Studies on large scale bridging in delamination of unidirectional and cross-ply laminates

Delamination propagation in fibrous laminates is often accompanied by development of a damage zone, within which intact fibers bridge the crack faces that constitutes a major toughening mechanism. A direct experimental assessment of the crack closure tractions due to bridging fibers is a challenging task. In addition, when the length of the bridging zone becomes comparable with the laminate linear dimensions, referred to as large scale bridging, the extent of bridging can be influenced by the laminate size and layup. Hence, traction-separation relations used in computational methods to predict fracture do not always rely on solid experimental evidence of large scale bridging and for numerical convenience simple linear traction-€“separation expressions are often considered. In this work, a semi-experimental methodology is employed to identify the contribution of large scale bridging to delamination resistance of unidirectional and cross-ply carbon fiber/epoxy laminates. Double cantilever beam specimens of different thickness are prepared and subjected to monotonic and fatigue mode I loadings. In selected specimens, optical fibers with multiplexed fiber Bragg grating (FBG) sensors are embedded. The strain distribution in the vicinity of the crack plane is monitored by FBG sensors. The strain data are used in an inverse identification procedure to quantify the bridging tractions and the energy release rate (ERR) due to the bridging. Subsequently, the identified tractions are implemented in a cohesive zone model to simulate delamination propagation. Employing the aforementioned approach, specimen thickness dependence of large scale bridging in delamination of unidirectional laminates is investigated. The results indicate that onset of delamination is independent of the specimen geometry. However, the extent of the bridging zone and the plateau level of delamination resistance both significantly increase with increasing the specimen thickness. The maximum traction at the crack tip is independent of the specimen size while the length of the steady state bridging zone scales linearly with the thickness of laminate. Moreover, the nonlinear rate of tractions decay is inversely proportional to the specimen thickness. It is shown that the tractions decay parameter is controlled by the curvature of the specimen arm. Evaluation of bridging effects on fatigue delamination of unidirectional laminates indicates that the stable phase of fatigue crack growth is more extended in the thicker specimens and the critical ERR increases by increasing the specimen thickness. Similar to the trend observed in monotonic delamination, fatigue delamination in the thicker specimens is accompanied by a longer bridging zone in which the bridging tractions decay with a lower rate. The maximum bridging traction at the crack tip is independent of the specimen thickness. It is demonstrated that the traction-separation relation and the ERR due to the large scale bridging depend on the ply orientations at the crack interface whereas ERR values at the onset of delamination are almost the same. As compared to delamination of a unidirectional specimen, crack propagation in cross-ply laminate is accompanied by a smaller bridging zone, but fiber bridging in the cross-ply specimens is more efficient. The adopted methodology in this work provides a robust assessment of large scale bridging and is a step towards a better understanding of delamination in fiber reinforced composites.

Botsis, John
Lausanne, EPFL
Other identifiers:
urn: urn:nbn:ch:bel-epfl-thesis6933-0

Note: The status of this file is: EPFL only

 Record created 2016-04-25, last modified 2018-05-01

Download fulltext

Rate this document:

Rate this document:
(Not yet reviewed)