Manufacturing and Characterisation of a Tungsten Fibre-Reinforced Polymer Composite

Abstract

Using metal fibres in fibre-reinforced polymers is a way to tailor not only the mechanical properties but also material properties like, e.g., electrical and thermal conductivity or toughness. While recent works focus on ductile steel fibres, this work demonstrates the manufacturability of tungsten fibre-reinforced polymers. The Vacuum Assisted Process works well to quasi-unidirectionally reinforce an epoxy matrix with tungsten fibres of 150 µm diameter, achieving a fibre volume content of 23 ± 1% (±standard deviation). Tensile tests of 10 mm-wide tungsten fibre-reinforced polymer specimens yield a Young’s modulus of 89 ± 5 GPa, an ultimate tensile strength of 615 ± 33 MPa, and a failure strain of 1.9 ± 0.2%. The fractured specimens are further investigated, revealing that 66% of the tungsten fibres fail in a dominantly ductile manner with a strongly localised region of plastic deformation. This is a unique feature of tungsten fibres with the potential to enhance the fracture toughness of fibre-reinforced polymers.

1. Introduction

Fibre-reinforced polymers (FRP) are essential for structural lightweight applications. They allow to save weight and resources compared to traditional construction materials due to their high specific strength, their high specific Young’s modulus, and the ability to adjust their structure, meaning the orientation and position of the reinforcement fibres, according to given load paths [1]. Besides the structure, the reinforcement fibre type is decisive for the material properties of an FRP. Glass fibres (GF) and carbon fibres (CF) are the most used reinforcement fibre types because of their high specific ultimate tensile strength and specific Young’s modulus [2]. Less used fibre types are natural fibres, polymer fibres, and metal fibres [1], each having advantages and disadvantages regarding a given use case scenario.

In order to maintain the advantages of one fibre type while compensating for its disadvantages, two or more different reinforcement fibre types can be combined and embedded in a polymer matrix [3,4]. This concept, called fibre hybrid composites, triggered research on metal reinforcement fibres. The intention was to improve functional properties like the toughness and damage tolerance, or the electrical, thermal, or magnetic properties, of glass and carbon fibre-based FRP (GFRP and CFRP) [5,6]. While the specific mechanical properties are of primary interest in lightweight applications, such functional properties are vital to meet additional requirements. On the pathway towards new fibre hybrid composites, studies on FRP containing just metal fibres are helpful in identifying advantageous functional properties and gaining insights about the manufacturability of such composites.

Focusing on a toughness increase as an advantageous functional property, the development of stainless steel fibres (SF) that fail in a ductile manner at failure strains of up to 20% initiated research on endless steel fibre-based FRP (SFRP) [6]. SF provide a relatively high Young’s modulus of 193 GPa [6], ranging between the one of typical GF and CF (Table 1). The specific Young’s modulus is below that of CF and in the same order of magnitude as that of GF. In contrast to CF and GF, SF exhibit ductile fibre failure, which is beneficial in terms of toughness. This motivated investigations on the behaviour of SFRP under quasi-static and compressive loads [7], the influence of ductile and brittle matrices [6], the effect of the textile structure on the tensile and impact behaviour [8], the fibre–matrix interphase [9], and the influence of the fibre distribution on the fracture behaviour [10].

Tungsten fibres (WF), compared to SF, provide a higher Young’s modulus of approx. 406 GPa [11]. The ultimate tensile strength is above 2 GPa for fibres with a diameter of 150 µm [12] and can reach up to 4.5 GPa in the case of 16 µm-thin fibres [13]. The specific Young’s modulus of WF is in the same order of magnitude as the one of SF, while WF exhibit a higher specific ultimate tensile strength. Even though the failure strain is in a similar range to the one of CF or GF (Table 1), WF fail in a ductile manner, showing a pronounced region of localised plastic deformation [12,13]. While initially developed for incandescent lamps [14,15], the use of WF as reinforcement fibre dates back to the 1960s and 1970s when WF were used in metal–matrix composites, e.g., embedded in a copper matrix [16] or in nickel-based alloys [17]. Recently, WF-reinforced bulk tungsten was examined for use as plasma-facing material in future fusion reactors [18] with more information on this topic being available in Riesch et al. [19]. In WF-reinforced bulk tungsten, WF toughen the brittle bulk tungsten material through extrinsic toughening mechanisms [12]. Given the localised plastic deformation behaviour of WF, WF potentially also increase the toughness when added to CF or GF in a fibre hybrid composite. However, WF are currently not used in lightweight-driven FRP applications because of their high density and high fibre price [20], which was considered a drawback already in the 1970s when WF were still used as core fibres in boron reinforcement fibres [21]. In recent decades, WF have only been embedded in an epoxy matrix to investigate different fibre–matrix interface characterisation methods [22,23,24]. The potential of WF regarding advantageous functional properties for fibre hybrid composites, which motivated research on SFRP, remains unexploited. Table 1 shows that a substantial differentiation of WF from SF lies in the lower failure strain, comparable to that of CF and GF. However, the failure is still ductile and characterised by a strongly localised region of plastic deformation, which is an interesting feature in terms of toughness.

As a first step towards WF in fibre hybrid composites, this work aims to demonstrate the manufacturability of a tungsten fibre-reinforced polymer (WFRP). For this purpose, a standard epoxy resin and infiltration process are used. The obtained WFRP is characterised regarding the achieved fibre volume content and the mechanical properties. The mechanical properties are then compared with model expectations based on the rule of mixture and set in contrast to CFRP, GFRP, and SFRP. Particular focus is put on the fibre fracture to reveal if and to what extent the strongly localised plastic deformation behaviour previously reported for WF [12,13] also occurs in a WFRP.

Table 1. Characteristic material properties of an example-wise chosen carbon fibre (T300 Standard Modulus, Toray Carbon Fibers America Inc., Decatur, AL, USA [25]), glass fibre (E-glass, Saint-Gobain Vetrotex Deutschland GmbH, Herzogenrath, Germany [26]), steel fibre (annealed stainless steel AISI 316L, Baekert N.V., Zwevegem, Belgium [27]), tungsten fibre (potassium doped as-fabricated, 150 µm diameter, OSRAM GmbH, Munich, Germany [12]), and an epoxy matrix (CR80/CH80-10, SIKA Deutschland GmbH, Stuttgart, Germany [28]). References for information from other sources are mentioned in the respective cell of the table. Fibre prices are estimated similarly to Swolfs et al. [4].

	Carbon Fibre	Glass Fibre	Steel Fibre	Tungsten Fibre	Epoxy
Ultimate tensile strength in MPa	3530	3400	667	2721	80
Young’s modulus in GPa	230	73	193	410 [29]	3
Density in kg/dm3	1.76	2.60	7.85 [30]	19.3 [29]	0.95
Specific ultimate tensile strength in MPa/(kg/dm3)	2006	1308	85	141	84
Specific Young’s modulus in GPa/(kg/dm3)	131	28	25	21	3
Predominant failure behaviour	brittle	brittle	ductile	ductile	ductile
Failure strain in%	1.5	2.2–2.5	19.5	approx. 2.0	6.5
Diameter in µm	7	3–26 [1]	30	150	-
Thermal conductivity in W/(m×K)	105	1.0	14	170 [31]	0.15–0.25 [32]
Electrical conductivity in S/m	1.7 × 105	1 × 10−13–1 × 10−14	1.4 × 107–1.8 × 107 [30]	1.9 × 107 [31]	1 × 10−11–1 × 10−13 [33]
Price estimation	€€-€€€€	€	€€€	€€€€€	-

Table 1. Characteristic material properties of an example-wise chosen carbon fibre (T300 Standard Modulus, Toray Carbon Fibers America Inc., Decatur, AL, USA [25]), glass fibre (E-glass, Saint-Gobain Vetrotex Deutschland GmbH, Herzogenrath, Germany [26]), steel fibre (annealed stainless steel AISI 316L, Baekert N.V., Zwevegem, Belgium [27]), tungsten fibre (potassium doped as-fabricated, 150 µm diameter, OSRAM GmbH, Munich, Germany [12]), and an epoxy matrix (CR80/CH80-10, SIKA Deutschland GmbH, Stuttgart, Germany [28]). References for information from other sources are mentioned in the respective cell of the table. Fibre prices are estimated similarly to Swolfs et al. [4].

	Carbon Fibre	Glass Fibre	Steel Fibre	Tungsten Fibre	Epoxy
Ultimate tensile strength in MPa	3530	3400	667	2721	80
Young’s modulus in GPa	230	73	193	410 [29]	3
Density in kg/dm3	1.76	2.60	7.85 [30]	19.3 [29]	0.95
Specific ultimate tensile strength in MPa/(kg/dm3)	2006	1308	85	141	84
Specific Young’s modulus in GPa/(kg/dm3)	131	28	25	21	3
Predominant failure behaviour	brittle	brittle	ductile	ductile	ductile
Failure strain in%	1.5	2.2–2.5	19.5	approx. 2.0	6.5
Diameter in µm	7	3–26 [1]	30	150	-
Thermal conductivity in W/(m×K)	105	1.0	14	170 [31]	0.15–0.25 [32]
Electrical conductivity in S/m	1.7 × 105	1 × 10−13–1 × 10−14	1.4 × 107–1.8 × 107 [30]	1.9 × 107 [31]	1 × 10−11–1 × 10−13 [33]
Price estimation	€€-€€€€	€	€€€	€€€€€	-

2. Materials and Methods

2.1. Raw Materials

Woven fabrics made from commercial potassium-doped WF (OSRAM GmbH, Munich, Germany), as described and characterized in Gietl et al. [34], were used as reinforcement material. The fabric consisted of warp fibres with a diameter of 150 µm and an ultimate tensile strength of 2721 MPa [12]. Tungsten weft fibres of 50 µm diameter were used at a weft fibre distance of approx. 3 mm. The fabric thickness was approx. 200 µm, and the fabric width was approx. 53 mm with a total number of 200 warp fibres, resulting in a measured space of 113 ± 8 µm in between the fibres. The fibre spacing of the warp fibres was one order of magnitude smaller than the weft fibre spacing, while the warp fibre diameter was three times that of the weft fibres. Therefore, the woven fabric is referred to as quasi-unidirectional. In tungsten fibre-reinforced tungsten, the maximal achievable fibre volume content with this fabric was 25.5%, according to Gietl et al. [34].

The matrix material was an epoxy system consisting of the resin Biresin CR80 and the hardener Biresin CH80-10, supplied by SIKA Deutschland GmbH. This resin system is suitable for resin infusion under flexible tooling and shows good fibre impregnation due to its low viscosity. It is used, amongst others, for composites in the wind energy and marine sector because of its long pot life of 330 min [28]. While in future, screening experiments to determine the ideal matrix choice for WFRP will be required, the selected epoxy system is a reasonable starting point for this proof-of-concept study due to its good handling properties. The adhesion of epoxy on metal, and thus also tungsten, is generally known as excellent [35]. Therefore, no coating was applied on the WF in this first study, knowing that the adhesion between epoxy and tungsten can be further improved through coatings or surface treatments in the future [35,36,37].

2.2. Production of Composite Specimens

The Vacuum Assisted Process (VAP), a process patented by Trans-Textile GmbH (Freilassing, Germany) and Airbus Defence and Space GmbH (Friedrichshafen, Germany), was used to produce five quasi-unidirectional WFRP plates. For this purpose, four layers of the woven fabric with a length of approx. 270 mm were stacked to a pile and placed on a glass plate. Then, the VAP setup, including a peel ply, a flow promoter, a semipermeable membrane, a vacuum suction fleece, and a vacuum foil, was built up on top.

For each WFRP plate, 200 g Biresin CR80 (resin) and 60 g Biresin CH80-10 (hardener) were manually mixed and degassed at 1–3 mbar until no more gas bubbles were present in the resin. Afterwards, the fabric was infiltrated at room temperature at an initial infiltration pressure of 3–5 mbar. A ramped/stepped cure cycle was performed after complete infiltration consisting of the following steps: heating by 0.2 °C/min to 45 °C, holding at 45 °C for 1 h, heating by 0.2 °C/min to 80 °C, holding at 80 °C for 8 h, and cooling to room temperature by −0.5 °C/min.

From each WFRP plate, three specimens of approx. 10 mm width and two specimens of approx. 3 mm width, to investigate a potential size effect, having a total length of 250 mm, were cut by water jet cutting. Water jet cutting was chosen due to the hard and abrasive nature of WF. Because of the quasi-unidirectional characteristic of the WFRP specimens, with a predominant reinforcement in the tensile direction, the specimens were equipped with 1 mm-thick GF-reinforced polymer end tabs, as specified in DIN EN ISO 527-5 [38]. Accounting for a free specimen length of 150 mm, the end tabs were attached to the WFRP plate using UHU endfest 300. This adhesive required an additional curing of 2 h at 70 °C.

An identical procedure was followed to obtain neat epoxy specimens from an unreinforced epoxy plate manufactured using the same process parameters.

2.3. Experimental Setup and Methodology

Before further testing, each specimen was visually inspected. The specimen width and thickness were averaged from measurements at three different positions: 10 mm away from the upper and lower end tab and in the specimen centre. Tensile tests were performed using the universal testing machine Inspekt 250 from Hegewald & Peschke GmbH, Nossen, Germany. The testing machine was equipped with a 250 kN load cell and mechanical wedge action grips, leading to sharp and clear imprints of the grips in the end tabs. This indicates that no slippage occurs. A constant displacement rate of 2 mm/min was used, as recommended in DIN EN ISO 527-5 [38]. Strain measurements were performed using the digital image correlation system ARAMIS from Carl Zeiss GOM Metrology GmbH, Braunschweig, Germany. Eight WFRP specimens and the neat epoxy specimens were globally monitored with an ARAMIS measurement field of 150 mm × 110 mm (large measurement field, Figure 1). To achieve a higher spatial resolution, 17 WFRP specimens were monitored with an ARAMIS measurement field of only 35 mm × 26 mm (small measurement field, Figure 1). The speckle pattern required for the digital image correlation was adapted to the size of the measurement field. An image frequency of 2–4 Hz, a subset size of 17 × 17 pixels, and a subset distance of 15 pixels were used. The average strain was determined via a virtual strain gauge of 30 mm length and 2 mm width close to the upper clamping. This was to cover potential clamping effects close to the end tabs.

Based on the obtained stress-strain curves, the Young’s modulus in the strain range from 0.05% to 0.25%, the ultimate tensile strength as the highest occurring stress, and the failure strain, defined as the strain when the stress drops to zero, were determined for all specimens with a thickness deviation of less than 0.2 mm.

The fractured specimens were examined using a Keyence VHX-600 digital microscope (Keyence Deutschland GmbH, Neu-Isenburg, Germany) and a Philips XL30 ESEM scanning electron microscope in secondary electron imaging mode. By counting the number of fibres m visible in the fracture surface of each specimen, completely monitored via digital microscopy, the fibre volume content φ of each specimen was determined based on the measured specimen width w, the specimen thickness t, the known WF diameter d, and the assumption of zero porosity:

$$\phi ={\displaystyle \frac{m{\left(\frac{d}{2}\right)}^2}{wt}}$$

(1)

This method is feasible due to the excellent visibility of the 150 µm-thick WF in the fracture surface and the absence of pores indicated by the microscopy analysis.

2.4. Modelling Using the Rule of Mixture

The experimentally determined Young’s modulus and ultimate tensile strength were compared to a model prediction based on the rule of mixture. Model predictions for GFRP, CFRP, and SFRP are provided for comparison using the material data provided in Table 1. The rule of mixture is a simple model that estimates the material properties of unidirectional reinforced FRP X based on the single fibre property X_f, the matrix property X_m, and the fibre volume content φ [39]:

$$X=\phi X_f+(1-\phi )X_m$$

(2)

A potential influence of the weft fibres is not represented in this model and neglected as the reinforcement in warp fibre direction is dominant for the WFRP studied in this work.

3. Results and Discussion

3.1. Production of Composite Specimens

Figure 2 shows a manufactured WFRP plate proving that the VAP is suitable for infiltrating WF fabrics. Visual inspection of the WFRP plates confirms the absence of pores. The average plate thickness is 1.27 mm, and the average specimen width after water jet cutting is 9.80 mm for the wide and 2.81 mm for the thin specimens. Some specimens show thickness variations between different measurement points of more than 0.2 mm and are excluded from further evaluation. Such thickness deviations are caused by the high stiffness and poor drapability of WF woven textiles made from 150 µm-thick fibres. The consolidation pressure achieved during the VAP is not always sufficient to flatten these textiles, which exhibited a slight curvature because they are provided on a reel.

The absence of pores allows for the determination of the fibre volume content using Equation (1). The achieved average fibre volume content is 23 ± 1%, which is close to the theoretical expectation of 25.5% [34]. This means that approx. 145 fibres are present in the wide specimens and approx. 41 in the thin specimens. Compared to other FRP materials, the obtained fibre volume content is low. Quasi-unidirectional SFRP has been manufactured with fibre volume contents of approx. 40% [7] to 45% [8]. In general, fibre volume contents of 50% to 70% are achievable for unidirectionally reinforced laminates [40]. While a low fibre volume content leads to inferior mechanical properties, it facilitates infiltration and, combined with the plate thickness of only 1.28 mm, explains the pore-free specimens.

Whereas the manufacturability of WFRP using the VAP can be confirmed, major challenges during the manufacturing process are related to the nature of the used WF textile, i.e., defects such as broken or misaligned fibres and poor drapability due to high stiffness. For future works, it is necessary to develop defect-free, more flexible WF textiles made from thinner WF with reduced fibre spacing. Such thinner WF with diameters down to 16 µm exist [13] and are of particular interest with regard to fibre hybrid composites where WF have to be handled in textile processes together with, e.g., CF or GF.

3.2. Tensile Properties

Table 2. Mechanical properties of the tested wide and thin WFRP specimens with thickness variations less than 0.2 mm, a fibre volume content of 23 ± 1%, and the neat resin specimens. Given are the mean values ± standard deviation (see Appendix A for Equation (A1)).

Specimen Type	Wide	Thin	Epoxy
Valid specimens	10	5	5
Young’s modulus in GPa	89 ± 5	91 ± 2	3.1 ± 0.1
Ultimate tensile strength in MPa	615 ± 33	576 ± 18	66 ± 2
Failure strain in %	1.9 ± 0.2	1.5 ± 0.4	4.4 ± 1.1

provides the mechanical properties deduced from tensile testing. Regarding the epoxy matrix, a good agreement exists between the experimentally determined Young’s modulus and the datasheet [28], while the ultimate tensile strength and the failure strain deviate. This discrepancy can be explained by the decision to, for consistency reasons, use the same specimen geometry and strain gauge characteristics as for the WFRP specimens, which are different from the testing standard for unreinforced polymers.

Figure 3 shows the normalised stress-strain curve of the tested WFRP specimens, which exhibit a linear section from 0% to approx. 0.3% strain, followed by a nonlinear section indicating plastic deformation. Specimen fracture occurs at the maximum stress ranging between approx. 550 MPa and 660 MPa at a failure strain of 1.9 ± 0.2% for the wide and 1.5 ± 0.4% for the thin specimens. This failure strain is below the one determined in single WF tensile tests [12] and also below the failure strain of the epoxy resin. Undulations of the WF in woven form, fibre misalignment, and weave defects are possible causes.

The determined Young’s modulus, 89 ± 5 GPa for the wide specimens and 91 ± 2 GPa for the thin specimens is close to the model prediction of the rule of mixture shown in Figure 4a.

The ultimate tensile strength depends on the specimen size. With 615 ± 33 MPa, the wide specimens have a higher ultimate tensile strength compared to the 576 ± 18 MPa of the thin specimens. For the wide specimens, the ultimate tensile strength is only slightly lower than the model expectation provided in Figure 4b, which gives the upper boundary for the expected material properties. In an actual specimen, undulations of the WF in woven form, fibre misalignment, and weave defects lead to a reduced ultimate tensile strength. However, the overall good match between experimental results and model prediction validates the assumption of a quasi-unidirectional WFRP material and the choice of a rectangular specimen geometry.

The size effect observed for the ultimate tensile strength can be explained by the influence of the water jet cutting. Water jet cutting causes a rough surface with visible scratches in the epoxy or cut-through fibres. The influence of such a rough surface at the specimen edges is more dominant in the case of a thin specimen width.

Figure 5 shows the strain field evolution derived via digital image correlation in comparison to the related stress–strain curve (Figure 5a–e) and after fracture (Figure 5f). The strain fields confirm edge effects and explain why the determined ultimate tensile strength of wide specimens is closer to the predicted strength. It is also visible that an influence of the clamping exists, as specimen failure often occurs close to the upper clamping (Figure 5f). However, comparing the stress at a point where still a homogeneous strain field exists (Figure 5d) with the point where a localised strain increase close to the clamping occurs (Figure 5e) reveals that the influence of the clamping does not significantly affect the determined ultimate tensile strength.

3.3. Fracture Analysis

Figure 6 shows a fractured WFRP specimen. A fracture plane perpendicular to the tensile direction is visible, which is typical for unidirectional reinforced FRP under tensile load [41]. In the surrounding area of the fracture plane, matrix cracks with a 45° orientation towards the tensile direction are present. The direction of those cracks is mirrored at the fracture plane. Such cracks in the fracture region indicate a shear failure of the epoxy matrix. For SFRP, shear stress has been associated with fibre misalignment [8], which also exists in the WFRP specimens. Cracks in an epoxy matrix close to weft fibres (arrow in Figure 6) have alternatively been explained with a poor adhesion between fibres and the matrix, which is the case in the WFRP specimens, and stress concentrations at the weft fibres [6].

Figure 7 provides a detailed view of the fractured surface, where the imprint of a weft fibre is visible in the polymer matrix (Figure 7a). This is typical, as in all investigated fracture surfaces a weft fibre or the imprint of a weft fibre is present. The imprint of the weft fibre in Figure 7 is surrounded by a rough surface with a river line structure evolving from the weft fibre imprint. This observation indicates a crack propagation initiated by the weft fibre. Weft fibres are a known source of failure initiation in FRP [7,42,43], as they cause undulations in the warp fibres. During tensile loading, the warp fibres are straightened, leading to zones of stress concentrations [7]. For SF in an epoxy matrix, Allaer et al. [7] conclude that such stress concentrations lead to microcracks parallel to the weft fibre. The microcracks then cause a fibre–matrix debonding and a strain concentration at the warp fibres, initiating plastic deformation. Callens et al. [6] set up a similar hypothesis for SFRP with a brittle matrix, assuming that cracks in the matrix induce stress concentrations at the warp fibres, leading to local debonding and strain magnification in the fibres. This would then initiate a plastic deformation of the fibres and a fibre failure at a weak point [6]. A fibre failure at a weak point following plastic deformation is also plausible for WFRP. Given the weft fibre-initiated WFRP failure observed in this work, it can be concluded that for future WFRP, WF woven textiles with weft fibres of smaller diameter and a larger weft fibre distance are beneficial because this would reduce the waviness of the textile and the number of weft fibres used.

Figure 7b,c shows the matrix fracture surface in detail. Regions with a rough, hackle appearance, typically associated with a fast crack propagation [44], and regions with a smooth, mirror appearance can be differentiated. This indicates that stress concentration and loading speed locally differ during the course of the fracture due to fibre undulations, fibre misalignment, and an inhomogeneous fibre distribution. In Figure 7d, scarps separating different fracture planes are visible. The tips of the fractured fibres exhibit different shapes. This is also reflected in a different surface appearance of the fractured fibres during light optical microscopy, e.g., in Figure 7b. Due to a different surface morphology, fibres that failed in a dominantly ductile manner have a darker appearance. In comparison, fibres that failed in a dominantly brittle manner have a brighter appearance. Averaged over all fracture surface images, approx. 66% of the WF fail in a dominantly ductile manner.

Figure 8 presents scanning electron microscopy images of the fractured fibres. For fibres failing in a dominantly ductile manner (Figure 8a), a significant diameter reduction compared to the initial fibre diameter and necking are visible due to plastic deformation in the fracture region. The diameters of the fractured fibres range between 100 µm and 120 µm, which is larger than the diameter reduction measured for single WF by Riesch et al. [45]. Figure 8b shows a fibre that dominantly failed in a brittle manner and only exhibits marginal necking. Only in a narrow region at the outer fibre edge, a similar surface morphology than observed for ductile failed fibres is visible. This underlines that fibres failing in a dominantly brittle manner exhibit a partial ductile fracture. To understand the occurrence of brittle fibre failure, it is worth noting that stress concentrations and loading speed influence the failure behaviour of a fibre [46]. Those parameters differ locally, e.g., due to undulations, fibre misalignment, and an inhomogeneous fibre distribution. For WF reinforcing bulk tungsten, fibres failing in a dominantly ductile and a brittle manner have been observed [47]. Gietl et al. [48] described that WF failed in a brittle manner if they did not debond. In the present WFRP, this is unlikely the case because a weak fibre–matrix adhesion can be deduced based on Figure 8. A gap between the fractured fibres and the surrounding matrix (Figure 8a,b) is visible, and no matrix material is attached to the fibres sticking out of the fracture plane (Figure 8c).

Figure 8d shows a fibre exhibiting multiple fibre necking, which rarely occurs. It can be explained by the existence of different levels of crack propagation separated by scarps (Figure 7d) seldomly allowing two crack fronts to reach a warp fibre simultaneously.

3.4. Potential of Tungsten Fibres in Fibre-Reinforced Polymers

In a fibre selection process focusing on lightweight only, WFs are not competitive with CF and GF despite their high Young’s modulus and ultimate tensile strength (Table 1, Figure 4). This is because their high density results in a poor specific Young’s modulus and specific ultimate tensile strength. In addition, the high fibre price is a disadvantage in commercial applications. To identify use case scenarios for WF, the focus must lie on functional properties such as the thermal and electrical conductivity or the toughness of FRP. Concerning an improved material toughness, research on SFRP, with SF showing ductile fibre failure and a high Young’s modulus, was motivated. This work points out that similar characteristics apply to WF. The key difference between SF and WF is that the provided Young’s modulus and ultimate tensile strength of WF and the specific ultimate tensile strength are higher, while the failure strain is lower. Due to the low failure strain of WFRP compared to SFRP, using WF to increase the toughness of a material, defined as the total energy dissipated during fracture, is not expected. When it comes to the fracture toughness of a material, i.e., the amount of energy dissipated during fracture due to crack propagation, beneficial effects through extrinsic toughening mechanisms, as found for WF-reinforced tungsten [18], are expected because 66% of the WF in the tested WFRP failed in a dominantly ductile manner showing plastic deformation. Exceptional for WF is that this plastic deformation is strongly localised. Therefore, WF have the potential to locally dissipate energy during WFRP fracture already at small crack openings and hence increase the fracture toughness. This makes WF a promising fibre type for tailored fibre hybrid composites focusing on a high fracture toughness, where disadvantageous WF properties like the high density or the high fibre price are less relevant.

4. Summary and Conclusions

In this proof-of-concept study, quasi-unidirectional epoxy-based tungsten fibre-reinforced polymer plates are successfully manufactured using the Vacuum Assisted Process, starting from a tungsten fibre-woven textile with 150 µm-thick tungsten fibres. The achieved fibre volume content of 23 ± 1% is close to the expected fibre volume content. Tensile tests of 10 mm-wide tungsten fibre-reinforced polymer specimens, using a digital image correlation system for strain measurements, reveal a Young’s modulus of 89 ± 5 GPa (±standard deviation), an ultimate tensile strength of 615 ± 33 MPa, and a failure strain of 1.9 ± 0.2%. The Young’s modulus is close to the rule of mixture prediction, while the ultimate tensile strength is slightly reduced due to manufacturing and sample preparation defects.

Optical microscopy and scanning electron microscopy of fractured tungsten fibre-reinforced polymer specimens confirm that most tungsten fibres (66%) fail in a dominantly ductile manner when embedded in an epoxy matrix, showing a strongly localised necking region. This unique local plastic deformation behaviour, combined with a higher Young’s modulus, ultimate tensile strength, and specific ultimate tensile strength compared to steel fibres, makes tungsten fibres an interesting fibre type to enhance the functional properties of conventional fibre-reinforced polymers by adding tungsten fibres in the form of a tailored fibre hybrid composite.