Determination of critical energy release rates for steel-cfrp interfaces considering residual thermal stresses

Abstract

The combination of composites and metals to Fibre Metal Laminates (FML) is employed in some applications to overcome specific weaknesses of the single constituents. However, as the hybrid interfaces can often not achieve the adhesion quality like in monolithic materials, delamination is a prevailing failure type. With increasing mismatch of Young’s modulus the interlaminar stresses in the interface increase, leading to specific challenges for specific material combinations. This publication deals with the characterization of fracture toughness in FML, especially the combination of epoxy based CFRP with steel is investigated. This combination leads to relatively high mismatches of several properties and is therefore a very interesting subject of investigation. Further, in a laminate CFRP plies adjacent to steel layers exhibit an angle leading to interfaces like steel-0°, steel-90° or steel-45°, showing different crack initiation and propagation behaviour. Under this purpose, the critical energy release rates (ERR) in mode I and mode II are determined in static tests for all mentioned interfaces. Conducted are the well-known Double Cantilever Beam (DCB) and End Notch Flexure (ENF) tests. The results are evaluated based on beam theory and compliance calibration methods. Further, the cracked interfaces are investigated using microscopy. Residual stresses resulting from the different thermal expansion of the constituents during curing in an autoclave lead to an initial loading on the interface. Its portion has to be considered to determine the true mechanical fracture toughness. This thermal portion can be calculated by analytics following methods provided in the literature. Required are specific material properties as elasticities and thermal expansion coefficients, but also the temperature the bonding between steel and CFRP layers takes place. This so called stress free temperature is determined in a temperature dependent test setup. Thin asymmetrical hybrid specimens manufactured the same way as the DCB and ENF specimens show a strong curvature at room temperature, which is a result of the residual stresses. At the stress free temperature the specimens become plane, what can be observed in the employed experimental test setup. The determined ERR adjusted by the thermal portion are required input parameters for the crack propagation simulation employing the so called Virtual Crack Closure Technique (VCCT). This method is used to simulate the DCB and ENF tests with a predefined thermal load step. Coincidence of crack initiation and propagation of simulation and test is a proof of correct input values. Therefore, the simulation helps to evaluate the methods to obtain the adjusted ERR and the values themselves.