The overall goal is to quantify permeation rates in composites based on measurement of thermal strains. Therein, the magnitudes of thermal strains and permeation are connected to the density of inter-fibre-failure (IFF). The integration of optical FBG-sensors into the composite laminate presents the enabling technology for this approach.
A single ply of CFRP with unidirectionally (UD) aligned fibers possesses transversal isotropic properties, meaning fiber parallel properties of the ply are dominated by fibers and fiber perpendicular properties of the ply are dominated by the matrix. As a result, stiffness in longitudinal direction of a UD ply is larger by 1-2 orders of magnitude compared to the transverse stiffness. On the other hand, the coefficient of thermal expansion (CTE) is much larger in transverse direction than in fiber direction. The difference in CTE results in process induced thermal residual stresses when a multidirectional laminate is cooled down from high temperatures during curing in an autoclave down to room temperature. Thermal residual stresses acting between fibers and resin on micro-scale are neglected at this point.
The concept of a stress-free temperature (SFT) refers to a condition during curing of a laminate that serves as a reference point regarding the development of thermal residual stresses. Above this temperature the matrix is assumed to be viscoelastic enough so that any residual stress built in the material is fully relaxed. Below this temperature the increase of thermal residual stresses with decreasing temperature is essentially linear. When considering a cross-ply laminate at SFT consisting of equal numbers of 0° and 90° plies, there are no stresses present in the laminate on ply level. Reducing the temperature of the laminate leads to residual stresses in the single plies of the laminate. The magnitude of thermal residual stresses in the individual plies is a result of the difference between the free elongation of the unconstrained single ply and the mean elongation of the composite laminate. As the mean elongation of the laminate is a function of stiffness and CTE of its constituents and depends on the layup, a single ply inside the laminate is forced to contract more in fiber direction than it would when unconstrained. At the same time, the contraction of a single ply in transverse direction is restricted compared to the unconstrained state. As a result. reducing the temperature of a cross-ply laminate leads to compressive residual stresses in longitudinal direction of a ply parallel to the fibers and tensile stresses in transverse direction. When considering the stress state of a single ply in transverse direction, a critical state denoted by tensile stresses combined with low tensile strength in said direction is present. Depending on the magnitude of thermal residual stresses, transverse matrix micro cracks can occur and have been reported repeatedly as the result of sole cooling of a laminate from curing temperatures to room temperature. Considering the fact that liquid hydrogen is stored at -253 °C, the temperature difference between curing and application is significantly larger. Thus, much larger densities of matrix micro cracks can be expected at cryogenic temperature.
The density of matrix micro cracks plays an important role when analyzing the permeation through composites. As single plies in a composite are generally subjected to thermally induced transverse tensile stresses, the micro cracks in the material are pulled apart and thus opened. At high micro crack density, these cavities present leak paths through which permeation can occur, e.g. by hydrogen permeation in a composite tank. Simultaneously, the formation of transverse matrix micro cracks in a single ply affects the stiffness in transverse direction of said ply. Due to the local discontinuity in the matrix, the area of opened micro cracks does not contribute to the tensile load bearing capacity of the ply. As a result, the restraining effect of a cracked ply on the adjacent plies is reduced significantly. In turn, this means that the contribution of the cracked layer to the mean CTE of the laminate is reduced since the mean laminate CTE is a function of CTE and stiffness of the individual layers of the laminate. Increasing IFF densities are expected to result in an increased impact on the mean laminate CTE. Therefore, by monitoring the thermal strain of a laminate using FBG-sensors, changes in CTE can be detected.
The novel approach that is introduced is based on the following hypothesis:
Thermal strains in a composite correlate with the permeation rate of the composite based on the interlinkage of both parameters to the IFF-density.