45 th International Congress IFKT, 27 – 29 May 2010, Ljubljana, Slovenia 1 3D MODELLING OF WARP KNITTED STRUCTURES – FROM GEOMETRICAL TO MORE REALISTIC MECHANICAL MODELS Wilfried RENKENS 1 , Yordan KYOSEV 2 & Marcel BEIß 2 1 Renkens Consulting, Aachen, Germany 2 Hochschule Niederrhein – University of Applied Sciences, Mönchengladbach, Germany Abstract: In this paper an overview of the modelling for warp knitting structures during the last years, starting from first work in the area of 3D geometry like [1] and [12] up to latest development, presented in [6,11] is presented. The advantages and challenges during the geometrical modelling of single and double needle bed structures, with weft insertions, spacer fabrics, tubular fabrics etc. are discussed. Accent is taken on the verification of the models using images of real structures. Various structures are analysed and compared with their simulated virtual copies, where the differences are analysed quantitatively and qualitatively. Some possibilities to improve the accuracy of the 3D models using computational mechanics are presented. Aim is to predict the main mechanical properties numercially, as for instance tensile behaviour of the structures and compression behaviour of the spacer structures. Finally, the geometry can be exported as initial state for extended FEM and CFD calculation of these structures. Keywords: warp knitted structures, 3D-simulation, spacer structures, behaviour, prediction 1 INTRODUCTION Warp knitted structures are widely used in car interior products, sport clothing, medicine, textile reinforced composites, etc. During the last years different steps in the development of computer based simulation of these structures were presented. Today a geometrically correct calculation and presentation in 3D space of single bed and double bed structures is possible. The natural question arising after building the model is – how well matched is this geometry to the reality. The goal of this work is the explanation of the most important differences between simulation and real objects. These differences present the front end of the current research and are the tasks, which have to be solved during the next research step. In order to answer these questions photographs of warp knitted structures are compared to their simulated counterparts. At first the comparison is conducted on yarn level to compare parts of single threads. Afterwards the comparison is performed on loop level to discuss the orientation of threads in the loops and their configuration. In a final step a macro-sized picture is compared to analyze the shifting and position of the loops between the real product and its simulation. The structure of the threads is of no interest in this last steps. 2 STATE OF THE ART Different mathematically descriptions of the loop geometry exist for more than 50 years [7] but a detailed form of the loop still remains an ongoing research problem. An explicit classification of the process to model loops is e.g. described in [6]. Geometrical, mechanical (including the subgroups based on energy- and elasticity theory) and empirical models have to be distinguished. Göktepe and Harlok [1] demonstrated a process for modeling of simple knitted structures that was insufficient for industrial use. Robatille, et al. [12] presented processes for modeling of reinforced knitted fabrics with significantly thinner and more flexible looping threads than the warp, weft or diagonal yarns (Rowings). Both models however were developed for a small group of structures only and thus not suited for broad application. The first geometrical models for the description of general knitted fabrics on single bed machines were reported in [5] and [9]. The advancement of spacer fabrics on double bed machines was described in [10]. The influence of forces and the refining of the models with mechanical effects were presented in [4] and [11]. An example of a spacer fabric is given in Fig. 1.