An official website of the United States government
Abstract The vacuum-assisted resin infusion mold (VARIM) process is widely used in wind blade manufacturing for its cost-effectiveness and reliability. However, the current method faces challenges such as long curing times and defects due to nonuniform heating across the blade structure. To address this, a multi-zone heated bed setup tailored to blade thickness has been considered. However, determining an optimal temperature for each zone poses a computational challenge, which can be tackled with a novel machine-learning approach. Using a digital twin based on a high-fidelity multiphysics solver, a time-distributed LSTM model was trained to understand complex resin curing dynamics. This eliminates the need for costly lab experiments, as the model learns heating patterns and curing behavior efficiently. Once trained, the ML model acts as a digital twin by predicting the degree of cure for a given temperature setpoint with 96.73% accuracy. This model, when used as a surrogate for a Nelder-mead optimization workflow, improves the curing time by roughly 12.5% and presents a more uniform curing rate throughout the part.
more »
« less
This content will become publicly available on November 12, 2026
Digital Twin – Driven Machine Learning Optimization Framework for Multizone Curing Control in Wind Turbine Blade Manufacturing
Abstract. The Vacuum-Assisted Resin Infusion Molding (VARIM) process is widely used in wind turbine blade manufacturing due to its cost-effectiveness and reliability. However, challenges such as prolonged curing cycles and defects caused by non-uniform cure remain persistent. To address these issues, multizone heating systems have been developed to enable independent temperature control across blade sections. Yet, optimizing the temperature profile in each zone is computationally intensive, requiring detailed modelling of curing kinetics and heat transfer mechanisms. To overcome these challenges, in this work, a machine learning (ML) based digital twin of the VARIM process was developed using a time-distributed long short-term memory (LSTM) network trained on data generated by a high-fidelity multiphysics solver. The model achieved a predictive accuracy of 96.7 % in replicating the resin curing behavior. Its time-distributed architecture effectively captures the spatial – temporal dependencies across multiple zones, allowing precise prediction of the degree-of-cure evolution. Paired with a gradient-free optimization algorithm, the digital twin reduced curing time by 12.5 % while improving cure uniformity. This AI-driven framework eliminates costly trial-and-error experimentation, and provides a scalable, adaptive solution for improving both quality and productivity in wind turbine blade manufacturing, with strong potential for extension to other composite manufacturing processes.
more »
« less
- Award ID(s):
- 1916776
- PAR ID:
- 10658503
- Publisher / Repository:
- Copernicus Publications
- Date Published:
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
null (Ed.)Cure shrinkage of the polymer matrix during the composite manufacturing process leads to residual stresses, which can adversely affect the structural integrity and dimensional stability of composite structures. In this paper, a novel approach is developed for measuring the resin shrinkage and strain evolution of an epoxy resin (EPON-862) in the composite manufacturing environment. The resin is cured in a custom-designed autoclave with borosilicate viewports, while digital image correlation (DIC) is used to analyze the strain evolution throughout the cure cycle. These processing induced strains are correlated to the cure-state using differential scanning calorimetery (DSC). The different mechanisms involved in the polymer strain evolution during composite processing are discussed.more » « less
-
This paper presents a method for the digital reconstruction of the geometry of a wind turbine blade from a point-cloud model to polysurface model. The digital reconstruction of the blade geometry is needed to develop computer models that can be used by architects and engineers to design and analyze blade parts for reuse and recycling of decommissioned wind turbine blades. Initial studies of wind-blade geometry led to the creation of an airfoil database that stores the normalized coordinates of publicly-available airfoil profiles. A workflow was developed in which these airfoil profiles are best-fitted to targeted cross-sections of point-cloud representations of a blade. The method for best-fitting airfoil curves is optimized by minimizing the distance between points sampled on the curve and point-cloud cross section. To demonstrate the workflow, a digitally-created point-cloud model of a 100 m blade developed by Sandia National Laboratory was used to test the reconstruction routine.more » « less
-
The purpose of this study was to develop a replicable methodology for testing the capabilities and characteristics of a wind turbine blade in a structural re-use application with the specific goal of creating and demonstrating an efficient and commercially viable wind blade pedestrian bridge design. Wind energy experienced a dramatic increase in popularity following the turn of the century and it is now a common source of renewable energy around the world. However, while wind turbines are able to produce clean energy while in service, turbine blades are designed for a fatigue life of only about 20 years. With the difficulty and costs associated with recycling the composite material blades used on the turbines, wind power companies choose to dispose of decommissioned blades in landfills instead. The Re-Wind BladeBridge project aims to promote a more sustainable life cycle for wind power by demonstrating that decommissioned wind turbine blades have the capability to be repurposed as structural elements in bridges. This paper presents an analysis and characterization of a LM 13.4 wind blade from a Nordex N29 turbine, along with a design for a pedestrian bridge using two LM 13.4 wind blades to create a 5-meter span bridge. Software developed by the Re-Wind team called “BladeMachine” was used to generate the engineering properties of the blade at multiple sections along the blade length. Resin burnout tests and mechanical testing in tension and compression were performed to determine the material and mechanical properties of the composite materials in the blade. Additionally, a four-point edgewise bending test was performed on a 4-meter section of the wind blade to evaluate its load carrying behavior. The results of these tests revealed that the LM 13.4 blades are suitable to be re-utilized as girders for a short-span pedestrian bridge. An overview of the design of the BladeBridge currently under construction in County Cork, Ireland is presented, including details on the architectural and structural design processes.more » « less
-
Radio-frequency (RF) heating of thermosetting epoxies is an agile method to decouple the extrudability of epoxy resins from their buildability for additive manufacturing. Through this method, the resin is extruded in the liquid state at the early stages of curing. Then, an RF applicator induces a rapid and uniform increase in temperature of the resin, accelerating the solidification of the printed feature. Understanding the evolution of the resin's RF heating response as it cures is therefore critical in meeting the demands of additive manufacturing. In this work, we show that the high-frequency dielectric loss, determined using in situ rheo-dielectric measurements, of both neat and carbon nanotube (CNT) filled resins is correlated to the heating response at different temperatures throughout curing. Furthermore, we show that the presence of CNTs within the resin augments the heating response and that their dispersion quality is critical to achieving rapid heating rates during the cure.more » « less
