*3.1. Microstructure Analysis*

The observations of the brittle fracture microstructures helped evaluate only the dispersion rate of a filler in a polymer matrix. The composite's observed structure can be considered inhomogeneous [47,48]. Numerous wood flour clusters were discovered, forming combinations of fibres and plates with diameters ranging from 50 μm to 100 μm (Figures 7 and 8). Because a fracture in a composite occurs typically in the sample's most weakened areas, material defects in the form of pores and voids became visible at the fracture, being a testimony to the plates and wood fibres being pulled out from the polymer matrix [48] (Figures 7 and 8). Further analysis of the brittle fractures' microstructure revealed the presence of other fillers' clusters, most likely being mineral fillers (talc or chalk) and relatively regular shape and size not exceeding 1 μm. They were generally well dispersed in the polymer matrix (Figures 9 and 10), but some cluster sizes from 5 μm to 10 μm (Figure 10b) were also discovered. The performed SEM analysis of brittle fractures did not reveal microstructural differences in the material in its original condition compared to the material after X-exposure for 2016 h (SZ-X-2016 and NSZ-X-2016 series). The data collected in the brittle fracture analysis, revealing the microstructure at the material crosssection, can suggest that the material's inner structure did not change under the influence of the applied weathering procedure.

**Figure 7.** Microstructure of the profile's fracture surface in the original condition: (**a**) SZ profile, magnification: 500×, (**b**) NSZ profile, magnification: 500×.

**Figure 8.** Microstructure of the profile's fracture surface after weathering: (**a**) SZ-X-2016 series profile, magnification: 500×, (**b**) NSZ-X-2016 series, magnification: 500×.

**Figure 9.** Microstructure of the profile's fracture surface in the original condition: (**a**) SZ profile, magnification: 20,000×, (**b**) NSZ profile, magnification: 20,000×.

**Figure 10.** Microstructure of the profile's fracture surface after weathering: (**a**) SZ-X-2016 series profile, magnification: 20,000×, (**b**) NSZ-X-2016 series profile, magnification: 20,000×.

An SEM surface analysis was carried out, taking into account the fractures' surface microstructure analyses and bearing in mind that the profiles' usable surface was directly exposed. The tests covered only the non-brushed profiles because of brushed profiles' high surface roughness, which prevented their observations. An analysis of NSZ samples' surface microstructure in the original condition revealed a uniform coating of the fibres with polymer (Figure 11). The surface was relatively smooth and uniform, characteristic of extruded NFPC profiles [48,49]. No exposed wood fibres were observed. Following the profiles' X-exposure, significant changes in the surface morphology were observed already after 300 h. The microscopic image revealed melting of the polymer's outermost layer, exposing the surfaces of fillers not wetted with the polymer, taking the form of large plates and wood fibre clusters (Figure 12a). Extending the exposure time to 2016 h significantly aggravated the top layer's degradation. Highly non-homogenous surface topography with molten areas was observed [50]. The revealed microstructure contained agglomerated wood fibres (Figure 12b).

**Figure 11.** Microstructure of NSZ profile's surface in the original condition (**a**) magnification: 500×, (**b**) magnification: 1000×.

**Figure 12.** Surface microstructure after weathering, magnification: 500× (**a**) NSZ-X-300 series, (**b**) NSZ-X-2016 series, (**c**) NSZ-F-336 series, (**d**) NSZ-F-2016 series.

F-exposure also contributed to the changes in the surface morphology. The microstructure changes were reported after 336 h of exposure (NSZ-F-336 series), and minor molten areas in the polymer's outermost layer became visible, exposing the filler's surface (Figure 12c). Still, the changes are noticeably more minor than those reported for samples after X-exposure for 300 h (Figure 12a). The F-exposure time extension to 2016 h aggravated the profile's outermost layer, making the filler much more visible (Figure 12d). It should be emphasised that the degradation rate of NSZ-F-2016 series samples was significantly lower than the degradation rate of samples after X-exposure for the same exposure duration (NSZ-X-2016 series).

Summing up the results of microstructural tests, it can be concluded that the applied weathering procedures performed with laboratory light sources did not affect the composite's internal structure. No differences that could be considered microstructure changes were observed in the brittle fracture analysis [48,49]. Under UV-irradiation influence the surface layers between PVC matrix and wood fibres became more brittle. Due to these factors create additional stresses at the interface of the components, causing development of the cracks on the weathered surface of the samples [49]. Exposing the profile to light emitted by fluorescent lamps (F-exposure) and xenon lamps (X-exposure) caused significant surface degradation. The surface morphology analysis revealed molten areas in the polymer matrix's outermost layer, exposing the surface of lignocellulosic fibres. Extended exposure aggravated the degradation of the profiles' usable layer, which corresponds to the literature data [45,50]. Weathering impacts exerted with a xenon lamp affected the surface properties more significantly than F-exposure for the same exposure time. The above can be explained by the differences between relative spectral intensity of radiation for UVA 340 fluorescent lamps (F-exposure) and relative spectral intensity of radiation for a xenon-arc lamp (X-exposure). It is supposed that wavelengths rays between 360 nm and 400 nm is the most important factor causing photodegradation to some organic substances such as PVC. F-exposure contains about 26% wavelengths rays between 360 nm and 400 nm (Table 1) while X-exposure contains about 54% (Table 3). Surface damage after X-exposure was more intensive and vast than after F-exposure.

#### *3.2. Mechanical Properties*

An analysis of the results suggests that the analysed material's flexural strength in the original condition was 60 ÷ 67 MPa; 63 ÷ 64 MPa after F-exposure, and 61 ÷ 64 MPa after X-exposure (Figure 13). These values are similar to those obtained for construction profiles made of wood fibre-reinforced composites with PE matrix and PVC matrix with rice husk fibre, for which the original condition values amounted to 71 MPa and 67 MPa, respectively [51]. They exceed the test results on composites with recycled high-density polyethylene matrix and rice husk fibre filler, which reached the flexural strength of 25 MPa for the filler content of 50% and 38 MPa for the filler content of 80% [28]. Still, they are lower than the results for composites with polymer matrix reinforced with sycamore, sisal or bamboo fibres, whose flexural strength ranged from 100 MPa to 134 MPa [52].

**Figure 13.** Flexural strength test results of brushed (SZ) and non-brushed (NSZ) samples in basic state (laboratory conditions), after F-exposure for 336 h (F-336) and 2016 h (F-2016) and after Xexposure for 300 h (X-300) and 2016 h (X-2016). The error bands represent the standard deviation (series size n = 12).

The analysis of F-exposure's influence on the flexural strength revealed a decrease for SZ series samples. No decrease was reported for NSZ samples. Still, it should be emphasised that a statistically significant change in the strength occurs already after the first exposure period T1, which lasts 336 h for F-exposure and 300 h for X-exposure. Further exposure up to 2016 h does not cause a significant change in the strength (Figure 14).

**Figure 14.** Diagrams showing the differences in the sample's flexural strength σf, MPa, after F-exposure and X-exposure for time T1 and T2: (**a**) brushed sample (SZ), (**b**) non brushed sample (NSZ). The error bars show the standard deviation (series size n = 12). The tables below summarise the relative change in the flexural strength Δσf, %, during exposure time (T1 − T0) (T2 − T1) and (T2 − T0) calculated according to Formula (5). Statistically insignificant differences Δσ<sup>f</sup> are highlighted in grey.

Different behaviour of SZ and NSZ profiles during F-exposure, where each weekly cycle starts with a 24 hours' phase of wetting through condensation (Table 2), can be explained by the difference in the surface's condition. As demonstrated in a previous study, mechanically treated profiles can be more susceptible because of lignocellulosic fibres' exposure in the process [7,28,31]. It is assumed that the exposed hydrophilic fibres swell due to their wetting, which weakens the interaction forces between the matrix and the filler and deteriorates the strength [20,53]. A similar effect was observed for planned profiles made of HDPE composite with a wood flour filling [30].

X-exposure did not deteriorate the flexural strength of either SZ or NSZ series samples (Figure 14). It can be concluded that short-term spraying used in the exposure, followed by long-term light exposure combined with an elevated temperature (Table 2), does not exert such a significant influence on the NFPC's strength as the exposure including longterm wetting. The results after X-exposure can even suggest that exposure to elevated temperature (BST 60 ◦C—see Table 2) could result in plastification of the polymer matrix and its better surrounding by the filler, and hence improvement in the interphase bonds [48]. A decrease in the flexural strength after X-exposure was observed in most of the previous papers, reaching 20–25% [27,30,31]. Still, it has to be pointed out that most of the papers concerned composites with HDPE matrix, which is less resistant to UV than PVC [13,32]. A lack of significant changes in the flexural strength corresponds to the results of brittle fracture analysis, which did not reveal any changes in the composite's microstructure as a result of weathering (Figure 8).

Flexural modulus is another mechanical property analysed in the study. It depicts the material's stiffness, which is a key feature for construction products installed with point support, e.g., on a grid, as happens with terrace and facade profiles [13]. The flexural modulus' value level determines the profiles' susceptibility to deformation under service loads [22]. The solutions examined in the study achieved the flexural modulus values of 3970 MPa in the original condition (Figure 15). As shown in a previous study, construction profiles made of composite with HDPE matrix and wood flour are characterised by the flexural modulus of 2530 ÷ 3600 MPa [36,38]; with PP matrix and wood flour—ca. 4500 MPa [54]; and with HDPE matrix and sisal and bamboo fibres—2500 MPa and 3700 MPa, respectively [52].

**Figure 15.** Results of flexural modulus tests of brushed (SZ) and non-brushed (NSZ) samples in basic state (laboratory conditions), after F-exposure for 336 h (F-336) and 2016 h (F-2016) and after X-exposure for 300 h (X-300) and 2016 h (X-2016). The error bars represent the standard deviation (series size n = 12).

The performed ageing procedures exerted a significant influence on the flexural modulus values (Figure 16). A decrease was observed in all tested series after the exposure. Similarly to flexural strength, increasing the time from 336 h for F-exposure and 300 h for X-exposure to 2016 h does not cause a statistically significant difference in the flexural modulus. The difference in the modulus of elasticity between SZ-X-300 and SZ-X-2016 is the exception for which the changes are noticeably lower than after 300 h of X-exposure.

**Figure 16.** Diagrams showing differences in the flexural modulus Ef, MPa, after F-exposure and X-exposure for time T1 and T2: (**a**) brushed sample (SZ), (**b**) non brushed sample (SZ). The error bars show standard deviation (series size n = 12). The tables under the diagrams summarise the relative change in the modulus of elasticity ΔEf, %, for exposure time (T1 − T0) (T2 − T1) and (T2 − T0), calculated according to Formula (5). Statistically insignificant differences ΔEf are highlighted in grey.

The influence of F-exposure on brushed profile samples was most significant. The modulus of elasticity amounted to 3270 MPa (SZ-F-336) and 3160 MPa (SZ-F-2016) after F-exposure. For unbrushed samples, the values reached 3640 MPa (NSZ-F-300) and 3520 MPa (NSZ-F-2016). Similarly to flexural strength, exposing the lignocellulosic fibres during brushing could play a decisive role [7,31]. The reduction in the interphase interaction on the composite' surface can determine the value of the modulus of elasticity much more than the flexural strength. Weakening of the top layer significantly increases susceptibility to strain [9,53]. Moreover, X-exposure reduced the modulus of elasticity's value, whereby the non-brushed sample series revealed more significant differences than the brushed ones. The results correspond to the results of experiments performed for profiles with HDPE matrix and wood flour filling [28], although the drops discovered in this study are much smaller.

Charpy impact strength is another mechanical property taken into account in the study (Figure 17). Because of the high risk of construction profiles' exposure to dynamic loads throughout their entire life, stable impact strength value expressing the material's susceptibility to fracture can be considered one of the key functional parameters.

**Figure 17.** Results of Charpy impact strength of brushed (SZ) and non-brushed (NSZ) samples in basic state (laboratory conditions), after F-exposure for 336 h (F-336) and 2016 h (F-2016) and after X-exposure for 300 h (X-300) and 2016 h (X-2016). The error samples represent the standard deviation (series size n = 8).

The difference between the samples' impact strength before and after F-exposure and X-exposure seems significant, but because of the dispersion of the results in each test series, the statistical significance for some of these changes cannot be confirmed. It applies, especially to SZ samples. An anomaly is observed for NSZ samples, involving a significant increase in the impact strength after F-336 exposure. After F-2016 exposure, the impact strength decreases significantly compared to F-336 exposure. The final impact strength change between the initial value and the value after F-2016 exposure is not statistically significant, although it amounts to over 6%.

The general trend observed for the change in the mechanical properties after weathering (Figures 14, 16 and 18) is a statistically significant change after weathering time T1. However, in most cases, the difference between T1 and T2 is minor. The course of the Charpy impact strength changes for NSZ under F-exposure is the only exception.

**Figure 18.** Diagrams presenting differences in Charpy impact strength, auC, kJ/m2, after F-exposure and X-exposure for time T1 and T2: (**a**) brushed samples (SZ), (**b**) non brushed samples (NSZ). The error bars represent the standard deviation (series size n = 8). The tables under the diagrams present the relative change in the impact strength ΔacU, %, for exposure time (T1 − T0) (T2 − T1) and (T2 − T0), calculated according to Formula (5). Statistically insignificant differences ΔacU are highlighted in grey.

In order to comprehensively evaluate the exposure type influence on the change in the mechanical properties Δ*Y*, the following equation was used:

$$
\Delta Y = 100\% \cdot \frac{Y(F\_i, T\_j) - Y(X\_i, T\_i)}{Y(T\_0)} \tag{7}
$$

where: *F*, *X*—exposure type (according to Table 2), *Y*(*F*,*Ti*), *Y*(*X*,*Ti*)—value of the mechanical property after F- and X-exposure in time *Ti Tj* . Exposure times: *T*0—zero hours (before exposure), *T*1—336 h for F-exposure and 300 h for X-exposure, *T*2—is 2016 h for both exposures. The analysis results of the exposure type's influence on the property changes are

summarised in Table 5. All applied exposure types and their times are compared.


**Table 5.** Differences between the mechanical properties after F- and X-exposure in time T1 and T2 according to Equation (7), %. Statistically insignificant differences are highlighted in grey.

The analysis of the data summarised in Table 5 indicates that in five cases (except for the impact strength for SZ samples), the difference between the mechanical properties after short-term F-exposure and X-exposure (F-336 and X-300) and after long-term F-exposure and X-exposure (F-2016 and X-2016) is statistically significant. The mechanical properties of the SZ surface material revealed the highest drop after F-exposure (negative values in the Table), while for the NSZ surface, it was after X-exposure (positive values in the Table). Hence, it can be concluded that for the SZ surface, more unfavourable changes can be expected after F-exposure, while for NSZ surfaces, it occurs after X-exposure. Both the exposure type and duration do not significantly contribute to the change in the Charpy impact strength for SZ samples.

On the other hand, an absence of a significant difference between the mechanical properties after F-exposure for 2016 h and X-exposure for 300 h can be observed. The modulus of elasticity for SZ samples makes the only exception, where significant differences in the modulus' value can be observed between all exposure types. Major changes in the modulus (lower modulus values after exposure) are caused by F-exposure, while ranking the mechanical property values after both exposures (from the highest to the lowest value of the modulus), we get X-300; X-2016; F-336 and F-2016.

#### **4. Conclusions**

An analysis of the experimental data collected under the study suggests that exposing construction profiles made of PVC composite with wood flour filling to light emitted by different laboratory sources of light alternately with wetting causes degradation of their usable surfaces. An SEM analysis of the surface microstructure revealed molten areas in the polymer matrix outermost layers and exposed surfaces of the filler fibres. The degree of the changes can be considered as significantly reducing the profiles' aesthetic and decorative properties. No microstructure changes were observed in the brittle fracture tests. However, the weathering procedure was discovered to impact the surface morphology. Influences involving irradiation with a xenon lamp and short-term wetting (X-exposure) caused much more significant surface degradation than exposure to fluorescent lamp's light and longterm wetting (F-exposure) for the same exposure duration. The observations applied only to the mechanically non-treated usable surfaces. The observations were not carried out for brushed surfaces, because of the surface roughness.

The applied exposures affected the mechanical properties. The influence of surface treatment on changes in mechanical properties during weathering was not statistically analyzed due to the different mechanical properties of SZ and NSZ profiles before weathering. In the drawings, however, differences can be observed were observed in the susceptibility to the exposure for profiles with mechanically treated usable surfaces (brushed—SZ) compared to non-brushed (NSZ) ones. Artificial weathering carried out with fluorescent lamps and long-term wetting, included by condensation (F-exposure), greatly influenced the brushed profiles' properties. In turn, the influence of artificial weathering by exposure to a xenon lamp and short-term wetting was more significant for the non-brushed profiles. Significant changes in the flexural modulus were observed after the exposures, especially after F-exposure. No significant decrease in the flexural strength occurred, and the impact strength changes are hard to assess because of the dispersion of the results in each series.

The exposure duration (time) affected the properties of interest. Although extending the weathering exposure time from 300 h (X-exposure) or 336 h (F-exposure) to 2016 h significantly aggravated the surface morphology changes, especially after X-exposure, the changes in the mechanical properties observed at the initial stage of ageing progressed only slightly.

The constancy of mechanical properties matters for the building fitness assessment. Based on the collected data, it can be concluded that for profiles made of natural fibrereinforced polymer composites, whose usable surface was developed in a standard way, by mechanical treatment (brushing), artificial ageing using fluorescent lamps and long-term wetting including condensation (F-exposure) seems to be the adequate procedure to assess the changes in the mechanical properties. The procedure causes more severe swelling of lignocellulosic fibres and weakens the interaction forces between the matrix and the filler, reducing the mechanical parameters. The influence was particularly evident for the

flexural modulus, which should be considered as a suggestion to select this parameter as a diagnostic feature of resistance to accelerated weathering. It needs to be emphasised that the changes in the modulus of elasticity can be determined only after short-term F-exposure (336 h), which can be used for quick diagnostics of new solutions.

This study does not exhaust the topic of artificial weathering resistance test methods for construction profiles made of plant fibre-reinforced polymer composites. Considering the dynamic development of this product group and its growing significance in civil engineering, further studies are planned. Future studies will cover other NFPC compositions and extended weathering exposure time.

**Author Contributions:** Conceptualisation, E.S. (Ewa Sudoł); Methodology, E.S. (Ewa Sudoł), E.K. and E.S. (Ewa Szewczak); Formal analysis, E.S. (Ewa Sudoł); Investigation, E.S. (Ewa Sudoł), E.K. and E.S. (Ewa Szewczak); Writing—original draft preparation, E.S. (Ewa Sudoł), E.K. and E.S. (Ewa Szewczak); Writing—review and editing, E.S. (Ewa Sudoł); Visualization, E.S. (Ewa Sudoł) and E.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Ministry of Education and Science as part of the project NZM-058/2020.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Acknowledgments:** Special thanks to Marcin Kupisz and Cezary Str ˛ak for technical support.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the study's design, in the collection, analyses, or interpretation of data, in the writing of the manuscript, or in the decision to publish the results.
