Experimental and Numerical Investigations on Fracture Mechanics of Composites
High fidelity failure prediction in composite materials and structures has been the subject of intellectual pursuit by both the industrial and academic communities for the past fifty years. To this end, we have adopted continuum and fracture mechanics of the problems as “forward analysis”, where linear and angular momentum is balanced, constitutive relations are perfectly described, and laws of thermodynamics are fully satisfied to obtain the governing equations of the system. Given the boundary conditions (loads and constraints), material properties and geometry, we can employ mesh-based/meshless approaches to solve the forward problem based on relevant simplifications of the kinematics. Such a solution provides full-field distribution of deformations, strains, stresses, and even damage/crack propagation for composite structures. As presented in Figure 1, we have proven that our peridynamics (PD) laminate model can capture excellent crack path predictions at macro scale as compared to the experimental fracture surfaces.
Additionally, we have thoroughly scrutinized the role of Halloysite nanotubes (HNTs) on the physics behind the deformation mechanisms and damage development in carbon fibre-reinforced polymer (CFRP) composites by using multi-instrument measurement and peridynamics. The nanotubes prevent the coalescence of microcracks by blocking crack propagation or diverting its path. Figure 2 presents an experimental-numerical mental image of this scenario, showing the hindrance of crack growth, crack tip splitting, and prevention of crack coalescence by HNTs clusters. In addition to experiments, by employing PD theory, we have purely numerically investigated that the branching and deflecting behaviour of macro (main) cracks in presence of multiple number of micro-cracks at the vicinity of the crack tip.
Multi-Scale Damage Analysis using Inverse-Forward Methods and Multi Measurements
As many composite structures often contain brittle orthotropic fibres, brittle/ductile isotropic matrices, and soft cores being able to observe and predict failure necessitates that we combine measurement (experimental) and prediction (numerical) of micro-to-macro phenomena. In this context, we are eager to answer the following vital questions: Where does the damage nucleate/initiate and why? How many damage/cracks will evolve and in what direction? How will the mechanical properties of the pristine material adapt/change in this condition? What will be the scale of the damage, e.g., micro/meso/macro?
Thus far, we utilized multi-instrument measurements systems such as Acoustic Emission (AE), Digital Image Correlation (DIC), Infrared Thermography (IRT), and Fibre Bragg Grating (FBG) sensors to understand the specific damage mode and accumulation in thin/thick hybrid fibre-reinforced laminates with various stacking sequences subjected to pure bending/tensile loads. As shown in Figure 1, the combination of DIC-SEA enables early prediction of susceptible interlaminar delamination transition zones (from Region I to II) at stress levels 30% below material strength.
In our investigation on thick hybrid carbon/glass composites depicted in Figure 2, compressive stresses accumulate under the loading tip during flexural tests, thus triggering fibre micro-buckling (or kinking) and/or matrix yielding in the critical failure region. These kink band formations result in onset of transverse cracks, thereby leading to delamination at plies closer to the mid-plane of the specimen. Besides, as sketched in SEM micrograph and DIC graphs, the kink band formation at the top layer is more prominent in specimens containing carbon plies on the faces due to the low compressive strength of carbon fibres. Additionally, mode-II delamination initiation is observed in the 3C specimen due to shearing forces. Further, the macro damage is triggered by matrix cracking and subsequent delamination onset in the carbon layer of the 2C sample according fractography/DIC results.
