Fiber Reinforced Polymers
Assembly Line
Vestas and Markforged
The Vestas team began researching alternative ways to improve their overall manufacturing process. Using Markforged’s cloud-based, AI-powered Digital Forge additive manufacturing platform, the company successfully launched its direct digital manufacturing (DDM) program in 2021. The program frees up manufacturing processes from relying on outside suppliers, and provides a knowledge base for collaboration.
The DDM program already includes 2000+ Vestas parts stored in a Markforged Eiger™ cloud-based digital repository. This allows employees at any Vestas location — with little to no additive manufacturing expertise — to quickly search for and print any number of fiber-reinforced composite parts on their local X7™ and composite parts on their Onyx One™ 3D printers.
According to Jeremy Haight, Principal Engineer — Additive Manufacturing & Advanced Concepts at Vestas, “Our approach is end-to-end. We provide the physical article in near real-time to a variety of places. It’s the closest thing to teleportation I think you can get.” Thanks to the repository, the Vestas team knows they will get consistent, up-to-spec parts at a moments notice, anywhere in the world, without the need for specialists at their global facilities. This has dramatically reduced shipping and freight costs, and manufacturing lead times.
SonoPrint: Acoustically Assisted Volumetric 3D Printing for Composites
Advancements in additive manufacturing in composites have transformed various fields in aerospace, medical devices, tissue engineering, and electronics, enabling fine-tuning material properties by reinforcing internal particles and adjusting their type, orientation, and volume fraction. This capability opens new possibilities for tailoring materials to specific applications and optimizing the performance of 3D-printed objects. Existing reinforcement strategies are restricted to pattern types, alignment areas, and particle characteristics. Alternatively, acoustics provide versatility by controlling particles independent of their size, geometry, and charge and can create intricate pattern formations. Despite the potential of acoustics in most 3D printing, limitation arises from the scattering of the acoustic field between the polymerized hard layers and the unpolymerized resin, leading to undesirable patterning formation. However, this challenge can be addressed by adopting a novel approach that involves simultaneous reinforcement and printing the entire structure. Here, we present SonoPrint, an acoustically-assisted volumetric 3D printer that produces mechanically tunable composite geometries by patterning reinforcement microparticles within the fabricated structure. SonoPrint creates a standing wave field that produces a targeted particle motif in the photosensitive resin while simultaneously printing the object in just a few minutes. We have also demonstrated various patterning configurations such as lines, radial lines, circles, rhombuses, quadrilaterals, and hexagons using microscopic particles such as glass, metal, and polystyrene particles. Furthermore, we fabricated diverse composites using different resins, achieving 87 microns feature size. We have shown that the printed structure with patterned microparticles increased their tensile and compression strength by ∼38% and ∼75%, respectively.
Inferring material properties from FRP processes via sim-to-real learning
Fiber reinforced polymers (FRP) provide favorable properties such as weight-specific strength and stiffness that are central for certain industries, such as aerospace or automotive manufacturing. Liquid composite molding (LCM) is a family of often employed, inexpensive, out-of-autoclave manufacturing techniques. Among them, resin transfer molding (RTM), offers a high degree of automation. Herein, textile preforms are saturated by a fluid polymer matrix in a closed mold. Both impregnation quality and level of fiber volume content are of crucial importance for the final part quality. We propose to simultaneously learn three major textile properties (fiber volume content and permeability in X and Y direction) presented as a three-dimensional map based on a sequence of camera images acquired in flow experiments and compare CNNs, ConvLSTMs, and Transformers. Moreover, we show how simulation-to-real transfer learning can improve a digital twin in FRP manufacturing, compared to simulation-only models and models based on sparse real data. The overall best metrics are: IOU 0.5031 and Accuracy 95.929 %, obtained by pretrained transformer models.