Behaviour and Modelling of Fibre-Reinforced Polymers
Master thesis
Permanent lenke
http://hdl.handle.net/11250/237520Utgivelsesdato
2014Metadata
Vis full innførselSamlinger
Sammendrag
This thesis evaluates the behaviour of polypropylene (PP) and glass fibre reinforced polypropylene with a glass fibre content of10% and 30% by weight. Reinforcement materials are added to plastics to improve their mechanical properties and to reduce cost when compared to materials of similar strength. The difference in mechanical behaviour with increasing fibre content was examined using uniaxial tensile tests, bending tests and tensile tests on a plate with a centric hole. The application of fibre reinforced polymers have increased in the last years, and so has the need for a good material model. SIMLab at NTNU have created a material model used for ductile polymers and are now in the process of making a brittle polymer model. This model was tested on the 30wt% PP. The model was calibrated using the tensile tests and validated using the plate tests. The experimental tests showed that unreinforced polypropylene and glass fibre reinforced polypropylene are two different classes of materials. Whereas PP is ductile and isotropic, fibre reinforced PP is brittle and displays anisotropic behaviour. The material model was quite simple and included only anisotropic elasticity and brittle damage and fracture. This model worked well on the material in the longitudinal direction. In the 45 and 90 degree direction however it underestimated the maximum stress and strain. The material parameters in the material model were dependent on the material direction, and the damage parameters were not. The experimental tests showed that there were different fracture mechanisms in the different directions, and therefore the damage parameters should also be dependent on the material direction. The experimental tests on the plate with a centric hole revealed that the fibre reinforced material is also strain-rate dependent. This is something the material model does not take into account, and therefore, the material model either has to be calibrated at the same strain rate as it should be used, or the material model should take strain rate into account.