Title: MICRO MECHANICS BASED SPRING BACK MODELING AND EFFECT OF DAMAGE ON NET SHAPE
Abstract: A micromechanics based Finite Element (FE) method integrated with Multi-Stage Spring-Back Analysis (MSA) is proposed to model spring back process and demonstrate the effect of damage on net shape. In theory, the micro mechanics approach is considered to be an enhancement to commercially practiced FE nonlinear/contact/thermal (implicit/explicit) solutions. The proposed FE modeling of composites during cure process contains two different steps; transient thermal and mechanical analysis. The methodology can be also able to predict the effect of springback/residual stress on net shape of thick composite sandwich (i.e., honeycomb core, adhesive layer) structure during the manufacturing process. The proposed analysis considers the following: a) matrix shrinkage, change in modulus, and coefficient of thermal expansion versus time/temperature; b) subsequent spring-back analysis by adding/removing thick sub-layers and film adhesive during second stage of co-cure/co-bond; and c) progressive delamination failure, and disbond of as-build part. In addition, MSA provides the user pertinent sensitivity information of key spring-back parameters to optimize the jig shape and the effect of defects (i.e., void shape/ size/distribution), and missing fiber, and resin rich, and part wrinkling. A building block validation strategy is adopted for prediction and validation of the proposed analysis against available experiments. The building block approach includes: 1) Material Modeling: micromechanics-based stiffness, strength and curing process prediction; 2) deflection and damage evolution of simple flat plate spring back analysis; 3) deflection and damage evolution of L-shape spring back analysis and 4) curved honeycomb sandwich structure evaluation including the effect of composite resin shrinkage and honeycomb. Flat panel and L-shape geometries are modeled in order to study the effect of stacking sequence on the stress distribution and spring-in-angle of the curved panel and compared with available test data.
Authors: Pooya Behroozinia, Harsh Baid, Frank Abdi, Dade Huang, Max Repupilli, Bob Minaie
Conference: CAMX 2017 –Orlando