Recent advances in modern manufacturing methods have been applied to traditional textile-based machinery that enable the construction of extraordinarily sophisticated structures. These structures are capable of effectively utilizing advanced materials precisely and efficiently with high quality and repeatability. Despite all the advances in manufacturing capabilities of modern braiding machinery, lack of agile predictive modeling tools is a limiting factor for designers. One important traditional textile technology that has been modernized is braiding and overbraiding; a process that emulates the traditional Maypole Dance by interlacing fibers into near-netshape composite preforms. Developments of computer controlled machinery along with high performance synthetic fibers such as carbon has enabled commercial manufacture of advanced composite structures in automotive and aerospace applications.
One of the most important advantages of braiding is that it forms a continuous reinforcement throughout complex shapes; an attribute that is highly challenging to achieve by other composite manufacturing methods. Unfortunately, the conventional design and analysis methodologies of composite materials, such as finite element analysis, are somewhat lacking to represent realistic reinforcement orientation—due to complexity of creating useful 3D representations of braided structures and process specific knowledge requirements. Nonetheless, mastering the Maypole Dance just got a bit easier.
The general strategy for modeling textile composites is to use multi-scale approaches to account for elastic constants at each hierarchy where material properties are ultimately homogenized and applied to the macro or part scale. If one could directly use the topology of the braided structure in finite element analysis, there would be no need to perform multi-scale analysis on the part. In this view, further research is necessitated to ease the difficulty of defining material orientations in finite element analysis, attain higher model fidelity, bolster design confidence, and eliminate costly trial-and-error often associated with braided composites.
Recently, Dr. Yang Shen and Dr. David Branscomb, of Highland Composites in the United States presented a new and general method for generating 3D representations of braided structures using explicitly expressed formulas. To elucidate further, in their study they proposed a method that rapidly generates braided topology on arbitrary shapes with any non-concave cross-section. Their work is currently presented in the research publication, International Journal of Engineering Science.
The scientists employed kinematic equations to generate the basic formation of braided structures including yarn interlacing. Next, they employ the model in such a way that the mandrel is decomposed along the braiding direction, segment by segment. Each segment is defined with a cross-section and a centroid, with the centroids aligned collinearly on the Z axis and the cross-section planes oriented perpendicular to the Z axis to form representative fiber architecture. Moreover, the third step in the model involved translation of all the nodal coordinates obtained from the kinematic equations in the radial direction such that they were conformed to the contour of each cross-section. Eventually, translations and rotations were performed on all the nodal coordinates until they matched the approximate shape of the overbraided mandrel surface. Various computer aided tools were used during this modeling effort.
The authors demonstrate that the developed model generates representative 3D geometry on different cross-sections including curved mandrels. Interestingly, the geometric model is solely kinematic in nature and did not involve complex braiding process simulations, such as a dynamic finite element method. All in all, a computational model to generate the desired 3D geometry of an overbraided structure was established.
In summary, the study by Highland Composites scientists presented a new facile technique of generating 3D geometry of an overbraided structure on a non-straight mandrel with a convex cross-section. The presented model showed outstanding capabilities as its layer-by-layer deployment enables fell-points to be discretized and consolidated on each subsequent braid layer, after which the layers are transformed onto the desired centerline of the mandrel. The model was made versatile to enable its implementation in any computer language. Altogether, the presented model has potential to promote further adoption of braided structural components by easing the complexity encountered in the design phase, improving design confidence, reducing trial-and-error, and augmenting detailed finite element investigations through generation of rapid and realistic geometry.
Yang Shen, David J. Branscomb. A general approach to fast prototype the topology of braided structures. International Journal of Engineering Science, volume 131 (2018) page 40–60.Go To International Journal of Engineering Science