Get This Paper

Real-Time Process Optimization Using In-Mold Dielectric Analysis and Machine Learning


Title: Real-Time Process Optimization Using In-Mold Dielectric Analysis and Machine Learning

Authors: Alec Redmann, Alexander Chaloupka

DOI: 10.33599/nasampe/c.23.0051

Abstract: The processing parameters for thermoset and composite parts are often provided by the material supplier or determined in a laboratory setting and then transferred to the manufacturing floor. However, deviations in material batches, humidity, machine calibration, among other variables, can cause unpredictable effects in manufacturing and final part quality. This leads to black box manufacturing with conservative cycle times, a reliance on extensive post-production quality control, and the potential for significant waste. This study demonstrates how specialized dielectric analysis sensors, machine learning, and material models enable the measurement of critical material and process information directly during manufacturing. The viscosity, degree of cure, and gel-point are calculated in real-time and are used to dynamically control the process parameters. As the data collection continues over time during serial production, the machine learning algorithm is constantly retrained and refined to ensure constant quality and detect material or process deviations. An explanation of the measurement principles for dielectric analysis is presented, and a direct comparison is made with the traditional analysis methods of rheology and differential scanning calorimetry (DSC). Examples are demonstrated with RTM6-1 epoxy used for infusion of composite parts. The results show that in-mold dielectric analysis compares well to traditional laboratory techniques, with the added benefit of being capable of process integration for serial production and quality monitoring. When combined with machine learning, the data collection continues over time during serial production. The algorithms can then be constantly retrained and refined for process optimization.

References: Setter, R., Schmolzer, S., Rudolph, N., Moukhina, E., & Wudy, K., “Modeling of the Curing Kinetics of Acrylate Photopolymers for Additive Manufacturing.” Polym. Eng. & Sci. (2023): 1-20. doi:10.1002/pen.26353 2. Redmann, A., & Osswald, T.A., “A Model for Modulus Development of Dual‐Cure Resin Systems.” Polym. Eng. & Sci. 61(3) (2021): 830-835. doi:10.1002/pen.25628C 3. García-Manrique, J.A., Marí, B., Ribes-Greus, A., Monreal, L., Teruel, R., Gascón, L., Sans, J.A., & Marí-Guaita, J., “Study of the Degree of Cure Through Thermal Analysis and Raman Spectroscopy in Composite-Forming Processes. Materials 12(23) (2019): 3991. doi:10.3390/ma12233991 4. Zuyev, K.S., & Castro. J.M., “Applications of Chemo-Rheology to Develop Process Windows in Reactive In-Mold Coating.” J. Polym. Eng. 22(4) (2002): 233-260. doi:10.1515/polyeng.2002.22.4.233 5. Pisupati, A., Bonnaud, L., Deléglise-Lagardère, M., & Park, C.H., “Influence of Environmental Conditions on the Mechanical Properties of Flax Fiber Reinforced Thermoset Composites.” Applied Composite Materials 28(3) (2021): 633-649. doi:10.1007/s10443-021-09885-z 6. Chaloupka, Alexander. Development of a Dielectric Sensor for the Real-Time In-Mold Characterization of Carbon Fiber Reinforced Thermosets. (Dissertation) University Augsburg, Augsburg, Germany. 2018. 7. Hexcel. “HexFlow RTM6/RTM6-2 Product Data Sheet.” Accessed May 29, 2023.

Conference: CAMX 2023

Publication Date: 2023/10/30

SKU: TP23-0000000051

Pages: 11

Price: $22.00

Get This Paper