Towards Digital Twin for Composite Parts Lifecycle: from Manufacturing to Performance
Composites manufacturing industry is facing the management of defects along the entire lifecycle of the product. Digital twin tools allow to control and minimize various defects and optimize manufactured component performance by establishing a continuous and unambiguous flow of information during both manufacturing and product lifecycle. This THERMOPEDIA Focus note outlines a digital twin framework for integrated composites manufacturing and product design. Specifically, we focus on the digital twin for two highly challenging and important composites manufacturing processes: (i) the high-pressure resin transfer molding (HP-RTM), which has the ability of manufacturing composite parts in a short forming cycle (< 5 min), and the vacuum-assisted resin transfer molding (VARTM) typically employed for wind turbine blades.
Figure 1 depicts an automotive component manufactured by General Motors using HP-RTM technology. The part was manufactured in the 3.1 min cycle and eliminated 53% CF wastage using a glass fiber outside the part boundary.
Figure 1. An automotive component manufactured by General Motors using HP-RTM technology
By continuously collecting machine, process and part data, it is possible to detect defects within the production. Measurements within the processes as well as sensor data in the tool or on its surface further improve the accuracy of the digital twin. An ultrasonic scanning system and optical microscope can be employed to furnish the simulation engine with information of defects, void characteristics and part distortion. Reducing defects, such as voids, residual stress, and material degradation, increases the part's operational life and reduces the need for repair. Digital twins have the potential to automate production and provide the industry with virtual tools for future designs and improved processes.
Applying machine learning to the HP-RTM and VARTM manufacturing processes is critical to increase manufacturing efficiency and quality by reducing cycle times and energy consumption. An overview of the digital twin's hybrid data-physics driven simulation engine is presented in Figure 2. A validated, cost-effective high-fidelity multiscale-multiphysics framework is employed to virtually recreate the infusion and curing process of composites and the curing process of resins. Machine learning enhances the fidelity and predictability of process simulations to optimize the process parameters and mold parameters. Inputs and outputs of the simulations across the scales are listed in Fig. 2. Virtually manufactured components are subsequently virtually tested to determine performance and determine the materials-manufacturing-performance relation.
Figure 2. Schematics of the digital twin's hybrid data-physics driven simulation engine
REFERENCES
Cui, J., La Spina, A., and Fish, J. (2023). Data-Physics Driven Multiscale Approach for High-Pressure Resin Transfer Molding (HP-RTM), Computer Methods in Applied Mechanics and Engineering, 417, 116405.
Fish, J. and Yu, Y. (2023) Data-Physics Driven Reduced Order Homogenization, International Journal for Numerical Methods in Engineering, 124(7): 1620–1645.
Fish, J., Wagner, G., and Keten, S. (2021). Mesoscopic and Multiscale Modelling, Nature Materials, 20: 774–786.
Yu, Y. and Fish, J. (2024). Data-Physics Driven Reduced Order Homogenization for Continuum Damage Mechanics at Multiple Scales, International Journal for Multiscale Computational Engineering, 22(1): 1–14.
Yuan, Z., Felder, S., Reese, S., Simon, J.W., and Fish, J. (2020). A Coupled Thermo-Chemo-Mechanical Reduced-Order Multiscale Model for Predicting Residual Stresses in Fibre Reinforced Semi-Crystalline Polymer Composites, International Journal for Multiscale Computational Engineering, vol. 18(5): 519–546.
Yuan, Z., Felder, S., Reese, S., Simon, J.W., and Fish, J. (2020). A Coupled Thermo-Chemo-Mechanical Reduced-Order Multiscale Model for Predicting Residual Stresses in Fibre Reinforced Semi-Crystalline Polymer Composites, International Journal for Multiscale Computational Engineering, 18(5): 519–546.
References
- Cui, J., La Spina, A., and Fish, J. (2023). Data-Physics Driven Multiscale Approach for High-Pressure Resin Transfer Molding (HP-RTM), Computer Methods in Applied Mechanics and Engineering, 417, 116405.
- Fish, J. and Yu, Y. (2023) Data-Physics Driven Reduced Order Homogenization, International Journal for Numerical Methods in Engineering, 124(7): 1620–1645.
- Fish, J., Wagner, G., and Keten, S. (2021). Mesoscopic and Multiscale Modelling, Nature Materials, 20: 774–786.
- Yu, Y. and Fish, J. (2024). Data-Physics Driven Reduced Order Homogenization for Continuum Damage Mechanics at Multiple Scales, International Journal for Multiscale Computational Engineering, 22(1): 1–14.
- Yuan, Z., Felder, S., Reese, S., Simon, J.W., and Fish, J. (2020). A Coupled Thermo-Chemo-Mechanical Reduced-Order Multiscale Model for Predicting Residual Stresses in Fibre Reinforced Semi-Crystalline Polymer Composites, International Journal for Multiscale Computational Engineering, vol. 18(5): 519–546.
- Yuan, Z., Felder, S., Reese, S., Simon, J.W., and Fish, J. (2020). A Coupled Thermo-Chemo-Mechanical Reduced-Order Multiscale Model for Predicting Residual Stresses in Fibre Reinforced Semi-Crystalline Polymer Composites, International Journal for Multiscale Computational Engineering, 18(5): 519–546.