Implementation of the action plane fracture criteria of Puck into the Finite Element code Abaqus

Abstract

The topic of this Master's thesis is the implementation of the action plane fracture criteria of Puck into the Abaqus (3DS Dassault Systems, Paris, France) environment. The proper code will be developed in Fortran code. The necessity of using this criterion follows that in Finite Element codes the strength of fibre-reinforced plastic is typically computed based on an analytical function describing the yield surface. Therefore, it cannot be distinguished between different failure modes. The Puck's failure criterion singles out the fibre failure and inter-fibre failure. Even though Abaqus options include Hashin theory, which also distinguish IFF and FF, Hashin didn't take into account the fracture plane and in Abaqus Hashin's criteria is only available for 2D model. That is why this criterion is one of the best failure criteria for fibre-reinforced plastic available today. The Puck's failure criterion code will be done in an implicit way. It is important to check whether this code is working properly in cooperation with Abaqus software. There is a need of validation this implementation. The first step to achieve it, is to prepare some example based on literature and calculate them using Compositor composite's calculator (program prepared and developed by Instutut für Kunsttoffverarbeitung, RWTH Aachen). These calculations will serve in the next step of this thesis, when this results will be checked with the one achieved with Abaqus software and implemented code. To obtain the validation of action plane fracture criteria of Puck, results from numerical calculations must be similar with experiments ones. Ensuring that the code works correctly and it provides acceptable results, action plane fracture criteria of Puck will be used in modeling and calculating degradation process in composite structure. By using action plane fracture criteria of Puck there is a wish to obtain a properly defined model and thanks to that receive correct results for numerical calculations. Thanks to this work we achieved a subroutine that can allow us to calculate a 3- dimensional model of a lamina. It is important, because now we can model more complex of laminate structure, not only 2-dimensional models, which in many cases are not sufficient. The implementation will improve researches about degradation processes in laminate structures, which was not available before.

-