeners, due to their availability and economical scales. Biaxial fabrics usually render higher resistance than unidirectional fibers because of their yarn-crimps or yarn-waviness towards the Z-direction, providing a certain degree of transverse load bearing. Based on this understanding, it was hypothesized that if adequate yarn waviness were created, namely by weaving a 2.5D fabric of certain structures, the resulting 2.5D composites might be able to provide interlaminar shear strengths and transverse strengths close or equivalent to that of the 3D composites. Since the 2.5D fabrics do not require Z-yarns, they can be produced on the most common weaving looms with minor modifications, thus avoid the technical difficulty and high costs in 3D fabric production, consequently providing a practical and economical alternative solution to the composite stiffener community.
In order to test this theory, a L-shape beam (L-beam) was selected as a model composite configuration. L-beams are essential components in any composite stiffeners, such as C, L, π, and I stiffeners. Their simple geometry allows the design of custom fixtures to generate tensile or compressive bending, so as to test their transverse properties and observe the corresponding failure modes. To our knowledge, there were no previous studies in the literature on the bending of 2.5D woven fabric composites.
Literature [
14,
15,
16,
17,
18] studied 3D composite L-beams and T-joints under tensile load, demonstrating their ability to carry a significantly higher load until failure than that of the 2D laminate counterparts. They further pointed out that the failure mode in 2D laminate L-beams was dominated by delamination due to out-plane tensile stress. These studies provided insights into the failure mechanisms of composite stiffeners, but the behaviors of the stiffeners under compression, which is the most common load, remained unknown. In the tensile mode, compression is generated by an “outward” bending moment, i.e., while the moment-induced compressive-stress is generated at the outer-surface (at R = Ro) and the tensile-stress generated at the inner-surface (at R = Ri).
Few studies were on the failure mechanisms of structures subjected to an “inward” bending moment, in which cases tension is created at the outer-surface and compression at the inner-surface. Springer and Chang [
19] showed, using the finite element method, that under “inward” bending, the beam failed primarily via delamination at small radius-to-thickness ratios (R
i/H < 0.3–0.5) and the in-plane failure was at large radius-to-thickness ratios. Helenon et al. [
20] used experimental and numerical methods to study the failure of 2D composite T-stiffeners at three out-plane bending angles (i.e., β = 0°, 45°, 90°). Their results revealed that the maximum free-edge principal transverse stresses occurred at the failure locations perpendicular to the fiber direction. A conference paper reported 2D L-beams with three different layups were subjected to a compressive load [
21], and using cohesive zone modeling, their results revealed that the maximum interlaminar shear strength occurred at the lower end of the laminate and the initial failures for all three layups were dominated by the interlaminar tensile stress. Burns et al. used three novel designs to increase the failure stress and damage limit of 2D T-joints, using bending tests and finite element modeling [
22]. Their results showed that the initial damage load was ~125% higher than that of a conventional T-joints. Furthermore, a similar progressive failure mode was observed in all T-joints, in which delamination initiated in the radius-bending region, followed by radius bend/delta-fillet interface cracking. There is an apparent lack of studies on 2.5D and 3D composite curved structures when they are subjected to compressive bending.
This study tried to provide an original and preliminary investigation of 2.5D woven fabric composites’ mechanical performances under compressive bending, along with the 2D plain weave (2DPW) and 3D orthogonal woven (3DOW) composites as references. Two 2.5D woven structures were selected and their fabrics produced for this work, namely a shallow-straight (3DSSW and a shallow-bend (3DSBW). The effect of woven structures on the strain distribution and failure modes was evaluated via Digital Image Correlation (DIC) and X-ray Computed Tomography (CT) scans. Considerable efforts were made to produce the carbon fiber woven fabrics and the epoxy composites so that their fiber volume could be controlled at the same level, therefore rendering meaningful and valid comparison of their properties.
5. Conclusions
This study demonstrated a viable option of using highly crimped 2.5D woven fabrics to fabricate composite stiffeners having compressive bending loads and interlaminar shear resistance equivalent to or better than that of the expensive 3D composite approach. L-beams were selected as representative stiffener components. L-beams of two types of 2.5D composites, namely 2.5D shallow-straight and shallow-bend woven fabric composites, together with a 2D plain weave and 3D woven fabric reference composites, were prepared from a standard modulus carbon fibers and subsequently infused with epoxy resins. The bending strength, deformation, and failure mode of these L-beams were obtained using a custom-built bending test fixture, and the data were acquired using DIC and Micro X-ray CT scans.
The study further revealed that the composite fabrication process, namely VARI, had a significant effect on the bending strengths of the L-beams, particularly at the through-thickness direction, by altering the fiber orientation and/or yarn waviness. Such alterations generally did not exist in 2D composite fabrications, and should be considered when designing 2.5D and 3D composite parts. While the presence of Z-yarns were favorable in term of achieving high through-thickness performances, similar effects could be obtained via the utilization of highly crimped warp yarns, such as in the case of 2.5DSSWC.
As factors controlling the stability of the stiffeners, the L-beams’ stiffness and strain distribution were controlled by their woven structures. The data in this work demonstrated that the conventional 2D laminate stiffeners that had their maximum strains in the supporting arms, therefore had low stiffness than the 2.5D and 3D L-beams, as expected.
The outcome of this study would provide guidance to the selection of appropriate stiffeners, as well as help designing higher performance composite structures, both effectively and economically. The data thus collected would enrich the database of 2.5D and 3D woven composites for further numerical simulation via finite element modeling.