**1. Introduction**

Natural fibre-reinforced polymer composites (NFPC) have been used in many industry branches for a number of years. Nowadays, it is hard to imagine the medical, automotive, aerospace and shipyard sectors and civil engineering without them [1–4]. NFPC, as well as carbon nanotubes-reinforced polymer composites rapidly growing [5].

Natural fibres nearly completely replaced synthetic fibres in polymer composites [6]. Nowadays, NFPCs include mainly lignocellulosic fibres obtained from different tree, grass and crop species. Their biodegradability, source renewability, low density at high strength and elasticity, and low cost and neutrality for humans and tools have been appreciated [7]. The fibres are obtained from hard and soft tissues (Figure 1)—wood, stalks, seeds, leaves, fruit, phloem, husks and shells being waste from agricultural production [2,7].

The most popular NFPC matrices include polypropylene (PP), high-density polyethylene (HDPE), polyvinyl chloride (PVC), and sometimes polystyrene (PS) [7–9]. Matrix selection depends on the composite's intended use [4]. Some matrices are made of biodegradable polymers, e.g., polyglycolic acid (PGA) and polyhydroxyalkanoates (PHA) [9].

Civil engineering applications are dominated by NFPC with PVC or HDPE matrix [4,6]. They are used in solid (Figure 2a) or cellular (Figure 2b) profiles intended for outdoor floors—on terraces and swimming pools (Figure 3) and ventilated façade cladding [10–12]. NFPC profiles are also employed in platforms, passages, landscape architecture and wet rooms [3–6].

**Citation:** Sudoł, E.; Kozikowska, E.; Szewczak, E. Artificial Weathering Resistance Test Methods for Building Performance Assessment of Profiles Made of Natural Fibre-Reinforced Polymer Composites. *Materials* **2022**, *15*, 296. https://doi.org/10.3390/ ma15010296

Academic Editor: Krzysztof Schabowicz

Received: 25 November 2021 Accepted: 28 December 2021 Published: 31 December 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

**Figure 1.** Types of plant fibres used in NFPC.

**Figure 2.** Sample NFPC building profiles for (**a**) facades, (**b**) floors. The dimensions are given in mm.

**Figure 3.** Sample application of NFPC profiles in civil engineering for outdoor floors: (**a**) at a swimming pool, (**b**) on the terrace.

The fitness of NFPC profiles for civil engineering applications, similarly to other construction materials, should be assessed according to the sustainable development concept, based on the usability criterion, by determining a collection of key features for the particular application [13]. The assessment is carried out from the angle of the product's influence on a building structure's fulfilling the seven essential requirements [14], according to the regulation of the European Parliament and the Council (EU) No. 305/2011 (CPR) [15]. The seventh essential requirement concerning the Sustainable use of natural resources states that building structures need to be designed and made so that natural resources are used sustainably and ensure the durability of building structures. Fulfilling the building structure's durability criterion depends on the construction products' resistance to operating factors, including the environmental ones [16,17]. The aspect of resistance to environmental conditions has special significance for outdoor products, such as terrace floors and façade cladding, which are directly exposed to long-term sun radiation, water impact, temperature changes and microbiological factors [12,18]. So far the weathering resistance test methods

for building performance assessment of profiles made of natural fibre-reinforced polymer composites have not been standardized.

Ensuring efficient interaction between lignocellulosic fibres and the polymer matrix is among the key challenges for natural fibre-reinforced polymer composites to achieve proper resistance to environmental factors [4,9]. The hydrophilic nature of the fibres makes them swell in an aqueous environment, which results in cracks formed in the hydrophobic polymer matrix [18]. As a result, the interaction between lignocellulosic fibres and the polymer deteriorates. Insufficient adhesion at the phase border leads to decreased mechanical parameters [9,19]. Hence, the fibres' surface is modified to improve the interphase interaction by increasing the fibres' wettability and reducing water absorption [4,9]. The most popular chemical methods involve employing substances whose particles react with cellulose hydroxy groups and introduce new groups linked with the polymer matrix's functional groups [20]. Similar treatments are used in the case of carbon nanotube reinforcement [21]. The number, shape, size and distribution of the fibres also affects the NFPC characteristics [4,7,8]. Proper scattering of fibres in the matrix promotes interphase adhesion by reducing voids and ensuring the fibres' surrounding by the matrix [2,4,21].

NFPC products are susceptible to sunlight [18,22]. Their exposure to UV light was discovered to contribute to a more significant decrease in the mechanical properties than exposure to microorganisms and high temperatures [23–28]. Tests on NFPC products' resistance to sunlight are typically carried out with accelerated methods, using laboratory light sources [27–36]. The application of accelerated methods involving product exposure to the relatively short but intensive impact of a factor or a set of service factors is standard for construction fitness assessment procedures. The impacts are selected according to the product's material characteristics, including the product's application scope, to simulate best the processes that occur during the product's use in real conditions [17]. As shown by previous studies, light and water impact cycles are the most burdensome exposure sequence in NFPC's accelerated weathering [32–38]. NFPC degradation progresses then much faster and more intensively than in the case of exposure to sunlight only [30]. In the wetting phase, the polymer matrix particles damaged as a result of UV impact are washed out, and successive ones are exposed [22], but hydrophilic lignocellulosic fibres swell too, which leads to reduced interphase adhesion, as was mentioned before [18,19]. As was already determined [34], the light exposure and spraying cycles cause much more intensive destruction of profiles whose surfaces were mechanically treated before than surfaces non-treated after extrusion. Composites with a higher share of lignocellulosic fibres on the surface, exposed during planing, show a higher drop in the flexural modulus [31,34]. NFPC was discovered to degrade faster than the polymer used as a matrix. At the initial weathering exposure stage, a pure polymer may be subjected to further cross-linking, while this property is physically limited in a composite by the filler [35]. The exposure time matters as well. The longer it is, the greater degradation occurs [34–36]. Light exposure reduces the mechanical properties and changes the NFPC products' colour [30,37].

Analysing previous studies on artificial weathering resistance of NFPC products addressed for civil engineering, it can be observed that different light sources are used in the exposure procedures, with diversified exposure sequence, including the dry and wet phase length and wetting method [34–37,39,40]. This paper contains a comparative analysis of the two most common methods used for construction products to determine the most relevant building performance assessment for natural fibre-reinforced polymer composite profiles. So far the comparative analysis of artificial weathering resistance test methods has not been performed. Tests were carried out for PVC and wood flour composite profiles as one of the most popular in civil engineering [10,11,26]. The influence was analysed of the applied exposure procedure on the changes in the composite microstructure and mechanical properties. A comparative exposure was performed, including the following:

• exposure to light emitted by xenon lamps (X-exposure) combined with alternate short spraying, with diversified exposure time,

• wetting through long-term condensation and then exposure to light emitted by fluorescent lamps (F-exposure), with diversified exposure time.

The influence was evaluated of the performed exposures on the usable surface's morphology and microstructure of brittle fractures, flexural strength, flexural modulus and impact strength.

#### **2. Materials and Methods**

#### *2.1. Profiles*

Commercial cellular profiles intended for outdoor floors were used for the tests. The profiles were made of PVC matrix composite with fine lignocellulosic fibre filler (wood flour) and plastifiers and modifiers as additives. The filler was recycled wood industry waste. The composite's formula is the manufacturer's trade secret and has not been revealed. The profiles were extruded in a plastic processing facility.

The profiles were 180 mm wide, 25 mm high, the front walls were 5 mm thick, and the chambers were 22 mm wide. The profiles had two usable surfaces: one grooved and one plane (Figure 4). As a standard, grooved and plane surfaces of construction profiles are mechanically treated (brushing) to provide a wood-like texture effect. Profiles with a standard usable brushed surface (SZ) and profiles with a non-brushed usable surface (NSZ)—for comparison—were used in the study.

**Figure 4.** Shape of the profiles used in the tests.

#### *2.2. Weathering Exposure*

The first weathering procedure (F-exposure) was carried out in UV Test apparatus (Atlas, Linsengericht, Germany) featured with 1A type (UVA-340) fluorescent lamps according to EN 16474-3 [41], emitting light in the wavelength range of 300 to 400 nm, with the maximum emission at 343 nm (Table 1). The exposure procedure complied with EN 927-6 [42]. The samples were exposed to cycles composed of a long condensation phase, followed by exposure to UV lamps, with the radiation intensity of 0.89 W/m2 measured at 340 nm wavelength, with alternated wetting cycles (water spraying) (Table 2).

Samples cut out from flat usable surfaces of brushed (SZ), and non-brushed (NSZ) profiles were exposed. The samples were 300 mm long, and the profiles' full width (180 mm) was maintained. During the exposure, the samples were arranged at ca. 80◦ angle, allowing free draining of water (Figure 5a). The exposure lasted 336 h—SZ-F-336 and NSZ-F-336 series, and 2016 h—SZ-F-2016 and NSZ-F-2016 series (Table 2).

The other weathering procedure (X-exposure) was performed in SunTest apparatus (Atlas, Linsengericht, Germany) featuring a xenon-arc lamp with a quartz shell, according to EN 1647-2 [43], which emits light from less than 270 nm in the ultraviolet range through visible spectrum up to IR, whereby a daylight filter was used, eliminating shortwave UV radiation (Table 3). Exposure was carried out according to EN ISO 4892-2 method A [44]. The cycles included exposure to light with the radiation intensity of 60 W/m2, measured in the band wavelength of 300–400 nm, combined with exposure to high temperature, and followed by water spraying (Table 2).

**Figure 5.** Weathering exposure procedure flowchart: (**a**) F-exposure, (**b**) X-exposure.

**Table 1.** Relative spectral intensity of radiation for UVA 340 fluorescent lamps [39] used in the UV test apparatus.




**Table 3.** Relative spectral intensity of radiation for a xenon-arc lamp with daylight filter [43] used in SunTest apparatus.


**Table 4.** Relative differences between the values of the mechanical properties of SZ and NSZ samples, calculated according to Formula (6), %. Statistically insignificant differences are highlighted in grey.


A brushed (SZ) and non-brushed (NSZ) flat usable surfaces were exposed. The samples' length ranged from 100 to 300 mm, and their width and thickness corresponded to the profile's dimensions. During the exposure, the samples were arranged horizontally, maintaining a ca. 10◦ slope to allow free draining of water (Figure 5b). The exposure lasted 300 h for the SZ-X-300 and NSZ-X-300 series and 2016 h for the SZ-X-2016 and NSZ-X-2016 series (Table 2).

Deionised water with pH 5.0 ± 7.5 and electric conductivity under 2 μS/cm measured at 25 ◦C were used for wetting in both weathering procedures.

#### *2.3. SEM Analysis*

The microstructure of composite profiles was examined with Sigma 500 VP cold-field emission scanning electron microscope (Carl Zeiss Microscopy GmbH, Köln, Germany), which allows reaching a high resolution at a low accelerating voltage. The tests were carried out at the accelerating voltage of 10 KeV inductive electron beam, using an SE detector on samples coated (sprayed) with a gold film.

At the first stage, the microstructure of brittle fractures obtained at 23 ◦C was observed. The observations covered samples cut out from brushed (SZ) and non-brushed (NSZ) profiles in their original condition and following X-exposure lasting 2016 h (SZ-X-2016 and NSZ-X-2016 series). The procedure was selected because it is expected to cause the most significant changes in the NFPC structure [45]. Observations were carried out at

500× and 20,000× magnification. At the second stage, the observations covered the usable surface microstructures in NSZ profiles in their original condition, following F-exposure (NSZ-F-336 and NSZ-F-2016 series) and X-exposure (NSZ-X-300 and NSZ-X-2016 series), at 500× magnification. The observations were not carried out for brushed profiles because of the high roughness of the usable surface, which made SEM examinations impossible.

#### *2.4. Testing Mechanical Properties*

Mechanical properties were tested on samples obtained from brushed (SZ) and nonbrushed (NSZ) profiles in their original condition and following a short- and long-term F-exposure and X-exposure (Table 2). The flexural strength, flexural modulus and impact strength were tested.

The flexural modulus was also tested according to EN ISO 178 [46], using a class 1 strength testing machine (Instron, Darmstadt, Germany). Three-point bending was performed according to EN ISO 178 [46], using samples sized 15 × 100 × 5 mm, cut out from the central part of the profile's front wall, parallel to vertical ribs (Figure 4). Supports with a 5 mm radius were used, spaced every 80 mm, corresponding to 16-times sample's thickness and a 5 mm radius pressing element placed in the middle of the span. The samples were freely supported (Figure 6a). The load was applied to the front surface at a constant rate of 5 mm/min. until destruction. Flexural strength *σ<sup>f</sup>* was calculated according to (1) and expressed in N/mm2. Twelve samples were tested in each series, giving a total of one hundred and twenty samples tested in the study.

$$
\sigma\_f = \frac{3FL}{2bh^2} \tag{1}
$$

where: *F*—maximum force, in N; *L*—support spacing, in mm; *b*—sample's width, in mm; *h*—sample's thickness, in mm.

**Figure 6.** Testing mechanical properties: (**a**) flexural strength, (**b**) impact strength.

The flexural modulus was also tested according to EN ISO 178 [46], using a class 1 strength testing machine (Instron, Darmstadt, Germany), in conditions identical to flexural strength tests. A load-deflection curve was recorded during bending in a linearly elastic range, including the force and deflection values corresponding to strain ε*f*<sup>1</sup> = 0.0005 and ε*f*<sup>2</sup> = 0.0025. The *f* <sup>1</sup> and *f* <sup>2</sup> deflection values were calculated according to Formula (2).

$$f\_1 = \frac{\varepsilon\_{f1}L^2}{6\hbar};\ f\_2 = \frac{\varepsilon\_{f2}L^2}{6\hbar} \tag{2}$$

where: *L*—spacing of supports, in mm; *h*—sample's thickness, in mm.

The force values recorded when ε*f*<sup>1</sup> and ε*f*<sup>1</sup> strain occurred were used for determining the values of σ*f*<sup>1</sup> and σ*f*<sup>2</sup> normal stress. The *Ef* modulus was calculated according to (3) and expressed in N/mm2. Twelve samples were tested in each series, giving a total of one hundred and twenty samples tested in the study.

$$E\_f = \frac{\sigma\_{f2} - \sigma\_{f1}}{\varepsilon\_{f2} - \varepsilon\_{f1}} \tag{3}$$

where: σ*f*1, σ*f*2—maximum normal stress corresponding to *f* <sup>1</sup> and *f* <sup>2</sup> stress determined according to (2).

The impact test was carried out with Charpy impact pendulum (ZwickRoell, Ulm, Germany) according to EN ISO 179-1 [47]. The samples used in the test had no notch, were sized 10 × 80 × 5 mm, cut out from the central part of the profile's front wall, parallel to the vertical ribs. The sample was freely resting on supports spaced at 62 mm and then hit with a 2J impact pendulum (Figure 6b). The load was exerted on the front surface. Charpy impact strength *acU* was calculated according to (4) and expressed in kJ/m2. Eight samples were tested in each series, giving a total of eighty samples tested in the study.

$$n\_{cII} = \frac{E\_c}{h \cdot b} \cdot 10^3 \tag{4}$$

where: *Ec*—energy absorbed by breaking the test specimen, in J; *h*—sample's thickness in mm; *b*—sample's width, in mm.

#### *2.5. Analysis of the Statistical Difference in the Mechanical Properties Test Results*

The changes in the tested materials' mechanical properties were analysed based on the characteristics' differences after F-exposure and X-exposures. Since in most cases, the differences between the results before and after the exposure were relatively low compared to the results' variability in the groups, the statistical significance of the differences was analysed with a one-way analysis of variance (ANOVA F-test).

The difference in the given mechanical property before and after weathering (Δ*Y*) was calculated with the following equation:

$$
\Delta Y = 100\% \cdot \frac{Y\left(T\_j\right) - Y\left(T\_i\right)}{Y\left(T\_i\right)}\tag{5}
$$

where: *Ti*, *Tj*—ageing times used; *Y*(*Ti*)—mean value of the given mechanical property after weathering for *Ti*, *Y*(*Tj*)—mean value of the given mechanical property after weathering for *Tj*.

Taking into account that two sample series—obtained from brushed (SZ) and nonbrushed (NSZ) usable surface of the profiles—were subjected to mechanical properties tests before and after weathering exposure, an analysis of the exposure influence on the properties of interest was preceded by an assessment of the differences between the properties of SZ and NSZ samples in their original condition. The following formula was used for calculating the relative difference:

$$
\Delta Y = 100\% \cdot \frac{Y\_{SZ} - Y\_{NSZ}}{Y\_{SZ}} \tag{6}
$$

where: Δ*Y*—difference between the mechanical properties of material *Y* with brushed *SZ* (*YSZ*) and non-brushed NSZ (*YNSZ*) surface.

The statistical significance of the differences was analysed with ANOVA F-test. The results are summarised in Table 4. No surface treatment influence was observed only for the modulus of elasticity. For flexural strength, the relative difference between the values obtained for SZ and NSZ samples amounted to 10.2%, while for the Charpy impact strength it was 9.6%. Both characteristics were higher for SZ than for NSZ samples. The exposure impact on all analysed mechanical properties was assessed separately for each surface type because of the statistically significant difference in the flexural strength and Charpy impact strength tests for SZ and NSZ surface samples.

#### **3. Results and Discussion**
