ANALYZING THE PROGRESSIVE DAMAGE OF COMPOSITES DUE TO MANUFACTURING INDUCED DEFECTS THROUGH A SEMI-DISCRETE DAMAGE MODEL

Abstract

During composite manufacturing processes, multiple steps are involved, each step introducing new physical and chemical processes. These processes alter the properties of the constituents (fiber and matrix), affecting the behavior of the composite materials. The changes in the properties of the fiber are not that significant. Whereas the mechanical and thermal properties of the matrix such as Young’s modulus, Poisson’s ratio, coefficient of thermal expansion etc. change significantly. Residual stresses are developed in the composite due to thermal expansion mismatch of the constituents and cure shrinkage of the resin. These resulting residual stresses have a considerable impact on the mechanical properties and performance of the composites. Also, cracks develop in the composite system during the manufacturing process which can affect the performance of the composite. When the composite system with residual stresses is mechanically loaded, the system exhibits a drop in the strength after a critical stress state is reached. To predict the critical stress at which the drop occurs we use the continuum damage method called the Smeared Crack Approach (SCA). Using SCA we can predict the stress-strain behavior of an RVE as the damage progresses. The critical stress value of the microscale RVE acts as the corresponding strength of the composite on a macro-scale level. To account for the variability in a composite system we randomly generate different RVEs. This is done by varying the number of fibers and the location of fibers using a statistic distribution for a fixed volume fraction. Then we conduct the simulations of manufacturing and progressive damage to identify the strength values of the RVEs. The strength values obtained are then used to assign to the different regions of a semi-discrete damage model of a composite laminate in macro-scale. The unique aligned meshing strategy of the model decomposes the bulk non-linearity and localization zones which provide a proper load transfer pathway. This random assignment of the strength values simulates the realistic behavior of a composite where each region has almost the same material properties but different strength values due to the uncertainties associated with the manufacturing processes.