A finite element based approach for the accurate determination of the shear behaviour of textiles with the picture-frame shear test

摘要

Picture frame shear tests are state of the art for determining the shear force vs. shear angle behaviour for in-plane deformation of most technical textiles, such as woven fabrics. Many publications describe this test and the used picture frames. Benchmark tests showed that the measured shearing behaviour for one sample depends on the picture frame used. The shearing rigidity of most textiles is very small compared to the in-plane tensile stiffness, so slight imperfections on the experimental setup have a significant effect on the measured results. During the picture frame test, wrinkles may form on the sample surface during the motion of the picture frame above a critical shear angle. These wrinkles can be described as local fabric buckling. If forming of wrinkles leads to a lower level of internal energy compared to a further shearing of the fabric, local wrinkles occur due to the principle of least action. Because of this effect, the measured shear force above the first formation of wrinkles is inaccurate for describing the exact shearing behaviour of textiles. Another possibility for measuring the shear force vs. shear angle behaviour is the bias-extension test. Here, higher shear angles can be achieved without the formation of wrinkles. Both methods are compared in this paper for different textile samples. The relationship of the shear angle vs. the applied shear force is one of the most important input data in numerical drape simulations. The analysis of wrinkles, which occur during textile draping, demands exact input parameters for the simulation. Most important for the drape simulation of technical high-performance textiles are accurate values for the bending and shear behaviours. This paper presents simulation results of the wrinkling during a picture frame shear test. Results show that the model input data for the shear rigidity measured with the picture frame shear test do not reproduce the formed wrinkles in the simulations compared to experiments exactly. Therefore, they are not suitable for an exact drape simulation. The underestimation of the shear force vs. shear angle behaviour is shown with a finite element simulation model. The adaptation of the picture-frame and bias-extension parameters for a proper use in numerical drape simulations is examined.