IKV: Research for Practical ApplicationShort-Fiber-Reinforced Thermoplastics: Targeted Analysis of Fatigue Behavior and Improvement of Service Life
A guest post by
Prof. Dr.-Ing. Christian Hopmann, Director of the IKV Institute and Chair of Plastics Processing; Roman Schmohl, M.Sc., Short-Fiber-Reinforced Thermoplastics; Jonas Regel, M.Sc., Partially Cross-Linked Epoxy Resin Systems, and Marie Hadenfeldt, M.Sc., Extrusion
| Translated by AI
6 min Reading Time
Short-fiber-reinforced thermoplastics are increasingly being used in high-stress applications—such as in automotive and equipment manufacturing. However, the inhomogeneous fiber orientation makes it difficult to reliably predict fatigue behavior. Researchers at the IKV demonstrate how targeted specimen fabrication, experimental analyses, and micromechanical simulations can lead to a better understanding of damage mechanisms and the development of more precise service life models.
However, a method developed at the IKV makes it possible to produce test specimens with homogeneous, highly oriented fibers, thereby allowing the damage mechanisms to be studied without any interactions.
(Bild: Toray)
Understanding the Fatigue Behavior of Short-Fiber-Reinforced Thermoplastics
Short-fiber-reinforced thermoplastics are increasingly being used in applications subject to fatigue loads, such as in automotive and equipment manufacturing. These materials are manufactured using injection molding, a process that results in a process-dependent distribution of fiber orientation across the wall thickness. This inhomogeneous anisotropy complicates damage analysis, as layer interactions cause a dependence on the fiber distribution. However, a method developed at the IKV makes it possible to produce test specimens with homogeneous, highly oriented fibers, thereby allowing for the investigation of damage mechanisms free from such interactions.
To investigate the fatigue behavior of short-glass-fiber-reinforced thermoplastics, experimental characterization, mechanical testing, and micromechanical simulation were combined in such a way that microstructure, stress, and model parameters for material modeling could be consistently mapped to one another. The investigations were conducted under tensile threshold loading, such that the material response is influenced not only by cyclic damage accumulation but also by time-dependent viscoelastic or viscoplastic effects, thereby representing a practical load case for structural applications. Creep during the loading phases thus superimposes on the cyclic damage and influences the micromechanical stress and strain distribution, and consequently also the damage development through crack initiation and growth. Test specimens based on polybutylene terephthalate (PBT) with 10, 20, and 30 wt% glass fibers were investigated.
A key methodological component was the use of test specimens with uniform fiber orientation to avoid the superposition of multiple orientation states across the thickness that is typical of conventionally injection-molded components. The microstructure was imaged using µCT and quantitatively described using fiber orientation tensors. This made it possible to conduct material tests at the mesoscale on test specimens with a defined orientation state, without interactions between different orientations masking the damage mechanisms.
Fiber ends were identified as the preferred initiation sites for damage, which is consistent with local stress concentrations and the clustering of microdamage in their vicinity. In contrast, fiber crossings that do not involve fiber ends do not consistently play a critical role, so their significance depends more on the local configuration. A qualitative comparison with the micromechanical simulations supports this interpretation. The phase-field fracture approach reproduces the location of typical damage localization as well as characteristic crack paths in the matrix, thereby enabling a mechanically grounded classification of the observed damage patterns.
This allows for a more precise identification of the key mechanisms involved in damage initiation and subsequent damage progression under tensile cyclic loading. The results provide a starting point for systematically quantifying the influence of defined fiber orientations and microstructural features in future studies and for incorporating them into microstructure-based service life models.
This work was supported by the German Research Foundation (DFG) as part of project HO 4776/82-1.
Thermoforming of Partially Cross-Linked Epoxy Resin Systems
The plastic formability of partially cross-linked thermosets opens up new possibilities for post-forming of fiber-reinforced composite components following a primary forming process. However, these possibilities have hardly been systematically exploited to date. In particular, there is no reliable process window that describes the permissible degree of deformation as a function of temperature, degree of cross-linking, and fiber content. In a new research project, the IKV therefore aims to identify the conditions under which partially cross-linked molding compounds can be hot-formed reproducibly and without structural damage, and to derive criteria for their industrial application. To this end, a process chain consisting of initial shaping, storage, the forming process, and a final cross-linking phase will be incorporated.
Date: 08.12.2025
Naturally, we always handle your personal data responsibly. Any personal data we receive from you is processed in accordance with applicable data protection legislation. For detailed information please see our privacy policy.
Consent to the use of data for promotional purposes
I hereby consent to Vogel Communications Group GmbH & Co. KG, Max-Planck-Str. 7-9, 97082 Würzburg including any affiliated companies according to §§ 15 et seq. AktG (hereafter: Vogel Communications Group) using my e-mail address to send editorial newsletters. A list of all affiliated companies can be found here
Newsletter content may include all products and services of any companies mentioned above, including for example specialist journals and books, events and fairs as well as event-related products and services, print and digital media offers and services such as additional (editorial) newsletters, raffles, lead campaigns, market research both online and offline, specialist webportals and e-learning offers. In case my personal telephone number has also been collected, it may be used for offers of aforementioned products, for services of the companies mentioned above, and market research purposes.
Additionally, my consent also includes the processing of my email address and telephone number for data matching for marketing purposes with select advertising partners such as LinkedIn, Google, and Meta. For this, Vogel Communications Group may transmit said data in hashed form to the advertising partners who then use said data to determine whether I am also a member of the mentioned advertising partner portals. Vogel Communications Group uses this feature for the purposes of re-targeting (up-selling, cross-selling, and customer loyalty), generating so-called look-alike audiences for acquisition of new customers, and as basis for exclusion for on-going advertising campaigns. Further information can be found in section “data matching for marketing purposes”.
In case I access protected data on Internet portals of Vogel Communications Group including any affiliated companies according to §§ 15 et seq. AktG, I need to provide further data in order to register for the access to such content. In return for this free access to editorial content, my data may be used in accordance with this consent for the purposes stated here. This does not apply to data matching for marketing purposes.
Right of revocation
I understand that I can revoke my consent at will. My revocation does not change the lawfulness of data processing that was conducted based on my consent leading up to my revocation. One option to declare my revocation is to use the contact form found at https://contact.vogel.de. In case I no longer wish to receive certain newsletters, I have subscribed to, I can also click on the unsubscribe link included at the end of a newsletter. Further information regarding my right of revocation and the implementation of it as well as the consequences of my revocation can be found in the data protection declaration, section editorial newsletter.
Figure 1: The results of material characterization of semi-finished products with defined degrees of cross-linking form the basis for material models, from which suitable forming windows can then be derived.
(Bild: IKV)
Semi-finished products produced using low-pressure processes and pultrusion are manufactured with defined degrees of cross-linking and characterized in terms of their thermomechanical behavior. Key parameters such as ductility, stress relaxation, and dimensional stability following semi-finished product manufacturing are determined experimentally. In addition, dynamic-mechanical analyses and structural mechanical testing methods are used to quantitatively characterize the viscoelastic-plastic behavior. The experimental results form the basis for material models designed to realistically describe the transition from the elastic to the plastic state of deformation (Figure 1). Through continuous comparison between the model and experimental data, a reliable parameter space is derived that enables the identification of a suitable forming window. At the same time, methodological approaches are being developed to determine the relevant degree of cross-linking during semi-finished product manufacturing.
This work is supported by the Federal Ministry for Economic Affairs and Energy (BMWE) as part of IGF Project 24012 N.
Numerical Simulations Using Neural Networks for Metamodeling in Extrusion
During the extrusion of long-fiber-reinforced thermosets, the flow of the material alters the original fiber orientation in the semi-finished product. To specifically adapt a component to its future load, both the placement position and the geometry of the semi-finished product must be selected in such a way that a load-appropriate fiber orientation distribution is achieved. Such a design using numerical simulations is extremely computationally and time-intensive due to the high complexity of the material and process behavior. The IKV has therefore investigated how artificial feedforward neural networks can be used as metamodels to develop efficient predictive models.
Figure 2: Overview of the approaches examined for a metamodel to describe extrusion and a summary of the main findings.
(Bild: IKV)
The research focused on fundamental questions regarding optimal model architecture, the selection and tuning of hyperparameters, the appropriate parameterization of input and output variables, and the assessment of the required data volumes. In addition, various strategies for generating training data, for systematic experimental design, and for dividing the data into training, validation, and test datasets were investigated (Figure 2). Another focus was the artificial expansion of the dataset to make the model more robust and improve its predictive performance. A comparison was made between a single-stage approach, in which the network structure directly maps the correlation between the insertion position and the resulting component deflection, and a two-stage model that additionally accounts for the resulting flow fronts, thereby embedding the physical mechanisms more deeply within the model.
For the strategy on systematic experimental design, the two most promising approaches for creating experimental designs were selected from a literature review and compared: An orthogonal Taguchi array and Latin Hypercube Sampling (LHS) as a space-filling design.
For both experimental designs, a comprehensive dataset of similar scope was generated through numerical extrusion simulations based on a parameterization of the insertion position and geometry (star-convex) using ten variables. The investigations showed that models trained on a Taguchi dataset initially achieve a very high model quality. However, if the same model is validated with data outside the experimental design, the model quality drops drastically, suggesting strong overfitting to the specific structure of the Taguchi design as well as insufficient coverage of the parameter space. Overall, these results illustrate that the choice of experimental design has a decisive influence on the generalizability and practical applicability of the metamodel.
The methodology developed in this project exemplifies the potential of artificial intelligence for plastics engineering, while also highlighting the challenges that arise during its implementation. The resulting metamodel can be applied independently of material and component geometry. It can be reliably used for optimization within the parameter ranges covered by the training dataset, as the model predictions correspond well with the target values derived from numerical simulation. However, this does not apply to the model accuracy outside this initial parameter range, which remains the subject of future research.
This work was supported by the German Research Foundation (DFG) as part of project HO 4776/74-1.