Ductile failure of Inconel 718 during flow forming process and its numerical investigation

Understanding the fracture mechanisms and predicting failure in forming processes is of great interest to increase formability and numerical methods have been proven effective. Incremental forming processes such as flow forming possess challenges due to material experiencing complex stress states and highly non-linear deformation histories during forming. Although some of the ductile failure criteria integrated into finite element (FE) simulations are found to be in better correspondence with the experimental findings, they mostly lack sufficient accuracy in predicting formability limits for flow forming and spinning processes. In this context, this work is concerned with the analysis and prediction of ductile failure of nickel-based super alloy Inconel 718 (IN718) in the flow forming process through FE analysis. Failure is modeled with three different models, namely Cockcroft-Latham (CL), Johnson-Cook (JC) and modified Mohr- Coulomb (MMC), implemented as a user subroutine in explicit FE solver Abaqus. Tensile tests are conducted with four different specimen geometries in order to identify parameters of the plasticity and damage models. Then, the developed framework is applied to a FE simulation of the flow forming process to predict formability limits and perform process parameter optimization to improve formability. Flow forming tests conducted at three thickness reduction values reveal that the forming limit of the material is at about 50% thickness reduction where radial cracks are observed on the outer surface. It is found that the CL model is able to predict formability limits and crack initiation locations while the MMC and JC models failed in both with significant over prediction of damage. Furthermore, the effect of temperature and process parameters such as the ratio of feed rate to revolution speed and roller axial and radial offsets are further examined with the developed FE frameworks. It has been shown that process parameter optimization requires accurate modeling of ductile damage since the behavior of different models is diverse.