Abstract

This paper presents the results of a numerical study in which different hardening models were compared. More specifically the following three hardening models were considered: isotropic hardening, combined isotropic-kinematic hardening and combined isotropic-kinematic-distortional hardening. Furthermore, all three models were used in combination with the same Hill 1948 yield surface. The isotropic hardening model was used to simulate tensile tests in different directions and a torsion test. It was then tried to reproduce the outcome of these simulations with the more advanced hardenings models. Therefore, the advanced models were calibrated based on one of the tests simulated with the isotropic hardening model. It was expected that different hardening models would predict the same behaviour for monotonous strain paths, but this is not the case for the models considered in this study. This seems to be an inconsistency complicating the calibration of more advanced constitutive models. In this paper it is shown that this inconsistency can be solved by scaling some parameters in the evolution equations of the kinematic and distortional hardening model with a ratio of two equivalent stresses.

Abstract

Manufacturing Process Simulation (MPS) is on the rise nowadays to predict manufacturing defects and deviations, especially during the design phases where it allows the engineers to take earlier decisions and helps to reduce the manufacturing trials to go for a Right at First Time philosophy. MPS, focusing on composite materials, lies on the physical and chemical behavior of the material during the different stages of manufacturing (e.g. forming, infusion, curing, etc.) to predict defects, distortion and residual-stresses in the final component.

Airbus Group has developed simulation tools and defined procedures to obtain an accurate prediction of the shape of a thermoset CFRP component. These simulations are run by using a high-fidelity description of the material considering its chemical and physical behavior during the full curing cycle. Curing distortion shape is obtained by simulating the cure cycle on the properly meshed component. This procedure is shown on an industrial component, where the final deformation is validated with experimental measurements and it is linked with additional manufacturing process simulations.

The experience gained shows that future research should focus on linking up simulations of successive manufacturing steps: start by simulating the lay-up process or the forming, then carry across predicted fields such as real local fiber angle and volume fraction into the curing distortion prediction and so on up to high-fidelity simulation of the assembly of the airframe.  A particular point of interest is the viscoelastic behavior in post-curing, as well as the analysis of the stress relaxation in ply overlaps during curing.