Experimental investigation of the dynamic response of graphite-epoxy composite laminates under compression

In this study, the dynamic stress–strain response of graphite-epoxy composite laminates is investigated. The laminates are interposed in a section of a split Hopkinson apparatus. A quasi-rectangular wave is generated at one end of the incident bar by striking it with another bar of known length. This bar is accelerated using a compressed air gun. Approximate average stresses and strains can be obtained by measuring the incident, reflected and transmitted waves in the split bar. The dynamic behavior is evaluated for a range of impact velocities. The dependence of the response on impact velocity is analyzed and discussed. Three different specimen thicknesses have been used. These are obtained by increasing the repetition factor of a base stacking sequence: (+45°, −45°, 0°, 90°). This process is called sublaminate scaling; it is preferred to ply scaling since it has been shown that the accumulation of layers of the same orientation decreases the failure load to such an extent that residual stresses may crack the specimen before any external load is applied. The laminates considered are: (+45°, −45°, 0°, 90°)_ns, n=2,3,4. The scale effects observed in the experimental response are analyzed and discussed.     

A micro-to-meso sublaminate model for the viscoelastic analysis of thick-section multi-layered FRP composite structures

Classical ply-by-ply analysis of multi-layered thick-section composite structures with tens of layers through the cross-section is often impractical, especially when material nonlinearity and time-dependent effects are included. This study introduces an integrated micromechanical-sublaminate modeling approach for the nonlinear viscoelastic analysis of thick-section and multi-layered composite structures. The sublaminate model is used to generate three-dimensional (3D) effective nonlinear responses at through-thickness material integration points with given spatial variations of strains determined from the trial strain increments of the standard displacement-based finite-element (FE). The number of material integration points is determined by the resolution of the FE discretization of the composite structure. The sublaminate model at a selected material point represents the effective nonlinear continuum behavior in its neighborhood using the 3D lamination theory with uniform in-plane strain and out-of-plane stress patterns through the representative layers. Therefore, the sublaminate has first-order stress and strain paths and cannot recognize the local sequence of the layers. While this approach is very effective approximation especially in the case of a very large number of repeating layers using relatively few elements (integration points) through the thickness, it cannot be used to represent the interlaminar stresses or bending/extension/twisting coupling effects within a sublaminate. A previously developed micromechanical model by the authors for a nonlinear viscoelastic unidirectional lamina is used for each layer in the sublaminate. The proposed modeling approach is first calibrated and verified against creep tests on off-axis glass/epoxy performed by Lou and Schapery (J. Compos. Mater. 5:208–271, 1971). Analyses for different thick-section laminated structures are presented using the integrated sublaminate with both shell and 3D continuum elements. The proposed 3D nonlinear time-dependent sublaminate model is computationally efficient and robust in analyzing multi-layered composite structures having large number of plies.