Process Modeling of Inductive Heating of CFRP

In addition to resistance and ultrasonic welding, induction welding is considered one of the most relevant methods for a thermal joining of thermoplastic carbon fiber-reinforced polymers (CFRP). Through contactless, intrinsic heating combined with high heating rates, complex shaped and large-scale structures can be efficiently joined. One challenge in achieving consistent joint quality with continuous induction welding is managing heat distribution, which depends significantly on various electromagnetic effects and numerous laminate parameters.

During inductive heating of CFRP, electrical eddy currents are induced into an electrically conductive CFRP laminate by an alternating electromagnetic field. Three heating mechanisms can be distinguished in the inductive heating of tape-based CFRP: Within the laminate layers, joule losses occur in the electrically conductive carbon fibers. When the electric current changes the laminate layer, contact losses between crossing fibers and dielectric losses in the matrix polymer lead to an interlaminar heating.

The complex heating behavior and the resulting temperature distribution in the joining zone are difficult to capture adequately with experimental approaches. A comprehensive experimental optimization of the continuous induction welding process is thus only possible with a multitude of laborious experiments. Numerical process simulations developed at Leibniz-Institut für Verbundwerkstoffe GmbH (IVW) can help to minimize the number of required experiments and provide new insights into the process.

A challenge in modeling the inductive heating behavior of CFRP is the choice of the degree of homogenization of the simulation model (Figure 1). Previous models at IVW can be classified into macro- and meso-scale models.

Macro-scale models, where the laminate is fully homogenized, are not capable of distinguishing between the different heating mechanisms and often cannot accurately simulate the material-specific heating behavior. However, due to their low complexity, they can be useful for a preliminary process design.

Meso-scale models, where individual laminate layers are discretely modeled with orthotropic material models, can capture the layup-specific heating pattern. However, distinguishing between the different heating mechanisms is still only partially possible at this scale.

One approach to solving this challenge is the discrete modeling of contact resistance between adjacent laminate layers. Figure 2 shows that fiber losses in the layers and contact losses in the interlayer can thereby be discretely calculated and analyzed. This approach allows the direct use of electrically measured properties of individual layers and full-scale laminate.

In the next steps, the electrical conductivities of individual tapes and the contact resistances between two laminate layers, depending on the angle difference and consolidation pressure, will be measured. The goal of this work is to predict the heating behavior of any laminate constructions dependent on basic process parameters without the currently necessary experimental calibration of the electromagnetic material model.