Automated braiding of non-axisymmetric structures

Using a simple virtual trial and error optimization loop improves manufacturing accuracy

Significance 

Composite materials have gained prominence in different industries owing to their remarkable mechanical properties. The aviation industry, for example, has leverage its weight reduction capabilities to reduce the fuel consumption of the new generation aircraft. The use of composite materials has skyrocketed in the past decade because they reduce not only the mass of the parts but also the number of assembled parts. Fiber-reinforced composites can be manufactured in a myriad of fabrication methods. The epoxy pre-impregnated unidirectional layup approach, although convenient, lacks repeatability and automation with limits its practical applications. Other methods that can improve the repeatability and automation of the composite production processes, like automated fiber placement, are not commonly used because they are expensive and time-consuming.

Recently, braiding of reinforced fibers using a mandrel has been identified as promising for addressing the above limitations to produce high-quality composite products. It comprises two yarn groups rotating, interlacing and ultimately resting on the mandrel moving through the braiding machine. and rotating that improve the impact resistance capabilities of the fabric. It is characterized by high-speed manufacturing due to the simultaneous deposition of several yarns, leading to improved automation and repeatable manufacturing process. Notably, the mechanical properties of the braided parts are significantly affected by a set of parameters, especially the braiding angle dependent on the mandrel and yarn rotation speed for a given mandrel geometry. The relationship has been termed as forward and inverse solution problems. While the forward solution computes the braiding angle as a function of mandrel speed, computing the mandrel speed as a function of the braiding angle is the inverse solution. The complicated relationship between these parameters makes the development of braiding models and processes a challenge.

Several methods have been presented to solve the forward and inverse solutions, including kinematic, analytical, mechanically-enhanced kinematic and finite element methods. Although these approaches have produced promising results, they are mostly limited to forward solutions. The methods fail to produce a desirable speed profile for a desired local braiding angle required by engineering requirements. To this note, researchers from the Polytechnique Montréal: Marc Gondran, Yasmine Abdin, Mr. Yohan Gendreau, Professor Farbod Khameneifar, and Professor Louis Laberge Lebel developed a novel Iterative Inverse Solution (IIS) approach equipped with the braiding angle control to enhance the automated braiding of non-axisymmetric structures. The work is currently published in the journal, Composites Part A.

In their approach, the research team’s proposed IIS utilized a simplified forward solution to rapidly compute the braid angle profile beginning with an arbitrary speed profile. The obtained results were compared to the desired braiding angle and the error was used as the feedback variable to modify the arbitrary speed profile. This feedback loop is repeated through an optimization scheme a until the speed profile accurately produces the desired local braid angle on the mandrel surface. As such, the iterative approach can be compared to an experimental trial and error process, where numerous virtual braiding experiments are done in the span of minutes, inside a desktop computer. Experimental and numerical case studies were presented, and their results were discussed to validate the feasibility of the novel approach.

The Canadian scientists demonstrated the efficacy and practicability of the presented method for efficient braiding of different structures on variable cross-section, curved and non-axisymmetric mandrels. The model exhibited a strong performance as it lowered errors associated with the braided angle in every case study, thereby producing a high-accuracy braid angle in a time-effrincient manner. For instance, the actual braid angle error with the desired braid angle was only 0.7 ° off when neglecting the unstable start-up zone.

In summary, a novel IIS approach for an improved automated braiding process was demonstrated. The approach allowed for efficient correction of the parameters to reduce the number of actual braiding needed to obtain the desired solution, thus improving the braided angle accuracy. The superior performance, improved accuracy, and time efficiency are some of the advantages of the proposed approach, making it a promising bet to meet the industrial need for a highly efficient and accurate inverse solution. In a statement to Advances in Engineering, Professor Louis Laberge Lebel, the corresponding author pointed out that the simplicity of their proposed inverse solution is likely to generate more applications of automated braiding of non-axisymmetric structures.

About the author

Louis Laberge Lebel
Polytechnique Montréal, Canada
JSPS Fellow, Gifu University, Japan

Louis Laberge Lebel is associate professor at Polytechnique Montreal and since 2014. He is leading the Advanced Composite and Fiber Structures Laboratory (ACFSlab). Currently, he holds a visiting professor position at Gifu University, Japan, as a Japanese Society for the Promotion of Science (JSPS) research fellow, until February 2022. Professor Laberge Lebel has also worked in the aerospace industry as a materials and process engineer and researcher for more than 3 years at Bombardier Aerospace and Pratt & Whitney Canada. Prior to that, he was a JSPS post-doctoral fellow at the Kyoto Institute of Technology. Prof Laberge Lebel has contributed 31 journal publications and 9 patent applications. His main research interests are textile composites, manufacturing, modelling, nano- and bio-composites.

Reference

Gondran, M., Abdin, Y., Gendreau, Y., Khameneifar, F., & Laberge Lebel, L. (2021). Automated braiding of non-axisymmetric structures using an iterative inverse solution with angle controlComposites Part A: Applied Science and Manufacturing, 143, 106288.

Go To Composites Part A: Applied Science and Manufacturing

Check Also

Moderation of thermoplastic composite crystallinity and mechanical properties through in situ manufacturing and post-manufacturing tempering: Part 1 – Mechanical characterisation - Advances in Engineering

Moderation of thermoplastic composite crystallinity and mechanical properties