Influence of process parameters on fiber orientation in reinforced thermoplastics

24 November 2016
Teresa Möbius, Rico Blei, Niels Modler, Camilo Pérez, and Tim A. Osswald
Simulations and experiments are conducted to determine how the fiber orientation of composite parts changes during a hot-press manufacturing process.

Textile-reinforced thermoplastics continue to find an increasing number of applications because of their highly productive manufacturing and good recyclability.1 Examples of current applications include seat panels, bike frames, and car front-ends that are made out of fiber-reinforced thermoplastic materials.2–4 To capitalize on the benefits of thermoplastic composites, a fundamental understanding of their material characteristics is required. In particular, it is necessary to find a way to achieve a reproducible fiber orientation after processing because changes in the fiber orientation of the composites can occur. This is especially the case during hot pressing processes when using tools that are open at the sides (e.g. during the manufacturing of semi-finished parts as organic sheets, and their later processing). These tool systems facilitate matrix flow to the outer areas of the cavities. In addition, changes in fiber orientation—and thus the mechanical properties of the resultant composite part—can arise.

During continuous development of composite parts along the whole process chain, virtual simulation of single process steps (e.g., impregnation or draping) is essential. The analysis and simulation of these process steps—together with the obtained results—are beneficial for predicting the mechanical properties of composite parts. These analyses can also be used for optimization of tool design, process parameter configurations, and to ensure a reproducible component quality.

In our work,5 we have thus analyzed the influence of selected process parameters on changes in the fiber orientation of unidirectional reinforced thermoplastics (made from a glass fiber/polypropylene hybrid yarn) during an isothermal hot pressing process. In detail, we used a mechanistic model—developed at the Polymer Engineering Center at the University of Wisconsin-Madison—to model the behavior of glass fibers. In this model, a fiber is represented as a chain of cylindrical, rigid, capped elements that are connected by nodes (see Figure 1). This allows us to set the fiber characteristics (e.g., length, diameter, and stiffness) to model various types of fiber behavior.6 By adjusting the number of fibers and nodes in the model, we can simultaneously obtain an acceptable simulation time and adequate-quality results.

Schematic illustration of a single fiber within the mechanistic model. The fiber is represented by cylindrical, rigid capped elements that are connected by a number (i) of nodes.7

By using the Hele-Shaw model (and hence the assumption of the matrix as a Newtonian fluid), we obtained an analytical solution for describing the flow field within the charge. We were thus able to adequately model the pressure distribution and velocity field (see Figure 2), as well as the vorticity fields. To make a comprehensive analysis of our simulation results, we created several MATLAB-based codes. These codes allow us to calculate the changes in fiber orientation by measuring the fiber angle after any desired time step, and at each position in the observed area (see Figure 3).

(a) Pressure distribution and (b) velocity field in the x–y plane, calculated using the Hele-Shaw model.

Schematic illustration of the simulation process and of measuring the fiber angle (α) along an analysis line.

We analyzed the effects of selected process parameters (e.g., applied pressure and temperature of the matrix system) on changes of fiber orientation. The pressure and temperature conditions we examined are given in Table 1 (where the temperature of the polypropylene is represented by its viscosity values, which were measured in advance). Overall, the results of these investigations show that higher vorticity occurs near the edges of the mold than in the center area and the larger changes in fiber orientation are thus more obvious in these outer areas (see Figure 3).

Analyzed process parameter values. The matrix temperature of the polypropylene is represented by its viscosity, which was previously measured.

Pressure (MPa)
Temperature (°C)/Viscosity (Pa·s)180/456210/219240/106

We also examined the dependency of process temperature/matrix viscosity on the fiber orientation, as well as the influence of process pressure. We find—see Figure 4(a)—that lower process temperatures (i.e., higher matrix viscosities) result in smaller changes of fiber orientation over the entire test plate. Furthermore, higher applied pressures—see Figure 4(b)—give rise to greater changes to fiber orientation at the edges of the plate. Moreover, differences between each of the pressure steps are not equal.

Influence of (a) matrix viscosity (at a pressure of 1.0MPa) and (b) applied pressure (at a viscosity of 219Pa·s) on fiber orientation (i.e., α), over the width of the plate. All measurements are for a y-axis position of 65mm.

In addition to our simulations of fiber orientation changes, we performed laboratory experiments to validate the simulation results. For these experiments, we used different process parameter combinations to manufacture various test samples. We subsequently used x-ray computed tomography to determine the fiber orientation of these samples: see Figure 5(a). Our experimental results confirm the results we obtained from our simulations, as shown in Figure5(b). Any differences can be explained from the assumptions we made in the simulations.

(a) X-ray image of a test sample (left) and the corresponding simulation result (right), for a temperature 240°C and applied pressure 1.0MPa. (b) Comparison of experimental (solid lines) and simulation (dashed lines) fiber orientation results, over the width of the plate.

In summary, we have conducted simulations and laboratory experiments to investigate the influence of selected process parameters (pressure and temperature/viscosity of the matrix) on fiber orientation within reinforced thermoplastics, during hot pressing processes. The results of our investigations can be used to derive recommendations for process setting during the manufacturing of unidirectional fiber-reinforced thermoplastics (with a focus on changes to fiber orientation). Within the ranges we analyzed, we find that higher viscosity and lower applied pressure cause a smaller change in fiber orientation. Nevertheless, with regard to the mechanical properties of composite parts, the impregnation quality (in addition to the importance of fiber orientation) is important. High impregnation qualities, however, are often achieved with low viscosities and high applied pressures. It is thus necessary to carefully consider these contradictory process parameter requirements. In our future work we need to conduct additional detailed investigations in which we analyze the parameter combinations with respect to the shear-rate-dependent viscosity of the thermoplastic matrix. We also plan to examine changes within the type of fiber reinforcement (e.g., unidirectional and bidirectional textiles).


Teresa Möbius
Institute of Lightweight Engineering and Polymer Technology, Technische Universität Dresden (TUD)

Teresa Möbius graduated with a degree in industrial engineering in 2011 from TUD. Since 2011 she has been working as a research scientist at the Institute of Lightweight Engineering and Polymer Technology. Her areas of research are technology development and integration of electronic elements.

Rico Blei
Institute of Lightweight Engineering and Polymer Technology, Technische Universität Dresden (TUD)

Rico Blei earned his diploma in mechanical engineering in 2015 from TUD. He currently works as a research scientist and focuses on lightweight constructions, as well as the area of bionics.

Niels Modler
Institute of Lightweight Engineering and Polymer Technology, Technische Universität Dresden (TUD)

Niels Modler is a professor, with the Chair of Function-Integrative Lightweight Engineering. He is also the general manager of the Collaborative Research Centre 639, the spokesman for Dresden within the Collaborative Research Centre/Transregio 39, and a member of VDI-FA310.

Camilo Pérez
Polymer Engineering Center, University of Wisconsin-Madison

Camilo Pérez graduated with a degree in chemical engineering in 2012 from the National University of Colombia, and obtained his PhD from the University of Wisconsin-Madison in 2016 under the direction of Tim Osswald. His area of research has been in the area of polymer composites processing, modeling, and simulation.

Tim A. Osswald
Polymer Engineering Center, University of Wisconsin-Madison

Tim Osswald is a professor in the area of polymer and composites engineering in the Mechanical Engineering Department. He holds honorary professor positions at the University of Erlangen-Nuremberg (Germany) and the National University of Colombia. He has authored 12 books and 300 papers in the field of polymer engineering.


  1. C. Cherif, Textile Werkstoffe für den Leichtbau, Springer, 2011.

  2. W. Hufenbach, N. Modler, M. Krahl, A. Hornig, H. Ferkel, H. Kurz and M. Ehlehen, Integrales Bauweisenkonzept: Leichtbausitzschalen im Serientakt, Kunststoffe 100, pp. 56-59, 2010.

  3. W. Hufenbach, A. Langkamp, F. Adam, M. Krahl, A. Hornig, M. Zscheyge and K.-H. Modler, An integral design and manufacturing concept for crash resistant textile and long-fiber reinforced polypropylene structural components, Proc. Eng. 10, pp. 2086-2091, 2011.

  4. M. Gude, M. Krahl, C. Garthaus and M. Stegelmann, Faser-Thermoplast-Verbund: neue Möglichkeiten zur Entwicklung von Leichtbauprodukten, Entwickeln Entwerfen Erleben: Beiträge zur virtuellen Produktentwicklung und Konstruktionstechnik, pp. 357-368, TUDpress, 2014.

  5. R. Blei, T. Möbius, C. Pérez, T. A. Osswald and N. Modler, Influence of selected process parameters on changes of the fiber orientation in unidirectional reinforced thermoplastics during a hot pressing process, Polym. Compos., 2016.

  6. C. F. Schmid, Simulation of Flocculation in Flowing Fiber Suspensions, PhD thesis, University of Wisconsin-Madison, 1999.

  7. D. Ramirez, Study of Fiber Motion in Molding Processes by Means of a Mechanistic Model, PhD thesis, Univerity of Wisconsin-Madison, 2014.

DOI:  10.2417/spepro.006772