Experiments, modeling and parameter identification for inhomogeneous distortion states of plastics with induced anisotropy

The project deals with the experimental characterization, modeling and parameter identification of polycarbonate (PC) films which receive a so-called "self-reinforcement" by cold-forming (pre-stretching). In this way, the initially isotropic material experiences a strain-induced anisotropy. Thus, properties such as strength and ductility can be influenced depending on the desired loading direction, which is utilized during cold-forming of plastics. The main aim is a comprehensive experimental characterization of the three-dimensional mechanical properties and on this basis to improve an own developed model which can simulate strain-induced anisotropy during cold-forming. The basic knowledge on cold-forming polymers, for example during manufacturing of stretched and thus self-reinforced thin films, will be extended by results of this project.

The experimental part deals with the extension of own developments for induced anisotropy in Dammann, Caylak, Mahnken (2014) for tensile bars and films made of PC by means of "sequential biaxial loading". Optical measurements support the determination of highly inhomogeneous strainfields. Further objectives are to investigate the local material behavior and the compressible inelasticity. For this purpose, the comparison of two independent methods for the determination of the volume-strain in preliminary experiments shows promising results. Additionally, findings on hardening and softening are essential, as preliminary investigations indicate. Instead of the globally observed softening, locally this effect hardly occurs. Different global strain rates will be characterized.

The material modeling part is concerned with the extension of own developments on strain induced anisotropy in Mahnken, Dammann (2014). Therefore, a structure tensor which develops in dependence on the loading direction is used to represent the anisotropy by means of weighting functions in dependence of the current so-called loading angle. Furthermore, the project aims to introduce a regularization by gradients to avoid FE-mesh dependencies and to include a volumetric flow function into the existing model. In doing so, new information from the experiments, indicated by preliminary investigations and not known from the literature, is considered.

Finally, parameter identification is intended. In addition to "homogenized" stress-strain curves also the inhomogeneous displacement data are used for the identification of the regularization parameter by means of an inverse finite element method. With the obtained material parameters for the material model cold-forming processes, such as the cold stretching of films, are simulated.