Next Article in Journal
On the Relationship between Lightning Strike Parameters and Measured Free Surface Velocities in Artificial Lightning Strike Tests on Composite Panels
Previous Article in Journal
Physico-Mechanical Property Evaluation and Morphology Study of Moisture-Treated Hemp–Banana Natural-Fiber-Reinforced Green Composites
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Conceptions and Feasibility Study of Fiber Orientation in the Melt as Part of a Completely Circular Recycling Concept for Fiber-Reinforced Thermoplastics

1
Kunststofftechnik Paderborn (KTP), Paderborn University, 33098 Paderborn, Germany
2
Institut für Textiltechnik (ITA), RWTH Aachen University, 52074 Aachen, Germany
*
Author to whom correspondence should be addressed.
J. Compos. Sci. 2023, 7(7), 267; https://doi.org/10.3390/jcs7070267
Submission received: 9 May 2023 / Revised: 16 June 2023 / Accepted: 21 June 2023 / Published: 25 June 2023
(This article belongs to the Section Fiber Composites)

Abstract

:
Fiber-reinforced thermoplastics are an important construction material for lightweight applications. The increasing use of especially glass fiber-reinforced plastics leads to growing amounts of not recyclable composite materials, which is commonly disposed of by landfilling. Hence, there is a need for a recycling concept for glass-fiber-reinforced plastics that enables their complete reuse over many recycling cycles. In this paper, such a recycling concept is presented, which is based on the idea of melting the whole glass-fiber-reinforced component without prior size reduction. The fiber-reinforced melt will be pressed through a nozzle in order to achieve a strand with highly oriented fibers that can then be applied in new components via a tape-like laying process. The feasibility of the recycling concept is proved in this paper. Therefore, investigations on the reorientation of fibers in the melt by pressing through a rectangular nozzle have been carried out with different nozzle diameters, shear rates and melt temperatures. The investigations result in a stable process, which enables an increase in fiber orientation of about 37% up to a mean fiber orientation of 67% in the flow direction. These findings are independent of the initial fiber orientation.

1. Introduction

1.1. Motivation and Research Aim

Fiber-reinforced plastics (FRPs) are increasingly used in high-tech applications such as aviation [1] or the automotive industry [2,3,4] and in the field of wind turbines [5]. In lightweight construction, they can reduce resource consumption, in particular due to their very good weight-specific, mechanical properties [6]. Thus, in Europe, increasing sales figures for FRPs can be recorded for years [7,8]. Furthermore, the German government mentions lightweight construction as a key technology that can reduce greenhouse gas emissions and thus contribute to the Climate Action Plan 2050 [9]. In view of the increasing global demand for raw materials [10], fiber-reinforced plastics can only be used sustainably if they are recycled at the end of the product life cycle. The recyclates need to be further used in new high-quality products, so that the finite resources are in a closed-loop material cycle.
At present, however, FRPs are mostly landfilled, if permitted, thermally recycled or, in the best case, downcycled into short-fiber-reinforced recyclates. These processes do not provide an opportunity for a circular economy, as the resources are not reused. Even though there are better recycling options for fiber-reinforced thermoplastics, high-quality FRP recycling is still not state-of-the-art, and thermal recovery of landfilling prevails. The goal must be complete recycling of both the matrix polymer and the fibers without any property losses. In order to utilize the properties of the reinforcing fibers in the new products, the preservation of the fiber length as well as the re-alignment of the fibers along the load paths in the component is of particular importance. This paper will therefore present a concept and demonstrate its general feasibility, which enables a closed-loop solution for recycling of fiber-reinforced thermoplastics by melting the fiber-matrix composite and re-aligning the fibers.

1.2. State of the Art

1.2.1. Effect of Fiber Length and Alignment

Strength and stiffness of fiber-plastic composites are determined in particular by the properties of the fibers, as long as they provide sufficient fiber length. In order to fully exploit the properties of the fiber, the forces acting on the composite must be transmitted from the matrix to the fiber. Due to the different Young’s moduli of the fiber and the matrix material, there is an increasing displacement between fiber and matrix towards the fiber ends, which causes increasing shear stresses. The shear stresses at the fiber-matrix interface also trigger stresses in the fiber [11]. In order to be able to use the fiber properties up to fiber breakage, there must be good bonding strength on the one hand, and the fiber length must be longer than the critical fiber length l c r i t   on the other. If the fiber length is > l c r i t , fiber breakage occurs; for fiber lengths < l c r i t , fiber pull-out takes place. The critical minimum fiber length l c r i t   is estimated with the aid of the fiber diameter d f ,   the apparent interfacial shear strength τ b and the fiber breaking stress σ f B as follows: [11]
l c r i t = σ f B d f τ b 2
where τ b is essentially affected by the applied glass fiber size, which promotes adhesion between fiber and matrix polymer. For subcritical fiber length, a rapid increase in composite strength is achieved by increasing the length. For supercritical fiber lengths, an asymptotic progression against the strength of a continuous fiber-reinforced plastic is achieved with an increase in fiber length [11]. Therefore, the fiber length represents an essential parameter for the mechanical properties of fiber-reinforced composites.
In addition, the fiber direction essentially determines the mechanical properties of the composite, because the load can just be transferred in fiber direction. The force absorption perpendicular to the fiber is low because the matrix is weakened by the increase in strain and the detachment of the fiber from the matrix. In the fiber-parallel direction, the fibers can still absorb large forces [11]. Therefore, the fiber length and fiber orientation in the load direction must be completely preserved on a high level in new component after several recycling process.
This demonstrates the need for a completely circular recycling solution for FRPs that prevents downcycling of the properties. In the following, the fundamentals of fiber orientation in the melt will be discussed, as these will be the fundamentals for the novel recycling concept (see Section 1.3).

1.2.2. Mathematically Description of Fiber Orientation

In order to be able to describe the fiber orientation in the melt mathematically, the orientation of a single fiber in three-dimensional space can be expressed by the orientation vector p (Figure 1), which is aligned along the fiber axis. The orientation vector results from the existing angular deviations with respect to the three main Cartesian coordinate axes [12].
p x = sin θ cos ϕ p y = sin θ sin ϕ p z = cos θ
While the use of the vector p works for the description of the orientation of single fibers, it is not able to describe the orientation of fiber clusters. To do so, the orientation is described by the probability density function using a second degree tensor according to the following [12]:
A i j = p i p j = A x x A x y A x z A y x A y y A y z A z x A z y A z z = p x p x p x p y p x p z p y p x p y p y 2 p y p z p z p x p z p y p z p z = sin 2 θ cos 2 ϕ sin 2 θ cos ϕ   sin ( ϕ ) cos θ sin θ   cos ( ϕ ) sin 2 θ cos ϕ   sin ( ϕ ) sin 2 θ sin 2 ϕ cos θ sin θ   sin ( ϕ ) cos θ sin θ   cos ( ϕ ) cos θ sin θ   sin ( ϕ ) cos 2 θ
The orientation tensor is symmetric, and the entries on the main diagonal sum to one. A x x ,     A y y   and A z z   represent the fiber orientations along the major axes, while the elements on the minor diagonal reflect the deviation of the major axis from the geometric axis of the fiber [12].

1.2.3. Behavior of Fibers in Plastic Melt

Plastic melt in a flow channel can be considered as laminar sheet flow. Assuming wall adhesion, the velocity at the channel wall is zero. The maximum flow velocity is formed in the center of the channel. This results in a velocity gradient in the flow which leads to a shear of the volume elements. The shear rate is the greatest close to the channel wall, since the greatest velocity gradient is present here. The resulting courses of the viscosity and the velocity profile as well as the shear rate are shown in Figure 2. The shear rate gradients lead to an alignment of glass fibers in the melt, which is the strongest in the boundary regions of the channel, forming the so called boundary layers with highly aligned fibers, and decreases towards the center of the channel, where the core layer with less aligned fibers is formed [13]. The thickness of the core and the boundary layers dependent of various parameters such as the structural viscosity of the material, the temperature of the melt, its viscosity, the thickness of the channel, the shear rate or the existence of fillers. The effect of the mentioned parameters on the fiber orientation in the melt and the thickness of the boundary and core layers will be described in the Section 1.2.4

1.2.4. Effect of Material and Process Parameter on Fiber Orientation

The degree of orientation of the fibers along the flow will depend on the fiber concentration in the melt, the fiber length, the viscosity of the melt as well as on the flow velocity, i.e., the shear rate acting on the fiber and therefore also on the geometry of the flow channel.
In the resting state, the macromolecules of the polymer are in a state of maximum entropy and are disordered and entangled. In a melt flow, the macromolecules have to slide past one another and require therefore a lot of energy at first. Through flow processes, the liquid layers move relative to each other, and shear stresses are transferred between the layers.
The shear stresses cause the molecular chains to align and stretch in the direction of flow. The more the molecular chains are stretched, the less energy is required to make them slide past each other. For this reason, the viscosity of melt decreases as the shear rate increases. The course of the shear stress is degressive for plastic melts, as can be seen in Figure 1 [14]. Another parameter influencing the viscosity of plastic melts is the temperature. An increase in temperature makes more energy available to the melt, the volume increases, and the chains are more mobile, resulting in a decrease in viscosity [13,14]. In addition, fillers significantly interact with the viscosity of plastic melts as they influence each other. Thus, not only viscosity influences the fiber orientation in the melt, but at the same time, fiber orientation also influences viscosity, with viscosity decreasing with increasing degree of fiber orientation in the flow direction [15]. Fibers in a melt are considered as rigid solids, which lead to a reduction in the channel cross-section and thus to rising velocity. Hence, the shear rate also increases with increasing fiber content (see Figure 3) [14].
Glass fibers strongly influence the flow behavior of plastic melts, as the viscosity of filled melts is increased by a factor k, which is a function of the volume fraction of the filler [16]. The higher viscosity of the filled melts results from the fact that the solid particles are entrained. However, shearing of the particles does not lead to a reduction in viscosity as in the structurally viscous plastic melts [11]. The influence of fiber content on viscosity decreases with increasing shear rate and is most pronounced at low shear rates [16]. At the same time, high fiber concentration in the melt leads to inhibition of fiber orientation due to interaction with adjacent fibers [16]. The aspect ratio of fiber length to diameter also affects the viscosity, with a higher aspect ratio leading to higher viscosity. Thus, fiber length must also be considered when modeling fiber orientation [12]. According to Kim and Song [15], the probability of fiber orientation in a low viscosity melt is higher than in a high viscosity melt because fiber motion is associated with higher stress.
To promote the process of fiber orientation, the glass fiber bundles have to dissolve in the polymer melt. This is commonly achieved by using a film former, which must possess good compatibility with the matrix polymer, in terms of wetting out and solubility [17]. For PA6-compatible glass fiber sizes, polyurethane based film formers have become established in the glass fiber industry, in conjunction with APTES utilized as adhesion promoter. For a long time, this simple combination of the two components have been used by glass fiber manufactures [18]. Other components and additives can also be present in the particular glass fiber size but often kept secret [19].
Another parameter influencing the melt viscosity as well as the shear stress and thus affecting the fiber orientation is the channel geometry. Due to the velocity profile in a channel, the formation of the boundary layer and the core layer occurs, with differences in fiber orientation. In the boundary layer, the fiber orientation in the flow direction is high due to the high shear rates. In the core layer, the degree of fiber orientation is lower. To maximize the fiber orientation, it is necessary to minimize the core layer width.
Optimizing the channel geometry, especially the thickness, is an effective way of influencing the alignment of the fibers and the formation of the boundary layer. In the case of thin-walled components, only a thin core layer is formed because the higher surface-to-volume ratio. The contact of the plastic melt with the channel wall is higher, and the boundary layer is dominant [12,20]. To achieve the highest possible fiber orientation, molds in the injection molding process are designed to let the melt pass through a thin undercut in which the fibers are oriented before they then flow into the actual mold [12,20].
In nozzles, in addition to shear flows, extensional flows also occur, which lead to pressure losses, in particular due to the dissipating secondary vortices at the inlet to the nozzle, which must be considered in the modelling [21]. The velocity increases cause the fibers to become increasingly oriented (Figure 4). Furthermore, the macromolecules are elastically sheared in the nozzle, whereby they just partly relax in the nozzle, and they will relax, contract, and the melt will expand as they exit the nozzle. The consequence is that die swell of the extrudate can occur, which can influence the fiber orientation [18].

1.2.5. Current State of Discontinuous Fiber Alignment

In the literature, several methods to align discontinuous fibers are described which use the fundamental principles of fiber orientation in the melt that have been described in the Section 1.2.4.
Such et al. [22] provide a historic review of different alignment methods, and they divide the methods into five categories: fluid-based, pneumatic, based on an electric field, ultrasonic or acoustic vibration. Pneumatic alignment processes are based on spraying chopped fibers through a nozzle. The resin can be added afterwards or simultaneously with the fibers. The problems are the limited throughput and the equal dispersion of the chopped carbon fibers [22]. Harper et al. [23] describe one of these processes in which they spray and align chopped carbon fibers through a nozzle with powdered binder to achieve a preform. Hot air consolidates the binder. They found good mechanical properties for highly aligned specimens. Neither acoustic vibration nor electric field orientation achieved the desired level of orientation [22]. Furthermore, electric field orientation just works for electrically conductive fiber; thus, it is not suitable for glass fibers. A dry alignment process is described by Miyake et al. [24]. In their investigations, they used fluffy synthetic thermoplastic fibers as a suspension medium and spread the carbon fibers on the synthetic fibers. A following drafting and combing process aligns the fibers as they follow the orientation of the synthetic fibers. After several process steps, they fabricate a yarn. With this process they could reach almost 70% fiber orientation in ±14° [24]. Most of the alignment processes mentioned in [22] utilize a fluid-based orientation. The fibers are then diluted into a suspension, often glycerin, and oriented through a nozzle by a converging flow. One recently described fluid-based process is called HiPerDiF (high-performance discontinuous fiber). The suspended fibers are led through a jet onto a perforated conveyor belt, and a vacuum helps to dry the aligned fibers before resin impregnation. An advantage of this principle is the upscaling option by using multiple nozzles [25]. Further research has been carried out [26,27] to improve this method for example by simulating the fiber orientation in the HiPerDiF process [28] or by applying the process to 3D-prinitng of composites [29].
One major drawback of the described fluid-based processes is the need of a drying process before fiber impregnation with the resin. Furthermore, most described processes are developed for short fibers in a thermoset resin, which offer just limited recycling possibilities. Therefore, there still is the need for a recycling concept of long fiber-reinforced thermoplastics that focuses on circularity.
In thermoplastic composites, the plastic melt itself can be used as fluid to align the fibers in flow direction. This will be used in the completely circular recycling concept that will be described in the Section 1.3. Subsequently, the feasibility of the concept is verified by melting long fiber-reinforced Polyamid-6 (PA6) in a reservoir and investigating the fiber alignment in a simple strand extrudate as a function of the parameters temperature, shear rate and nozzle geometry.

1.3. The Circular Recycling Concept

The holistic recycling concept is based on the idea that fiber-reinforced end-of-life components are melted in a reservoir without any prior size reduction. The reservoir already consists melt in which the new end-of-life components will be placed, so that direct heat transfer can take place. In a timed process, the melt is conveyed from this reservoir into the strand deposition unit (shown in Figure 5), from where the fiber-reinforced melt is pressed through a specially shaped, rectangular nozzle. The melt flow from the reservoir into the deposition unit is coordinated according to the pressing process and the piston movement. The geometry of the strand deposition unit as well as the nozzle have to be designed to guarantee maximum fiber orientation in the direction of flow. The strand can either be deposited directly on the new component in the load direction via a robot unit or first wound up and then deposited in load direction in a following process step. The deposited stands will be pressed to consolidate the semi-finished product. This work will focus on the fiber orientation in the first step. The fact that no screw machines are used to convey the fiber-reinforced melt means that fiber shortening is kept as low as possible. The fiber content of the new products will therefore equal the initial fiber content of the melted end-of-life composite parts. Future research on optimizing the process will concentrate once on the geometry of the deposition unit and the nozzle but also on the temperature and pressure control in the process in order to hinder any thermal degradation of the material. Another focus will be the degradation of the size during several lifecycles, since the state-of-the-art film formers based on polyurethane enable good composite performance for virgin fibers but undergo thermal degradation during compounding [30]. The overall concept from the deposition unit to the pressing process is shown in Figure 5.

2. Materials and Methods

In order to test the feasibility of this circular recycling concept, a high-capillary rheometer (RG50, Göttfert, Buchen, Germany) with a rectangular nozzle is used to form the fiber-reinforced strand. It offers the advantage that the shear rate as well as the temperature can be controlled during the entire process. The tests are carried out with PA6-LGF50 granules from BASF (Ultramid Structure B3WG10 LFX BK23215, BASF SE, Ludwigshafen, Germany). The high capillary rheometer has a capillary with a diameter of 15 mm in which the granules are melted. Via a piston, the fiber-reinforced plastic melt is pressed through the rectangular nozzle. The nozzle length is 100 mm, and the thickness is varied between 1 mm and 3 mm. In these investigations, the temperature, the shear rate and the thickness of the nozzle are varied (see Table 1). Due to limitations of the high-capillary rheometer, using the 3 mm nozzle the maximum achievable shear rate is 470 1/s instead of 1000 1/s. Furthermore, it is determined to what extent a prior mixing of the fiber bundle structure in the capillary influences the fiber orientation process in the nozzle. Therefore, the fiber bundle structures have been varied from the “normal” orientation in the capillary to “unstructured”. In this context, “normal” describes the orientation of the fiber bundles that results from filling the fiber-reinforced granules in the capillary and melting them, without any further disturbance. The “unstructured” fiber bundle orientation is achieved when the fiber-reinforced melt in the capillary has been manually stirred before it enters the nozzle. An overview of the varied process steps is given in Table 1.
Three representative pieces are taken from the strand, and their fiber orientation is examined by using the Nanotom S computed tomography (µCT) scanner from GE Inspection Technologies phoenix|X-rays. The µCT has a 180 kV X-ray tube with a detail detectability of 0.2–0.3 μm. The mean orientation tensor as well as the functional plot of the fiber orientation is analyzed using the VGSTUDIO MAX 2.2.2. software. The percentage of fiber alignment in flow (xx-) direction will be given as the xx-value of the mean orientation tensor, which is calculated according to Equations (2) and (3) by the software. The coordinate system is placed in the sample section for evaluation as shown in Figure 6.
Furthermore, sample sections are ashed in a muffle furnace (phoenix, CEM GmbH, Kamp-Lintfort, Germany) for 45 min at 550 °C so that the fiber structure becomes visible. The fiber lengths of the ashed samples are measured by suspending the fibers in water. The suspension is digitized with a scanner (EPSON Perfection V850 Pro, EPSON Deutschland GmbH, Meerbusch, Germany). The resulting images are analyzed with the software FiVer V1.80 (SKZ—KFE gGmbH, Würzburg, Germany), and the weighted average fiber length L w t is calculated. It is given by the single fiber length L f i and the number of fibers n f i :
L w t = n f i L f i 2 n f i L f i

3. Results and Discussion

3.1. Fiber Orientation along the Melt Flow

The fiber orientations in the sample have been evaluated along the y-axis (sample thickness) and are depicted in Figure 7. The different colors in the diagrams represent the process parameters temperature and shear rate. The lines show the value of A x x , A y y ,   A z z of the mean orientation tensor along the sample thickness. It can be seen that 60–70% of the fibers are oriented in the xx-direction along the flow, that 30–40% of the fibers are oriented according to the zz-direction, and that nearly no fibers are oriented in yy-direction. Due to the low nozzle thickness of 1 mm and the length of the fiber-reinforced granules of 10 mm, an alignment along the yy-direction is hindered. The different process parameters show no clear effect on the fiber alignment over the sample thickness as they just vary slightly. Therefore, it can be said that no definite core and boundary layers are formed during the pressing process through the 1 mm nozzle. This is advantageous, since the suppression of the core layer leads to a higher proportion of aligned fibers. Only for the first sample (Figure 7a), a tendency to form a core layer can be recognized for the samples pressed at 300 °C and 1000 1/s, due to the decrease in the proportion of aligned fibers in the xx-direction in the center of the sample. However, since this cannot be detected for the following two samples (Figure 7b,c), it can be assumed that the effects are negligible. As there are just little differences between the three samples, it can be stated that the fiber orientation is nearly constant along the melt strand, which is important for the future application.
In order to visualize the achieved fiber structure, the CT-scans as well as the ashed samples are depicted in Figure 8 for the investigations with the 1 mm nozzle. The high alignment of the fibers according to the melt flow direction is obvious and shown in the blue color in the CT-scans. Figure 9 depicts the measured pressure at the inlet and at the outlet of the 1 mm nozzle, and a nearly constant pressure can be observed during the extrusion period. The differences between the different samples represent the influence of the shear rate or the viscosity as a function of temperature. Slight fluctuations are obvious for all samples, which result from the high fiber content affecting the pressure sensors.
As already described before for the 1 mm nozzle, also for the 3 mm nozzle, the fiber distribution in the strand is equal, as there are no differences in Figure 10. The achieved fiber structure for the 3 mm nozzle can be seen in the CT-scans as well as the ashed samples in Figure 11. Again, no boundary and core layers are formed in the strand. Comparing the proportion of the oriented fibers between the 1 mm and the 3 mm nozzle, it can be seen that the thinner nozzle leads to a higher proportion of oriented fibers. This is due to the higher diameter reduction which leads to an increase in flow velocity and thus to a better fiber orientation [21]. This can be seen in Figure 12, which shows the mean xx-value of the orientation tensor through the whole sample, and the used colors clarify the process parameters. In this diagram, the values of the three samples are averaged, and differences are shown in the standard deviation. It is obvious that the melt temperature as well as the shear rate show nearly no effect on the proportion of xx-oriented fibers. Due to the viscosity reducing effect of a higher melt temperature [15], it has been expected that a higher temperature would lead to a higher proportion of oriented fibers. Moreover, an increasing shear rate was expected to increase the orientation of the fibers along the flow, as results from Foss et al. [31], Laun [13] and Azaiez et al. [21] show an increasing fiber orientation with rising shear rate.
The main reasons why these effects cannot be shown here are the unadapted nozzle geometry and the comparatively high fiber content. The capillary geometry, the passage between the capillary and the nozzle, as well as the nozzle geometry itself, should be designed to enhance fiber orientation and reduce fiber breakage. In these investigations, the nozzle geometry still does not have an adapted inlet, so that the fibers are deflected abruptly; thus, there is no continuous increase in flow velocity. In addition, the nozzle should be made longer, so that the time for orientation of the fibers is increased, and at the same time, the elastic portions of the stretching of the macromolecules relax, thus reducing die swelling. Die swelling can be considered another major factor reducing the fraction of oriented fibers, especially for the 3 mm nozzle, since the fibers change their orientation again as they exit the nozzle. The extrudate from the 1 mm nozzle expanded to a thickness of 3.18 mm, and the extrudate from the 3 mm nozzle expanded to a thickness of 11.7 mm. This high die swell supports the finding of a disturbing rearrangement of the fibers after they exit the nozzle.
Another significant difference compared to the above-mentioned studies from Foss et al. [24], Laun [13] and Azaiez et al. [21] is the high fiber content of 50 wt%. This high fiber content also leads to a strong hindrance to the freedom of movement of the individual fibers in the melt, so that the orientation is reduced. In addition, the high fiber length of several millimeter also causes steric hindrance of the fibers to align according to the flow direction.
The µCT-images in Figure 13 show the different fiber bundle structures in the capillary, representing the reference orientation of these investigations. The results shown above in Figure 7, Figure 8, Figure 9 and Figure 10 and in Figure 12 have been obtained for the normal fiber bundle structure in the capillary. Especially the µCT-images in Figure 8 and Figure 11 depict that throughout the orientation process in the nozzle, the fiber bundles dissolve, and the fibers orient in the flow direction. Figure 12 shows an increase in fiber orientation from the reference orientation in the capillary with 45% fiber orientation to the highly oriented fibers in the strand with up to 67% fiber orientation.
Furthermore, it has been investigated whether an unstructured fiber bundle orientation in the capillary hampers the fiber orientation process in the nozzles. Therefore, the fiber bundle structure has been manually disturbed in the capillary before pressing the fiber-reinforced melt through the 1 mm nozzle at 300 °C. The shear rate is altered between 100 and 1000 1/s. The results are shown in Figure 14 and depict again an increase in fiber orientation from the reference orientation with 47% fiber orientation to the samples pressed at 300 °C and 1000 1/s of up to 59% aligned fibers.
It can be concluded that a fiber alignment in xx-direction of up to 67% with 6.3% of the fibers deviate in ±5° has already been achieved with the unadjusted nozzle geometry, and that the process of fiber orientation in the nozzle is independent of the prior fiber orientation in the capillary.

3.2. Fiber Shortening

In order to evaluate if the high shear rate leads to a fiber breakage in the nozzle, the fiber length of the strands have been evaluated and compared to the fiber length in the fiber-reinforced granules. The results of the fiber length are shown in Figure 15. Compared to the initial fiber length in the granules, a fiber length reduction of 24% occurs for the 1 mm nozzle. For the 3 mm nozzle, a fiber length reduction of 37% could be measured, but for the investigations with the 3 mm nozzle and the high shear rate, no fiber shortening occurs. It can be seen that neither an increasing shear rate nor a change in temperature leads to any increase in fiber length shortening. The observed fiber length reduction will therefore occur mainly due to the redirection of the fibers in the melt towards the nozzle. A future improvement of the nozzle geometry will therefore help to further reduce the fiber shortening and at the same time enables an increase in fiber orientation. Nevertheless, a high fiber orientation and a reduction of the fiber breakage is an optimization problem as a high shear rate will increase the fiber orientation as well as increase the fiber breaking. Therefore, a compromise will be necessary to have a maximum of oriented fibers and at the same time a minimum of fiber breaking. This will be subject of future research.

4. Conclusions and Outlook

In this paper, we have discussed a new concept for a completely circular recycling process for fiber-reinforced thermoplastics, whose feasibility has been proven by the presented investigations. It was shown that even with an unadapted nozzle geometry, in concerns of fiber orientation and fiber breakage, up to 67% of the fibers could be aligned in the direction of flow, while at the same time, material damage could be largely avoided.
Future research work will focus on modelling the process in order to obtain a more precise understanding of the individual parameters and to develop an optimized nozzle geometry. In future recycling processes, glass fiber sizes will play a major role, since they are the source of adhesion between the glass fiber surface and the polymer, which gives FRPs their mechanical properties. Therefore, new glass fiber sizes and novel methods of adhesion promotion must be devised. A switchable glass-fiber–polymer interface, often referred to as debonding on demand, could enable an easier separation of the components of end-of-life glass-fiber-reinforced plastic. This would elude a costly resizing process for the reuse of recovered glass fibers for many lifecycles. At least, temperature-stable glass fiber sizes, which do not undergo thermal and mechanical degradation, are needed to avoid resizing of the glass fibers. Furthermore, investigations will be carried out with fibers of different lengths.
Overall, this innovative recycling concept keeps all used materials in circulation and will therefore enable a sustainable use of natural resources.

Author Contributions

Conceptualization, E.M., L.T., M.H. and C.G.; methodology, L.T.; data curation, L.T.; writing—original draft preparation, L.T. and M.H.; writing—review and editing, C.G. and E.M.; supervision, C.G. and E.M. All authors have read and agreed to the published version of the manuscript.

Funding

We acknowledge support for the publication costs by the Open Access Publication Fund of Paderborn University.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Meng, F. Environmental and Cost Analysis of Carbon Fibre Composites Recycling. Ph.D. Thesis, Universtity of Nottingham, Nottingham, UK, 2017. [Google Scholar]
  2. Seidel, S. Organoblech-Verschnitte recyceln: Faserverstärkte Thermoplaste mechanisch rezyklieren und in die Produktion zurückführen. Kunststoffe 2021, 1, 25–27. [Google Scholar]
  3. Harmina, T.; Hartikainen, J.; Lindner, M. Long Fiber-Reinforced Thermoplastic Composites in Automotive Applications. In Polymer Composites: From Nano- to Macro-Scale; Friedrich, K., Fakirov, S., Zhang, Z., Eds.; Springer: New York, NY, USA, 2005; pp. 255–262. [Google Scholar]
  4. Krauklis, A.E.; Karl, C.W.; Gagani, A.I.; Jørgensen, J.K. Composite Material Recycling Technology—State-of-the-Art and Sustainable Development for the 2020s. J. Compos. Sci. 2021, 5, 28. [Google Scholar] [CrossRef]
  5. Zotz, F.; Kling, M.M.; Langner, F.M.; Hohrath, P.; Born, H.; Feil, A. Entwicklung Eines Konzepts und Maßnahmen für Einen Ressourcensichernden Rückbau von Windenergieanlagen; Federal Environment Agency: Dessau-Roßlau, Germany, 2019.
  6. Ashby, M.F. Material Selection in Mechanical Design, 2nd ed.; Butterworth-Heinemann: Waltham, MA, USA, 1999. [Google Scholar]
  7. AVK Industrievereinigung Verstärkter Kunststoffe. Der Markt für Glasfaserverstärkte Kunststoffe 2019; AVK Industrievereinigung Verstärkter Kunststoffe: Frankfurt am Main, Germany, 2019. [Google Scholar]
  8. Witten, E.; Mathes, V. The Market for Glass Fibre Reinforced Plastics (GRP) in 2020; AVK Industrievereinigung Verstärkter Kunststoffe: Frankfurt am Main, Germany, 2020. [Google Scholar]
  9. Klimaschutzprogramm 2030 der Bundesregierung zur Umsetzung des Klimaschutzplans 2050, 2019. Available online: https://www.bundesfinanzministerium.de/Content/DE/Downloads/Klimaschutz/klimaschutzprogramm-2030-der-bundesregierung-zur-umsetzung-des-klimaschutzplans-2050.pdf?__blob=publicationFile&v=4 (accessed on 16 June 2023).
  10. OECD. Global Material Resources Outlook to 2060; OECD: Berlin, Germany, 2019. [Google Scholar]
  11. Ehrenstein, G.W. Faserverbund-Kunststoffe. Werkstoffe—Verarbeitung—Eigenschaften, 2nd ed.; Hanser: München, Germany, 2006. [Google Scholar]
  12. Tucker, C.L., III. Fundamentals of Fiber Orientation. Description, Measurement and Prediction; Hanser Publishers: Munich, Germany, 2022. [Google Scholar]
  13. Laun, H.M. Orientation effects and rheology of short glass fiber-reinforced thermoplastics: Polymer Science. Colloid Polym. Sci. 1984, 262, 257–269. [Google Scholar] [CrossRef]
  14. Schröder, T. Rheologie der Kunststoffe. Theorie und Praxis, 2nd ed.; Hanser: München, Germany, 2020. [Google Scholar]
  15. Kim, J.K.; Song, J.H. Rheological properties and fiber orientations of short fiber-reinforced plastics. J. Rheol. 1997, 41, 1061–1085. [Google Scholar] [CrossRef]
  16. Becraft, M.L.; Metzner, A.B. The rheology, fiber orientation, and processing behavior of fiber-filled fluids. J. Rheol. 1992, 36, 143–174. [Google Scholar] [CrossRef]
  17. Thomason, J.L. Glass fibre sizing: A review. Compos. Part A Appl. Sci. Manuf. 2019, 127, 105619. [Google Scholar] [CrossRef]
  18. Thomason, J.; Adzima, L. Sizing up the interphase: An insider’s guide to the science of sizing. Compos. Part A Appl. Sci. Manuf. 2001, 32, 313–321. [Google Scholar] [CrossRef]
  19. Becker, T.; Haag, M.; Gries, T.; Pico, D.; Wilms, C.; Seide, G.; Kleinholz, R.; Tiesler, H. Fibers, 12. Glass Fibers. In Ullmann’s Encyclopedia of Industrial Chemistry, 6th ed.; Bohnet, M., Ullmann, F., Eds.; Wiley-VCH: Weinheim, Germany, 2003; pp. 1–24. [Google Scholar]
  20. Skrabala, O.; Koslowski, T.; Haufe, S. Method for Influencing the Orientation of Fillers in a Plastic Melt During Production of a Moulding in an Injection Moulding Process; Universität Stuttgart: Stuttgart, Germany, 2017. [Google Scholar]
  21. Azaiez, J.; Chiba, K.; Chinesta, F.; Poitou, A. State-of-the-Art on numerical simulation of fiber-reinforced thermoplastic forming processes. Arch. Comput. Methods Eng. 2002, 9, 141–198. [Google Scholar] [CrossRef] [Green Version]
  22. Such, M.C.; Ward, C.; Potter, K. Aligned Discontinuous Fibre Composites: A Short History. J. Multifunct. Compos. 2014, 3, 155–168. [Google Scholar] [CrossRef]
  23. Harper, L.T.; Turner, T.A.; Martin, J.; Warrior, N.A. Fiber Alignment in Directed Carbon Fiber Preforms—A Feasibility Study. J. Compos. Mater. 2009, 43, 57–74. [Google Scholar] [CrossRef]
  24. Miyake, T.; Imaeda, S. A dry aligning method of discontinuous carbon fibers and improvement of mechanical properties of discontinuous fiber composites. Adv. Manuf. Polym. Compos. Sci. 2016, 2, 117–123. [Google Scholar] [CrossRef] [Green Version]
  25. Yu, H.; Potter, K.D.; Wisnom, M.R. A novel manufacturing method for aligned discontinuous fibre composites (High Performance-Discontinuous Fibre method). Compos. Part A Appl. Sci. Manuf. 2014, 65, 175–185. [Google Scholar] [CrossRef]
  26. Longana, M.L.; Yu, H.; Lee, J.; Pozegic, T.R.; Huntley, S.; Rendall, T.; Potter, K.D.; Hamerton, I. Quasi-Isotropic and Pseudo-Ductile Highly Aligned Discontinuous Fibre Composites Manufactured with the HiPerDiF (High Performance Discontinuous Fibre) Technology. Materials 2019, 12, 1794. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  27. Tang, J.; Swolfs, Y.; Longana, M.L.; Yu, H.; Wisnom, M.R.; Lomov, S.V.; Gorbatikh, L. Hybrid composites of aligned discontinuous carbon fibers and self-reinforced polypropylene under tensile loading. Compos. Part A Appl. Sci. Manuf. 2019, 123, 97–107. [Google Scholar] [CrossRef]
  28. Huntley, S.; Rendall, T.; Longana, M.; Pozegic, T.; Potter, K.; Hamerton, I. Validation of a smoothed particle hydrodynamics model for a highly aligned discontinuous fibre composites manufacturing process. Compos. Sci. Technol. 2020, 196, 108152. [Google Scholar] [CrossRef]
  29. Krajangsawasdi, N.; Nguyen, D.H.; Hamerton, I.; Woods, B.K.S.; Ivanov, D.S.; Longana, M.L. Steering Potential for Printing Highly Aligned Discontinuous Fibre Composite Filament. Materials 2023, 16, 3279. [Google Scholar] [CrossRef] [PubMed]
  30. Thomason, J.L. Glass Fibre Sizing: A Review of Size Formulation Patents; James, L., Ed.; Thomason: Glasgow, UK, 2015. [Google Scholar]
  31. Foss, P.H.; Tseng, H.-C.; Snawerdt, J.; Chang, Y.-J.; Yang, W.-H.; Hsu, C.-H. Prediction of fiber orientation distribution in injection molded parts using Moldex3D simulation. Polym. Compos. 2014, 35, 671–680. [Google Scholar] [CrossRef]
Figure 1. Vector and angle positions to describe fiber orientation with cartesian coordinates.
Figure 1. Vector and angle positions to describe fiber orientation with cartesian coordinates.
Jcs 07 00267 g001
Figure 2. Viscosity, shear stress and velocity profile of a viscoelastic plastic melt in a channel.
Figure 2. Viscosity, shear stress and velocity profile of a viscoelastic plastic melt in a channel.
Jcs 07 00267 g002
Figure 3. Melt velocity of an unfilled fluid and a suspension with rigid fillers.
Figure 3. Melt velocity of an unfilled fluid and a suspension with rigid fillers.
Jcs 07 00267 g003
Figure 4. Fiber orientation due to the melt flow in a normal channel (a) and in a nozzle with a converging flow (b).
Figure 4. Fiber orientation due to the melt flow in a normal channel (a) and in a nozzle with a converging flow (b).
Jcs 07 00267 g004
Figure 5. Schematic figure of the strand deposition process with highly oriented fibers in the strand.
Figure 5. Schematic figure of the strand deposition process with highly oriented fibers in the strand.
Jcs 07 00267 g005
Figure 6. Definition of the coordinate system in the sample with regard to the flow direction.
Figure 6. Definition of the coordinate system in the sample with regard to the flow direction.
Jcs 07 00267 g006
Figure 7. Proportion of oriented fibers in the strand pressed through the 1 mm nozzle according to the xx-, yy-, zz-orientation in the sample: (ac) represent the results of the three representative samples out of one strand. The colors show the respective process parameters.
Figure 7. Proportion of oriented fibers in the strand pressed through the 1 mm nozzle according to the xx-, yy-, zz-orientation in the sample: (ac) represent the results of the three representative samples out of one strand. The colors show the respective process parameters.
Jcs 07 00267 g007
Figure 8. Representative µCT-images showing the fiber orientation in the strand pressed through the 1 mm nozzle at different temperatures and shear stresses. On the right, an ashed part of the strand is depicted.
Figure 8. Representative µCT-images showing the fiber orientation in the strand pressed through the 1 mm nozzle at different temperatures and shear stresses. On the right, an ashed part of the strand is depicted.
Jcs 07 00267 g008
Figure 9. Pressure course over the extrusion time. The lines depict the measured pressure at the inlet and at the outlet of the 1 mm nozzle for the different process parameters.
Figure 9. Pressure course over the extrusion time. The lines depict the measured pressure at the inlet and at the outlet of the 1 mm nozzle for the different process parameters.
Jcs 07 00267 g009
Figure 10. Proportion of oriented fibers in the strand pressed through the 3 mm nozzle according to the xx-, yy-, zz-orientation in the sample: (ac) represent the results of the three representative samples out of one strand. The colors show the respective process parameters.
Figure 10. Proportion of oriented fibers in the strand pressed through the 3 mm nozzle according to the xx-, yy-, zz-orientation in the sample: (ac) represent the results of the three representative samples out of one strand. The colors show the respective process parameters.
Jcs 07 00267 g010
Figure 11. Representative µCT-images showing the fiber orientation in the strand pressed through the 3 mm nozzle at different temperatures and shear stresses. On the right, an ashed part of the strand is depicted.
Figure 11. Representative µCT-images showing the fiber orientation in the strand pressed through the 3 mm nozzle at different temperatures and shear stresses. On the right, an ashed part of the strand is depicted.
Jcs 07 00267 g011
Figure 12. Comparative overview of the averaged xx-value of the mean orientation tensor of three respective samples out of one strand for the different process parameters. The filled bars show the results of the 1 mm nozzle, the striped bars represent the results of the 3 mm nozzle. The results are compared to the normal fiber bundle structure in the capillary, shown here as the reference orientation.
Figure 12. Comparative overview of the averaged xx-value of the mean orientation tensor of three respective samples out of one strand for the different process parameters. The filled bars show the results of the 1 mm nozzle, the striped bars represent the results of the 3 mm nozzle. The results are compared to the normal fiber bundle structure in the capillary, shown here as the reference orientation.
Jcs 07 00267 g012
Figure 13. Representative µCT-images showing the normal fiber bundle structure in the capillary (a) and the manually unstructured fiber bundle structure (b).
Figure 13. Representative µCT-images showing the normal fiber bundle structure in the capillary (a) and the manually unstructured fiber bundle structure (b).
Jcs 07 00267 g013
Figure 14. Comparative overview of the xx-value of the mean orientation tensor. The results show the proportion of the oriented fibers for the samples with manually disturbed fiber orientation in the capillary, so that the fiber bundles are unstructured. The unstructured fiber bundle structure in the capillary is shown here as the reference orientation.
Figure 14. Comparative overview of the xx-value of the mean orientation tensor. The results show the proportion of the oriented fibers for the samples with manually disturbed fiber orientation in the capillary, so that the fiber bundles are unstructured. The unstructured fiber bundle structure in the capillary is shown here as the reference orientation.
Jcs 07 00267 g014
Figure 15. Weighted average fiber length of the fibers in the strand. The different colors represent the different process parameters. The gray bar shows the fiber length of the unprocessed granules as reference.
Figure 15. Weighted average fiber length of the fibers in the strand. The different colors represent the different process parameters. The gray bar shows the fiber length of the unprocessed granules as reference.
Jcs 07 00267 g015
Table 1. Varied process parameters for the described investigations.
Table 1. Varied process parameters for the described investigations.
ParameterLow LevelHigh Level
Temperature260 °C300 °C
Shear rate100 1/s1000 1/s or 470 m/s
Nozzle thickness1 mm3 mm
Fiber bundle structureNormalUnstructured
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Moritzer, E.; Tölle, L.; Greb, C.; Haag, M. Conceptions and Feasibility Study of Fiber Orientation in the Melt as Part of a Completely Circular Recycling Concept for Fiber-Reinforced Thermoplastics. J. Compos. Sci. 2023, 7, 267. https://doi.org/10.3390/jcs7070267

AMA Style

Moritzer E, Tölle L, Greb C, Haag M. Conceptions and Feasibility Study of Fiber Orientation in the Melt as Part of a Completely Circular Recycling Concept for Fiber-Reinforced Thermoplastics. Journal of Composites Science. 2023; 7(7):267. https://doi.org/10.3390/jcs7070267

Chicago/Turabian Style

Moritzer, Elmar, Lisa Tölle, Christoph Greb, and Markus Haag. 2023. "Conceptions and Feasibility Study of Fiber Orientation in the Melt as Part of a Completely Circular Recycling Concept for Fiber-Reinforced Thermoplastics" Journal of Composites Science 7, no. 7: 267. https://doi.org/10.3390/jcs7070267

APA Style

Moritzer, E., Tölle, L., Greb, C., & Haag, M. (2023). Conceptions and Feasibility Study of Fiber Orientation in the Melt as Part of a Completely Circular Recycling Concept for Fiber-Reinforced Thermoplastics. Journal of Composites Science, 7(7), 267. https://doi.org/10.3390/jcs7070267

Article Metrics

Back to TopTop