Inverse detection of constituent level elastic parameters of FRP composite panels with elastic boundaries using finite element model updating

Fiber reinforced plastics (FRP) panels are extensively used in ship structures due to their superior specific stiffness and strength as compared to metals. However, their prolonged use may result in degradation of material properties as well as the boundary conditions; thus, affecting the dynamic performance. Moreover, the changes in material properties are mostly at constituent level, i.e. fiber or matrix. A vibration based inverse identification technique is proposed using finite element model updating to estimate the constituent elastic material parameters of FRP panels having elastically restrained boundary. The objective function is formed from the difference of experimental as well as finite element prediction of dynamic responses. A gradient based optimization viz. the inverse eigensensitivity method is implemented. A set of numerically simulated examples is presented to demonstrate that the prediction of material parameters can be grossly erroneous if the boundary elasticity is overlooked. The algorithm is found to be robust even when the 'experimental' data is sparse and contains random noise. The method can be used for condition assessment and damage detection of FRP ship panels. The technique is novel as for the first time constituent level elastic parameters of FRP panels having elastic boundaries are estimated from dynamic responses.