T-RTM Thermoplastic Resin Transfer Molding

 

Thermoplastics, with all their advantages, are difficult to use as matrices for high-performance composites, as their high viscosity complicates the impregnation process. Overcoming the high viscosity can be reactive thermoplastics, such as PA-6, which is synthesized from ε-caprolactam (ε-CL) monomers by anionic ring opening polymerization (AROP). The synthesis of PA-6 allows the manufacture of parts using thermoplastic resin transfer molding (TRTM).

 

Reactive thermoplastic systems consist of a mixture of monomers and/or oligomers (in cyclic or linear form), which are converted into polymers by the addition of initiators and activators. Due to their low molecular weight, monomers and oligomers exhibit a water-like viscosity and can easily impregnate dry reinforcements, thus overcoming the main problem of thermoplastic matrices in composite manufacturing.

 

Due to their low viscosity, reactive thermoplastics can be used in liquid composite molding (LCM-liquid composite molding) techniques, such as thermoplastic resin transfer molding (T-RTM) or vacuum infusion. Reactive thermoplastics polymerize in situ, i.e. in the mold.

 

Currently, many reactive thermoplastics suitable for in situ polymerization are available, including thermoplastic polyurethanes, polyesters, polyamides, polycarbonates and polymethylmethacrylates. The most promising of these polymers is PA-6.

 

The following is a report on the research and development of the European “CosiMocomposites for sustainable mobility” project

 

 

The project “Composites for Sustainable Mobility” (CosiMo-composites forsustainable mobility) was launched in 2018 by Faurecia Clean Mobility (Nantes, France) to develop a smart thermoplastic composite resin transfer molding (T-RTM) process that uses an integrated sensor network, including data from the smart mold and machine data from the injection molding machine and hot press.

 

The project will be completed by the end of 2021, and the partners of CosiMo have gained knowledge and experience in closed-loop control that combines real-time sensor data with material parameters from laboratory data and simulation data. CosiMo is funded by the Bavarian Ministry for Economic Affairs, Regional Development and Energy within the Campus Carbon 4.0 program.

 

The demonstrator part is 1100 mm long and 530 mm wide and was designed at Faurecia Clean Mobility’s research center in Augsburg, Germany. The goal was to explore material and process limits, including metal and foam cores, complex geometries, various radii and thicknesses ranging from 2.5 to 10 mm. The Institute of Textile Technology Augsburg (ITA) offers glass fiber nonwovens up to 1000 g/m².

 

The local reinforcement properties of glass fiber noncrimp fabrics (NCF-noncrimp fabrics) and carbon fiber unidirectional tapes were investigated. The glass fiber material allows the use of recycled fibers from end-of-life composite parts.

 

 

Based on part designs from Faurecia Clean Mobility and preform tools made in Augsburg, CosiMo’s T-RTM process uses a KraussMaffei (Munich, Germany) 3K (three-component) injection molding machine. The process is a derivative of the 2K process patented by Tecnalia (Donostia San Sebastián, Spain), which KraussMaffei demonstrated at the 2016 K Show in Düsseldorf, Germany. The 3K injection molding machine is combined with a 4,400-kilonewton Wickert (Landau, Palatinate, Germany) hot press, in which the steel RTM tool is installed. The RTM tool is made by Siebenhurst in Dietfurt, Germany.

 

 

The main goal of the project was to investigate the full automation and sensor-based process control of a single-step T-RTM process using a sensor RTM mold and a hot press. The German Aerospace Center (DLR) Center for Lightweight Production Technology (ZLP, Augsburg) manufactured around 100 demonstrator parts in the project. “The parts were made of different reinforcements with high quality and reproducibility,” says Jan Faber, project leader for CosiMo manufacturing at DLR ZLP.

 

As project leader for the work package “HAP 3 – Smart Tooling”, DLR ZLP also provided the key interface between basic material characterization and “HAP 4 – Data-driven process control” in the work packages “HAP 1 – Customized Nonwovens” and “HAP 2 – Reaction Systems”.

 

Sensor Network

 

More than 70 sensors were integrated into the steel RTM mold to monitor process parameters during resin injection and in-situ polymerization. The network includes various sensors, including combined pressure/temperature sensors from Kistler (Winterthur, Switzerland), dielectric sensors from Netzsch (Selb, Germany), and ultrasonic sensors developed at the University of Augsburg. The latter are based on the concept of widely used commercial piezoelectric sensors, but modified for the high temperatures and in-mold integration requirements of thermoplastic composite processing. A central data acquisition system provided by iba AG (Fürth, Germany) collects, processes, and then publishes material, process, and machine data to a machine learning platform.

 

Process Simulation and Optimization

The sensor network enables monitoring of resin flow and polymerization, and tracking of part parameters in complex tool geometries. Resin filling and polymerization behavior are analyzed in real time using process parameters such as temperature and pressure.

 

 

 

Process parameter data from sensors were also used to optimize the process simulation model. Based on a previously defined manufacturing demonstrator, subprocesses of part manufacturing (e.g., resin filling simulation) and properties of polymerized components (e.g., deformation simulation) were modeled using ESI PAM Composites software (ESI Group, Rungis, France). DLR ZLP compared these simulation results with actual process data to optimize the T-RTM process as part of the HP3Workpackage activities.

 

The project then developed automated and simulation data-driven process control based on machine learning methods. The Institute for Software and Systems Engineering (ISSE, Prof. Reif) at the University of Augsburg generated the machine learning training data and developed artificial intelligence (AI) models that predict the polymerization state, the time required for polymerization, and potential problems in the subprocesses of resin injection, filling, and polymerization. This was done in collaboration with Kuka (Augsburg, Germany), another key partner in the project.

 

A kinetic model describing the polymerization process was developed at the Institute for Materials Resource Management (MRM) at the University of Augsburg. Netzsch was also a key collaborator, transferring his knowledge of thermoset process monitoring to thermoplastic process monitoring to help build the ability of the predictive AI model to predict when PA6 polymerization should be complete and in what state.

 

Part Manufacturing and Testing

For part production, the steel mold is heated to 170°C and the reactive caprolactam resincomponents are heated to 120°C. The preform is placed in the mold, which is then closed. After a brief dwell to allow the preform to reach mold temperature, the caprolactam monomer is injected, which takes 20-25 seconds. Process simulations showed that in-situ polymerization after injection took 5.7 minutes with a conversion of 98.5%. After this, the press was opened and demolding was performed at 170°C.

 

The degree of polymerization is confirmed by infrared spectroscopy, rheometer data and polymerized DEA sensor analysis. Finished parts are evaluated using nondestructive testing including microscopy, thermal imaging and air-coupled ultrasound. Part quality is correlated with process sensor data and compared to similar fiberglass/thermoplastic parts on the market.

 

Future Developments

 

For DLR ZLP, the project has been a success, says Faber. “We have completed our part and developed a lot of knowledge about using sensors to achieve fully digital closed-loop control of composite processing,” he says. “For industrial series production, nobody would install so many sensors. That’s not what we expect.

 

But for this research investigation, this large sensor network is very precise and helps us gain a comprehensive understanding of the process and material behavior. We can see where the process is changing due to part thickness or integrated materials such as foam cores.” He points out that this project has already generated a lot of knowledge and sees potential in the future to apply it to slower RTM and infusion processes, where the risk of part quality issues has historically been very high. Faber will also present progress and results from the CosiMo project on Wednesday at “Sensor-based In-situ Polymerization Process Monitoring in Caprolactam T-RTM Production.”

 

DLR ZLP is also one of the three key partners in the Augsburg AI Production Network, which was founded in January 2021 and also brings together the University of Augsburg and the Fraunhofer Institute for Casting, Composites and Processing Technology (Fraunhofer IGCV).

 

As Dr. Markus Sause, Director of the AI ​​Production Network and a researcher developing the ultrasonic sensors in the CosiMo project, explains, “We will use CosiMo The blueprint for the collaboration highlighted in the project is expanded to a larger scale, developing AI technologies for production, with a focus on composite materials.

 

Our new 5,000 m2 plant in Augsburg will open in 2022 and be equipped with various machines throughout the next year, allowing the company to see real processes in a production environment with AI demonstrations. ”

 

At the same time, Faurecia Clean Mobility will internally expand the knowledge gained through the CosiMo project, which is fully aligned with Faurecia’s sustainability strategy and ecological transition. Faurecia also thanks the Bavarian Federal Ministry for Economic Affairs, Regional Development and Energy for its financial support for the research project.

 

 

Note: This article is compiled based on “CosiMo: Smart thermoplastic RTM process demonstrated for battery box cover challenges imulator” 2021.9.24 and other online materials.