The Role of Abrasion Resistance in Determining Suitability of Low-Density Plantation Timber for Engineered Flooring

Abstract

Abrasion resistance is an important property for the functional performance and serviceability of timber floors. Although hardness is the conventional criterion used in selecting species for flooring applications, it shows greater variations and restricts the use of low-density species, whereas abrasion resistance could generate a more reliable indication of a product’s surface performance. Eucalyptus nitens is a fast-grown global plantation species extensively available in Tasmania, Australia. Until recently, this material has been perceived as unsuitable for appearance applications such as flooring. This study assesses several engineered flooring prototypes comprised of E. nitens—sawlog managed and fibre-managed resources—compared to an existing market product (E. obliqua and a commercial engineered timber flooring product with UV-cured coating). Tests were performed in accordance with the EN 14354:2016, sandpaper method using Taber abraser and further modified to test flooring prototypes. The highest abrasion resistance was observed in the E. nitens veneer composite product. Fibre-managed E. nitens resulted in the greatest level of abrasion, while sawlog-managed E. nitens was comparable to native regrowth E. obliqua, a commonly used flooring species historically used in Australia. Therefore, the findings from this research suggest there are suitable flooring applications for plantation E. nitens as engineered wood products in some domestic and residential dwellings when compared to existing native products.

1. Introduction

The wearing and abrasion of a material can be described as the progressive loss of substances from its surface due to mechanical loads [1]. The functional performance of a timber floor can be significantly influenced by its resistance to wear [2,3,4]. Wear is usually caused by the abrasion action caused by traffic moving over the floor, which is intensified by abrasive components such as dust, sand and grit [5]. The evaluation of the wear resistance of timber floors is challenging, as wear is the result of many uncontrollable variables [6,7,8]. According to Armstrong [6], hardwoods show great variations in behaviour under in-service traffic conditions; some species retain surface smoothness while others tend to develop roughness on surface fibres. Hence, laboratory testing approaches of abrasion properties of timber flooring were suggested as early as 1937, adapting methods used for evaluating materials such as tiles, concrete and stone [6]. Youngquist and Munthe [9] report conducting abrasion resistance testing on timber using a US Navy wear test machine. This testing machine was primarily designed for the testing of walkway materials for ships, and the wear resistance of teak (Tectona sp.) timber and modified timber materials. It was later accepted to use the testing machine to evaluate the wear resistance of solid timber. The results were acceptable in terms of the general conception of species’ performance. These are the origins of ASTM D 2394 [10], which provides guidelines for the testing of finish films on timber and timber-based floor surfaces using a Navy-type wear tester machine in the US. ASTM D 2394 suggests comparing the results against the values reported in 1948 for timber species by Youngquist and Munthe [9]. In addition, the Armstrong [6] report contributed to the guidelines by classifying common timber species for different levels of traffic exposure, based on qualitative analyses of the surface and quantitative mass loss caused by abrasion. The abrasion test adapts the use of a planning machine originally designed to test metal.

Although early data on the abrasion testing of timber is available, a specific method for testing the abrasion resistance of solid timber is still lacking [2,5,11]. As a consequence, the Taber abraser method, as referred to in EN 14354, EN 438-2, ASTM D 1044, ASTM D 4060 and ISO 9352, is occasionally used to determine timber abrasion resistance in comparative studies [2,5,12,13]. However, the Taber abraser method was originally developed to determine the abrasion resistance of high-pressure laminated papers/wood veneer floor coverings with coatings. It should be noted that different adaptations of abrasion tests are reported in the literature to be conducted on solid timber, as shown in Table 1, since most standards address materials with coated surfaces. Therefore, conclusions are generally reached by comparing an unknown species against a control species with known performance capacity subjected to the same method of testing.

With the depletion of and restricted access to high-density native forest timbers worldwide, fast-grown plantation Eucalypts can be a potential alternative resource for the flooring market due to their high availability, material properties and price [15]. Despite the increasing demand for timber flooring in the world market, only a few traditionally used hardwood species with known performance are still commonly used for flooring applications [15,16]. For many timber applications, the reason for using a certain species is generally inexplicable or non-quantifiable, and usually governed by tradition [17]. To diversify the timber species used for flooring requires an understanding of the material properties so as to determine its in-service behaviour for a specific end-use application. In Australia, E. nitens is widely used as a plantation species [18], particularly in Tasmania, due to the species’ ability to withstand frost and low mean annual temperatures [19,20]. The majority of E. nitens plantations in Tasmania were intended for pulpwood production. These are known as fibre-managed plantations intended to be harvested around 15 years of age (density range 429–669 kg/m³ [21]) without being subjected to thinning or pruning silviculture [22]. Alternatively, sawlog-managed plantations undergo silvicultural practices and are harvested around the age of 25 years (density range 440–755 kg/m³ [23]), and are thinned and pruned for the appearance sawn timber market [22,24]. In this paper, E. nitens is distinguished by the terms fibre E. nitens and sawlog E. nitens to reflect the differences. Due to the lower densities and presence of natural features such as knots in fibre-managed plantation timber, the resource does not comply with Australian standards for appearance grading [25,26], and requires an understanding of its specific properties [27,28,29]. This study is conducted to determine the potential of utilising plantation E. nitens in engineered timber flooring products, facilitating the use of this emerging resource for a higher-value application instead of being converted to wood chips for offshore processing, and thus supporting the local timber industry in manufacturing a more local product. The study also captures the specific properties of a native Tasmanian species, E. obliqua (density range 510–820 kg/m³ [23]), which forms part of the widely known Tasmanian Oak used in Australian flooring applications. The authors have observed that decisions of flooring species choices in Australia are generally influenced by the knowledge regarding how these species have performed in-service for generations, and colour preferences [30,31]. For instance, Tasmanian Oak (a combination of E. obliqua, E. regnans and E. delegatensis) has been used extensively for flooring applications, although the species falls in the lower range of hardness and density when compared with other commonly used Australian flooring hardwoods. Users are confident in the performance of Tasmanian Oak floors as a result of experience, in addition to its wide and local availability.

Hardness is typically used as a major criterion to qualify the performance of a timber species in flooring applications. Janka hardness and Brinell hardness methods are the most commonly used tests for this purpose [32]. In Australia, the Janka hardness test, conducted according to ASTM D.143 [33], is used to determine the hardness property of timber [34]. As part of this project, a regime of engineered timber flooring prototypes consisting of plantation E. nitens top layers was developed. Two of the prototypes consisted of 6 mm-thick top layers of sawlog E. nitens and fibre E. nitens laminated onto a 6 mm-thick commercial marine plywood substrate. The other prototype had a 1.2 mm-thick sawlog E. nitens top layer laminated onto a 12 mm-thick locally produced E. nitens plywood substrate. As reported in Millaniyage, Kotlarewski [23], Janka hardness tests per ASTM D.143 were conducted on prototypes in comparison with a 12 mm-thick solid E. obliqua flooring and a commercially engineered flooring product with a 3 mm-thick Tasmanian Oak top layer and an 11 mm-thick rubberwood backing. The results show that the prototype with a 1.2 mm-thick top layer achieved the highest Janka rating because of the impact of its high-density substrate. These observations were supported by Grześkiewicz, Kozakiewicz [35], suggesting that using the hardness rating of the top layer to predict the performance of an engineered timber flooring product might not always be accurate due to the multilayer compositions of the products, and the dependence on the thicknesses of each layer. Therefore, the authors identified the importance of assessing alternative material properties, such as abrasion resistance, to quantify the surface performance of engineered flooring and thus validate context-specific applications. However, the abrasion resistance properties of these timbers have not been quantified to the authors’ knowledge, as hardness and density have been key factors in flooring species selection up to now. This study was conducted on timber species and composite engineered flooring prototypes to understand the abrasion resistance properties of each configuration.

Therefore, the intent of the present study is to generate a body of new knowledge on the abrasion resistance of plantation E. nitens in comparison with a commonly used flooring species, E. obliqua, and a commercially available Tasmanian Oak engineered flooring product using a modified Taber abraser method (EN 14354). The abrasion test results obtained in the present study have been compared with the Janka hardness values reported by Millaniyage, Kotlarewski [23] for the same engineered flooring prototypes to identify the potential for using abrasion resistance as a more reliable indicator to quantify the surface performance of the engineered timber flooring. Determining the abrasion resistance of these materials facilitates the exploration of new applications and increasing value addition for the lesser-known plantation timber, encouraging the timber industry to diversify its product range.

2. Materials and Methods

2.1. Material Origins

Abrasion tests were conducted on five types of specimens: three solid timber specimens and two engineered flooring products, as shown in Table 2. Timber products were supplied from three different sawmills located in the North of Tasmania:

• Sawmill A provided the timber boards for sawlog E. nitens and regrowth E. obliqua, as well as veneers. E. obliqua is the dominant Eucalyptus species of the three and is marketed as Tasmanian Oak (E. obliqua, E. delegatensis and E. regnans). E. obliqua is commonly used for flooring, and is harvested from native and regenerated forests [22];

• Sawmill B processed the fibre-managed E. nitens boards;

• Sawmill C locally manufactured an E. nitens plywood, which was used as a substrate in one of the engineered flooring prototypes. This prototype included an E. nitens veneer (Sawmill A) on top. The E. nitens veneers were randomly selected from a commercial production line specialised in producing 0.6 mm veneers. In order to understand the potential of using veneers, and to test their performance in flooring and their in-state capabilities, the developed engineered flooring prototype consisted of a 1.2 mm-thick top layer, which was obtained by laminating two 0.6 mm veneers together. The veneers were produced by slicing, and were laminated onto an E. nitens plywood substrate by hot-pressing with a polyvinyl acetate (PVA) glue.

The material was used to develop the engineered flooring prototypes consisting of 6 mm-thick top layers of sawlog E. nitens and fibre E. nitens and 1.2 mm-thick sawlog-managed veneer top layer. The prototypes were developed based on interview insights gained from several architects, timber specifiers and flooring manufacturers familiar with using Tasmanian timbers in their projects (Human ethics approval H0026194). The prototypes were tested and compared with solid E. obliqua flooring, and a commercial Tasmanian Oak engineered flooring product. This existing flooring product was prefinished with a UV-cured coating system. The prototypes were subjected to Janka Hardness tests in their product form, as reported by Millaniyage, Kotlarewski [23], while the top 5 mm of each prototype was subjected to abrasion testing in the present study. As the prototypes with 6 mm top layers and E. obliqua panels consisted only of solid timber when dressed to 5 mm thickness, these are referred as solid specimens in the present study.

2.2. Abrasion Resistance Test Using Taber Abraser Method

The resistance to abrasion was determined according to Taber abraser test EN 14354 [36], using the sandpaper method. Since the standard addresses the abrasion resistance of lacquered surfaces, a modified methodology was adapted aligning with Brischke, Ziegeler [5] to accommodate the testing of solid timber. For solid timber and veneer specimens, 30 specimens from each type were tested and 10 specimens were tested from the engineered Tasmanian Oak product. Each specimen was obtained from a single board to get a wider representation of material properties. The floor panels used in the study were dressed to 5 mm thickness at the workshop and a hole saw was used to obtain round timber discs with a diameter of 100 mm to facilitate the fixing of the specimen to the Taber abraser machine (Figure 1).

The specimen’s surfaces being subjected to abrasion testing were set in a random orientation including both radial and tangential sections. The intent was to generate a combined variation in both surfaces. After dressing, the specimens were allowed to stabilise for two weeks. For the solid timber materials used, density specimens were also prepared to calculate the equilibrated air-dried density (ADD), and oven dried at 103 ± 2 °C until constant weight to obtain the oven-dry density (ODD). The moisture contents (MCs) and densities were determined in accordance with Australian/New Zealand Standards as per Equations (1)–(3) [37,38]. For the engineered flooring specimens with substrates, the ADD values were calculated, as they consisted of commercially available products dried and processed to industry standards.

Calculation of MC:

(1)$MC=\frac{mi-mo}{mo}\times 100$

where MC—the percentage moisture content of the specimen, mi—initial mass of the specimen, mo—oven-dried mass of the specimen.

Calculation of ADD:

(2)$ADD=\frac{mw}{aw\times bw\times lw}=\frac{mw}{vw}$

where mw—the mass (kg), aw—the width (m), bw—the depth (m), lw—the length (m) and vw—the volume (m³) of the tested specimen at moisture content w.

Calculation of ODD:

(3)$ODD=\frac{m0}{a0\times b0\times l0}=\frac{m0}{v0}$

where m0—the mass (kg), a0—the width (m), b0—the depth (m), l0—the length (m) and v0—the volume (m³) of the tested specimen at moisture content w.

Each timber disc was weighed, and the thickness was measured at four points immediately before subjecting to the abrasion test. The discs were clamped to the Taber Abraser and the abrasion test was conducted using S-42 sanding paper attached to rubber wheels rotating at 60 cycles per minute. After 1000 revolutions, the loss of mass was recorded and the decrease in thickness due to abrasion was measured at the same four points previously measured. The loss of thickness due to abrasion was calculated as a percentage for each specimen per Equation (4), and an average for each species/product was determined.

(4)$∆t=\frac{t0-t1}{t0}\times 100$

where ∆t—abrasion percentage, t0—thickness before abrasion (mm), t1—thickness after abrasion (mm).

The abrasion results were compared with the Janka hardness values reported in Millaniyage, Kotlarewski [23] for the tested prototypes. The Janka hardness test was conducted as specified in ASTM D.143 using a Hounsfield universal testing machine. The specimens (dimensions: product thickness × product width × 150 mm in length) were subjected to the projecting of a steel ball with a 11.284 mm diameter at a speed rate of 6 mm/min and the load required to penetrate the ball into half of its diameter (5.642 mm) was recorded as the relevant Janka hardness rating.

2.3. Statistical Analysis

A randomised experimental design was conducted, and the abrasion data were analysed with 1-way ANOVA test using R statistical software. Statistical significance was determined at a p value less than or equal to 0.05.

3. Results and Discussion

A summary of the moisture content (MC) and density (ADD and ODD) of the solid and composite timber flooring specimens tested is shown in Table 3. Accordingly, the MC values for all solid timber specimens were in the range of 9% to 13% at the time of testing. Among the three types of solid timbers tested, E. obliqua had the highest density (ODD = 615 kg/m³), followed by sawlog E. nitens (ODD = 570 kg/m³). Fibre-managed E. nitens had the lowest density (ODD = 445 kg/m³). The E. nitens veneer product had the highest ADD of 885 kg/m³ in comparison with all the tested samples. Figure 2 shows the surfaces of samples subjected to the Taber abrasion test.

Figure 3 shows the results of the abrasion tests conducted both on solid and engineered flooring specimens. The relative order of abrasion resistance in descending order is: E. nitens veneer engineered flooring (EV), E. obliqua (O), sawlog E. nitens (SN), Tasmanian Oak engineered flooring (TO), and fibre E. nitens (FN).

The summary statistics of the abrasion resistance of specimens tested as a percentage of thickness reduction when subjected to abrasion in a Taber abraser for 1000 revolutions are shown in Table 4. Among the three types of solid timber specimens tested, fibre E. nitens with the lowest density showed the highest abrasion at 4.94%, while the sawlog E. nitens abrasion was 3.90%. E. obliqua (3.55%) reached a value lower than both types of plantation E. nitens.

There is strong evidence that the mean abrasion values differed between the five tested groups (ANOVA, F_3,116 = 11.79, p ≤ 0.001). In accordance with Table 5, post-hoc pairwise comparisons using Tukey’s test show significant differences between fibre E. nitens in comparison with sawlog E. nitens, E. obliqua and E. nitens engineered flooring products. No differences in the average abrasion ∆t% values occurred for the seven pairs of specimens (SN–EV), (TO–O), (SN–EV), (O–EV), (TO–EV), (TO–SN) and (TO–FN). However, fibre E. nitens did not show significant differences with the Tasmanian Oak engineered flooring product. Yet, the results generate a fair comparison of how all the specimens might behave when subjected to similar levels of abrasive action. The tested pairwise comparisons were grouped to assess the impacts of different top-layer species and different thicknesses of the same species, as shown in Table 5.

The assessments of groups show that the top layer species had a significant impact on abrasion resistance, as pairs O–FN (p value ≤ 0.001 *) and SN–FN (p value = 0.005) showed significant differences. In terms of thickness differences, the 5 mm-thick E. obliqua and 3 mm-thick Tasmanian Oak did not show significant difference (TO–O, p value = 0.339). Although the fibre E. nitens and sawlog-managed E. nitens veneer product showed significant differences (FN–EV, p value ≤ 0.001 *), the 5 mm-thick sawlog-managed E. nitens did not show a significant difference from the sawlog-managed 1.2 mm-thick veneer product (SN–EV, p value = 0.522). This indicates that lower top layer thicknesses did not reduce the abrasion resistance of engineered composites when compared with solid specimens of the same species. In contrast, E. nitens engineered flooring specimens with 1.2 mm thickness showed higher abrasion resistance than the solid sawlog-managed E. nitens samples.

It can be observed that the E. nitens veneer product with an E. nitens plywood substrate resulted in the lowest thickness loss due to abrasion, and none of the samples penetrated into the substrate. This prototype consisted of two 0.6 mm-thick veneers laminated together to form the top layer. In order to further investigate this, five more samples with only one layer of 0.6 mm-thick veneer laminated onto E. nitens plywood substrate under the same production process were subjected to abrasion tests. The abrasion results obtained range between 2.13% and 5.96%, with an average value of 4.64% for the five samples. The lower abrasion observed in the 1.2 mm top layer might have been caused by the glue penetrating into the lower layers of the veneer. Although an average higher abrasion was observed in the 0.6 mm top layer, none of these samples penetrated to the substrate, confirming the observations of the veneer flooring prototype.

Figure 4 shows the results obtained in the present study in comparison with some species reported by Brischke, Ziegeler [5], Wentzel, González-Prieto [14] following similar testing methodologies. The thickness reductions due to abrasion obtained in the present study for both Tasmanian plantation sawlog E. nitens (3.90%, 570 kg/m³, age 27 years) and fibre E. nitens (4.94%, 445 kg/m³, age less than 20 years) are higher than the values published for fast-grown plantation E. nitens in Chile (3.21%, 617 kg/m³, age 19 years) and in Spain (3.06%, 574 kg/m³, age 16 years) by Wentzel, González-Prieto [14]. It should be noted that the densities reported in both Spain and Chile are higher than those of the Tasmanian E. nitens.

As observed in similar studies [2,5], the results in this study suggest that species with lower densities tend to show more abrasion (Figure 5). Although strong correlations were not observed, the best relation between density and abrasion was observed in sawlog E. nitens (R² = 0.2081) in comparison with fibre E. nitens (R² = 0.0299) and E. obliqua (R² = 0.0004).

However, the literature also suggests that the anatomical differences between materials strongly correlate with their resistance to abrasion [39,40]. For instance, Brischke, Iseler [11] reported that although the abrasion resistance of a timber species generally increases with higher densities, exceptions such as Douglas fir (Pseudotsuga menziesii Franco) have shown high abrasion resistance levels at a moderate density due to the positive impact on abrasion resistance caused by the regular alternation of earlywood and latewood in the species. Another study conducted on thermally and chemically modified timber suggests that the cell wall modification of timber might yield a significant and mostly negative impact on the abrasion resistance of wood as a result of increased surface brittleness [5]. Similarly, Khademibami, Shmulsky [1] reported that Hickory (Carya sp.) with a density of 801 kg/m³ showed higher thickness loss due to abrasion when compared with Red Oak (Quercus sp.) (731 kg/m³) and White Oak (Quercus sp.) (760 kg/m³). Therefore, it should be noted that the abrasion resistance of timber can vary considerably based on density, rate of growth, proportion of sapwood/latewood, origin and position in tree, and microstructural changes. The present study conducted a macroscopic characterisation of the timber available on the market with a view to understanding its abrasion properties, and therefore consideration of the above aspects is not within the scope of this research.

The choice of a suitable flooring species is primarily governed by its appearance and cost [6,31], especially where the in-service conditions are of domestic/light traffic exposure. Yet, the ability of the timber to withstand abrasive action is of importance in the long-term performance of a floor. Although in-service tests could provide an idea of how the timber might behave in a similar situation, these tests have many uncontrollable variables, such as the number and weight of people walking over the floor, the types of shoes worn, and the amount and size of abrasive elements [3], and it takes considerable amounts of time to generate results. For instance, Armstrong [6] reported conducting an in-service trial over nine years in the subway of a London Underground Station, while Németh, Molnar [3] conducted a similar trial on the mezzanine of a dormitory of the University of West Hungary for five years. Armstrong [6] is one of the earliest reported studies that identified the importance of conducting laboratory analysis to determine the abrasion resistance of solid timber, both qualitatively by analysing the surface, and quantitatively via mass loss. According to Armstrong [6], Tasmanian Oak can be categorised as suitable for light traffic applications, such as residential and domestic buildings, flats, small classrooms and offices. Due to the lack of an existing standard to determine the abrasion resistance of solid timber, the Taber abraser method is commonly used for this purpose. In the Taber abraser method, the timber specimens are subjected to cutting actions caused by the aluminium oxide abrasive in the sandpaper. The scratched surfaces of two samples tested are shown in Figure 6.

It should be noted that the loads caused by the Taber abraser tests with rotating sandpaper under a specific load do not necessarily replicate the exposure conditions of a floor [5]. In general, the qualitative aspects of a species’ response to abrasion, such as surface roughening, splintering or surface breakdown, can be observed through an in-service test. Quantitative thickness loss in current laboratory tests may allow for comparisons of species with similar anatomical features. Although high-density species tend to give higher quantitative abrasion resistance results, they might show inferior behaviours when in service, depending on their anatomical structure [6]. In addition, some species may show higher abrasion resistance due to the presence of natural volatiles and extractives, while the presence of brittle or hard resinous substances may increase the abrasion rate [6]. Hence, developing a reliable methodology to test the abrasion resistance of solid timber would be useful when introducing novel timber species into the flooring market.

The industry benchmarks typically employed to determine the suitability of a timber species for flooring applications are density and hardness [32,35,41]. In Australia, the Janka hardness test is used for this purpose, although the test results may be less reliable due to the multiple layers used in an engineered timber flooring product [23]. See Millaniyage, Kotlarewski [23] for detailed information about the Janka hardness tests conducted on the engineered flooring products in the present study, including Tasmanian plantation E. nitens and control specimens. The tests were conducted in accordance with ASTM D. 143 [33]. Table 6 compares the results of the present study with the Janka hardness values of the prototypes tested flooring by Millaniyage, Kotlarewski [23].

Overall, there is similarity between the Janka hardness results of the prototypes and the abrasion results of the specimens. The only exception is the Tasmanian Oak prefinished flooring product; a higher Janka hardness value was observed in the Tasmanian Oak prefinished product than in the sawlog E. nitens top layer, which was not seen in the abrasion test. Although density is closely related to hardness, the Janka test is not a suitable criterion to determine the abrasion resistance [6]. This is due to the relatively deep penetration of the steel ball used in the Janka hardness test, producing a gross hardness value that does not offer a reliable measure of the actual hardness of the wearing surface. As observed by Millaniyage, Kotlarewski [23], the Janka hardness values of top layers of 6 mm thickness were more dependent on the top layer species, while when the veneer product penetrated into the substrate, this resulted in a higher Janka hardness value. Therefore, the Janka hardness values for engineered timber flooring products might not always reflect their in-service performance, as they have multilayer structures with limitations in terms of re-coating and re-sanding the top layer [15,42]. In terms of engineered flooring products, surface abrasion tests might be more reliable, since the top layer’s performance can be quantified, as observed in this study.

4. Conclusions

This investigation of the abrasion resistance of fast-grown plantation E. nitens of solid and veneer types in comparison with that of commonly used E. obliqua and Tasmanian Oak prefinished engineering flooring indicates that the abrasion resistance values observed in sawlog plantation E. nitens were similar to those of both controls used. Out of the five types of specimens tested, the E. nitens veneer product showed the lowest abrasion. Further research is required to understand if this product would show similar results when subjected to in-service testing over a longer period under general traffic exposure. The abrasion resistance of most Australian species commonly used in flooring applications have not been evaluated under laboratory conditions, or documented. The available literature suggests that abrasion is generally correlated with density, but in some instances, the anatomical characteristics might impact the abrasion, causing some lower-density species to behave well when subjected to abrasion. The observations in the present study reflect a positive relationship between abrasion resistance and density. The preliminary observations of this study show that fast-grown plantation E. nitens have the potential to be used in purpose engineered flooring applications aimed at domestic/light commercial end use.

The functional suitability of a species used for flooring applications under most in-service conditions depends considerably on its wearing surface. Despite a comparatively high measure of abrasion loss, the floor would be acceptable to consumers as long as the surface remains reasonably smooth and aesthetically pleasing [6]. The selection of a species to be used for flooring should be based on its intended end-user application. For normal traffic conditions, such as domestic/light commercial, most hardwoods will provide satisfactory performance when given proper maintenance, as observed in Armstrong [6] long-term in-service trials, where relatively lower quantitative abrasion losses were observed. However, the characteristics of the native high-density timber species tested in these early studies are different to those of fast-grown plantation and regrowth timbers entering the current market. Therefore, it is important to understand the properties of this emerging resource, with the intention of introducing it for use in novel value-added products. With restricted access to high-density native timbers, it is important to educate consumers to make sustainable choices in terms of timber species that are suitable for their intended end-use. With the emergence of new engineered timber flooring products, it is important to acknowledge that the limited thickness of the top layer will restrict the capacity for the re-coating and re-sanding of the product. Traditional hardness testing methods will not address the in-service performance of multilayer composites, as the test results are impacted by the thickness of each layer and the materials used. Abrasion resistance would give a better picture of the wearing surface, and requires a consensus methodology for solid timber testing. The lack of a standardised method to evaluate the abrasion resistance of solid timber flooring restricts the comparison of values published in the literature. Most of the current abrasion resistance testing methods only focus on quantitative analyses of the surface thickness/mass loss. However, since timber flooring is an appearance application, the assessment of qualitative characteristics is also of importance. Therefore, improving the present testing methods in order to assess the abrasion resistance of solid timber by incorporating both quantitative and qualitative analyses and specified thresholds differentiating the requirements for specific end-user applications could facilitate the introduction of novel timber flooring products to the market.

The observations of the study show that abrasion resistance can be used as a more reliable indicator in assessing the surface performance of an engineered flooring product in comparison with traditional hardness tests. The lower-thickness top layers in engineered flooring products generate hardness values related to the characteristics of the substrate, suggesting that the current market practice of specifying the performance of an engineered flooring product based on the top layer species might not always be effective. In such instances, abrasion testing can be used as an alternative method to validate context-specific performance, thus overcoming the restrictions caused by the lower hardness values observed in low-density, fast-grown plantation timber.

The abrasion resistance of timber is correlated with its microstructural properties, and therefore assessing these properties at the cellular level via sawing orientation is identified as an area for future engagement. In this area, improving the abrasion resistance of plantation timbers through surface finishing or modifications is also identified as an important area to be researched in future studies.

Author Contributions

Conceptualisation, K.M.; methodology, K.M., N.K., A.T. and L.W.; formal analysis, K.M.; investigation, K.M.; writing—original draft preparation, K.M.; writing—review and editing, K.M., N.K., L.W. and A.T.; supervision, N.K., L.W. and A.T. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

Data will be made available on request.

Acknowledgments

The National Institute for Forest Product Innovation (NIFPI) is acknowledged for funding this research (Grant number: NT015/NIF079). Britton Timbers, Neville Smith Forest Products and CUSP Building Solutions are acknowledged for providing timber materials and enabling prototype development. Stuart Meldrum and Malcolm Liehr are acknowledged for their technical assistance and support.

Conflicts of Interest

The authors declare no conflict of interest.

Footnotes

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.

Figures and Tables

Figure 1. Set up of the Taber abrasion test used in the study.

Figure 2. Specimens after Taber abrasion test (SN—sawlog E. nitens, FN—fibre E. nitens, EV—E. nitens veneer on E. nitens plywood, O—E. obliqua, TO—Tasmanian Oak engineered product).

Figure 3. Abrasion as thickness loss% of tested timber specimens.

Figure 4. Comparing the abrasion resistance of tested species with values published in the literature (Red columns show the results of the present study, green columns show results for plantation E. nitens from a different geographical area [14], blue columns show results of similar testing on two other species [5]) * Approximate values obtained from the graph illustrated in [5].

Figure 5. Correlation between density and thickness loss% due to abrasion for tested solid timber species samples (FN: fibre E. nitens; O: E. obliqua; SN: sawlog E. nitens).

Figure 6. Scratched surfaces of timber specimens due to abrasion caused by the Taber abraser ((a) sawlog E. nitens; (b) E. nitens veneer)).

Table 1

Comparison of abrasion resistance studies (it should be noted that the values cannot be directly compared between studies due to the different methodologies used).

Reference	Brischke, Ziegeler [5]	Wentzel, González-Prieto [14]	Khademibami, Shmulsky [1]	Swaczyna, Kedzierski [12]	Knapic, Machado [2]
Standard	EN 438-2 High-pressuredecorative laminates (HPL)	EN 438-2 High-pressure decorative laminates (HPL)	ASTM D2394-17 Simulated Service Testing of Wood and Wood-Based Finish Flooring	PN EN ISO 5470-1 Rubber- or plastics-coated fabrics	EN 14354 Wood-based panels and Wood veneer floor coverings
Methodology summary	Samples subjected to abrasion with S-42 sanding paper attached to Taber abraser at 60 revolutions per min for 1000 revolutions	Samples subjected to abrasion with S-42 sanding paper attached to Taber abraser at 72 revolutions per min for 1000 revolutions	Specimen mounted on Navy-type wear tester and subjected to 500 rotations under a constant flow of 80 grit aluminum oxide media	Specimens subjected to 1000 revolutions of Taber abraser under 1000 g load	Specimens subjected to destructive action by passing loose aluminum oxide grit under a pair of leather clad brass wheels for 6000 revolutions
Measurement used	Thickness loss (%)	Thickness loss (%)	Thickness loss (mm)	Thickness loss (mm) and loss of mass (g)	Thickness loss (%)
Species abrasion (density kg/m3 in parentheses)	* European Beech (Fagus sylvatica L.) = 3.0% (725)* Scots pine sapwood (Pinus silvestris L.) = 5.7% (555)	Chile Plantation E. nitens = 3.21% (617)Spain Plantation E. nitens = 3.06% (574)	Sweetgum (Liquidambar styraciflua) = 0.234 mm (602)Hickory (Carya sp.) = 0.212 mm (801)Ash (Fraxinus sp.) = 0.185 mm (682)Red Oak (Quercus sp.) = 0.175 mm (731)White Oak (Quercus sp.) = 0.147 mm (760)	Oak (Quercus sp.) 0.18 mm (606)Fossil Oak (Quercus sp.) = 0.19 mm (595)Pine (Pinus sp.) = 0.28 mm (529)Cherry (Prunus sp.) = 0.17 mm (662) Sycamore Maple (Acer sp.) = 0.17 mm (545)Walnut (Juglans sp.) = 0.14 mm (553)Ash (Fraxinus sp.) = 0.19 mm (550)Mahogany (Swietenia sp.) = 0.23 mm (546)	Limba (Terminalia superba) = 13.5% (499)Mupanda (Brachystegia speciformis) = 3.5% (786)Rodhesian teak (Baikiaea plurijuga) = 2.5% (931)Cork Oak (Quercus suber L.) = 3% (783)

* Approximate values obtained from graph illustrated in [5].

Table 2

Composition of tested specimens and experimental design.

Specimen Type	Composition	No of Replicates
Sawlog E. nitens (27 years old)	Solid	30
Fibre E. nitens (below 20 years old)	Solid	30
E. nitens engineered flooring	1.2 mm-thick sawlog E. nitens veneer top layerFibre E. nitens plywood substrate	30
Regrowth forest E. obliqua	Solid	30
Tasmanian Oak engineered flooring *	3 mm-thick Tasmanian Oak top layerRubberwood (Hevea sp.) substrate	10

* 10 samples from control were tested for comparison.

Table 3

Moisture content and density of the tested hardwoods.

	Solid Sawlog E. nitens			Solid Fibre E. nitens			Solid E. obliqua			* E. nitensEngineered Flooring	* Tasmanian OakEngineered flooring
	MC%	ADD(kg/m3)	ODD(kg/m3)	MC%	ADD(kg/m3)	ODD(kg/m3)	MC%	ADD(kg/m3)	ODD(kg/m3)	ADD(kg/m3)	ADD(kg/m3)
Mean	11.3	630	570	10.8	495	445	10.3	680	615	885	640
SD	0.91	84.10	73.59	0.50	41.17	36.58	0.47	78.58	70.31	34.93	34.70
Min	9.6	460	415	9.6	445	400	9.5	535	485	800	575
Max	12.9	745	670	11.8	630	570	11.2	835	760	980	690

* Only ADD values are reported since the products were obtained and are commercially available.

Table 4

Abrasion test results of solid and engineered timber flooring specimens.

Specimen	Mean Abrasion ∆t%	Median	SD	COV%	SEM	Min	Max
E. nitens veneerengineered flooring (EV)	3.44	3.45	0.84	24.56	0.15	2.06	5.31
E. obliqua (O)	3.55	3.41	1.03	28.97	0.18	1.98	5.9
Sawlog E. nitens (SN)	3.90	3.77	1.17	30.03	0.21	1.9	6.63
Tasmanian Oak engineered flooring (TO)	4.33	4.03	1.77	40.88	0.55	1.75	7.66
Fibre-managed E. nitens (FN)	4.94	4.80	1.27	25.8	0.23	2.52	7.25

Table 5

Pairwise comparison for top layer species and thicknesses.

	Different Top Layer Species			Different Thicknesses of Same Species			Other Comparisons
	SN–O	O–FN	SN–FN	TO 1–O 2	SN 2–EV 3	FN 2–EV 3	O–EV	TO–EV	TO–SN	TO–FN
Estimate	0.354	−1.392	−1.038	0.782	0.463	1.501	0.108	0.891	0.427	−0.610
SE	0.298	0.298	0.298	0.421	0.298	0.298	0.298	0.421	0.421	0.421
p value	0.752	<0.001 *	0.005 *	0.339	0.522	<0.001 *	0.996	0.215	0.844	0.591

¹ Composite with 3 mm-thick top layer, ² 5 mm-thick solid sample, ³ composite with 1.2 mm-thick top layer. * difference is statistically significant (p value < 0.05).

Table 6

Comparison of thickness loss% due to abrasion with Janka hardness.

Present Study		Surface of Tested Panels *	[23]
Specimen	Abrasion ∆t%		Flooring Prototype Layers	Layer Thickness (mm)	Layer Density(kg/m3)	JankaHardness (N)
E. nitens veneer engineered flooring	3.44	[Image omitted. Please see PDF.](EV)	1. E. nitens veneer	1.20	430	6623
			2. E. nitens plywood	12.00	765
			3. E. nitens veneer	0.60	430
E. obliqua	3.55	[Image omitted. Please see PDF.](O)	1. Solid E. obliqua	12.00	715	5512
Sawlog E. nitens	3.90	[Image omitted. Please see PDF.](SN)	1. Sawlog E. nitens	6.00	565	4468
			2. Marine plywood	6.00	495
			3. E. nitens veneer	0.60	430
Tasmanian Oak engineered flooring	4.33	[Image omitted. Please see PDF.](TO)	1. Prefinished Tasmanian Oak	3.20	627	5323
			2. Rubberwood (Hevea)	11.0	587
Fibre-managed E. nitens	4.94	[Image omitted. Please see PDF.](F)	1. Fibre E. nitens	6.00	495	4188
			2. Marine plywood	6.00	495
			3. E. nitens veneer	0.60	430

* Shows the surface of the flooring panels containing the material tested on the present study as top layers labeled in align with Figure 2 and Figure 3 and Table 4 and Table 5.

References

1. Khademibami, L.; Shmulsky, R.; Snow, D.; Sherrington, A.; Montague, I.; Ross, R.J.; Wang, X. Wear Resistance and Hardness Assessment of Five US Hardwoods for Bridge Decking and Truck Flooring. For. Prod. J.; 2022; 72, pp. 8-13. [DOI: https://dx.doi.org/10.13073/FPJ-D-21-00074]

2. Knapic, S.; Machado, J.S.; Pereira, H. Properties of cork oak wood related to solid wood flooring performance. Constr. Build. Mater.; 2012; 30, pp. 569-573. [DOI: https://dx.doi.org/10.1016/j.conbuildmat.2011.11.014]

3. Németh, R.; Molnárné Posch, P.; Molnár, S.; Bak, M. Performance evaluation of strip parquet flooring panels after long-term, in-service exposure. Drewno; 2014; 57, pp. 119-134.

4. Welzbacher, C.; Brischke, C.; Rapp, A.; Koch, S.; Hofer, S. Performance of thermally modified timber (TMT) in outdoor application–durability, abrasion and optical appearance. Drv. Ind.; 2009; 60, pp. 75-82.

5. Brischke, C.; Ziegeler, N.; Bollmus, S. Abrasion Resistance of Thermally and Chemically Modified Timber. Wood Ind. Drv. Ind.; 2019; 70, pp. 71-76. [DOI: https://dx.doi.org/10.5552/drvind.2019.1813]

6. Armstrong, F. Timbers for flooring. Forest Product Research. Bulletin; 1957; 40, 31.

7. Harper, F. The abrasion resistance of flooring materials: A review of methods of testing. Wear; 1961; 4, pp. 461-478. [DOI: https://dx.doi.org/10.1016/0043-1648(61)90303-9]

8. Ohtani, T.; Kamasaki, K.; Tanaka, C. On abrasive wear property during three-body abrasion of wood. Wear; 2003; 255, pp. 60-66. [DOI: https://dx.doi.org/10.1016/S0043-1648(03)00219-9]

9. Youngquist, W.; Munthe, B.P. The Abrasive Resistance of Wood as Determined with the US Navy Wear-Test Machine; U.S. Department of Agriculture Forest Service, Report No. R1732 USDA Forest Service, Forest Products Laboratory: Madison, WI, USA, 1948.

10. ASTM D 2394 Standard Methods for Simulated Service Testing of Wood and Wood-Base Finish Flooring; American Society for Testing and Materials (ASTM): Conshohocken, PA, USA, 2017.

11. Brischke, C.; Iseler, N.; Meyer, L.; Sawyer, G. Testing the mechanical resistance of timber used for construction in the marine environment. Int. Wood Prod. J.; 2014; 5, pp. 39-49. [DOI: https://dx.doi.org/10.1179/2042645313Y.0000000050]

12. Swaczyna, I.; Kedzierski, A.; Tomusiak, A.; Cichy, A.; Rozanska, A.; Policinska-Serwa, A. Hardness and wear resistance tests of the wood species most frequently used in flooring panels. Ann. Wars. Univ. Life Sci. SGGW For. Wood Technol.; 2011; 76, pp. 82-87.

13. Coelho, M.U.; Del Menezzi, C.; Souza, M. Abrasion resistance of Pinus wood subjected to thermomechanical treatments. Pro Ligno; 2017; 13, pp. 94-100.

14. Wentzel, M.; González-Prieto, Ó.; Brischke, C.; Militz, H. Physico-Mechanical Properties of Thermally Modified Eucalyptus nitens Wood for Decking Applications. Wood Ind. Drv. Ind.; 2019; 70, pp. 235-245. [DOI: https://dx.doi.org/10.5552/drvind.2019.1838]

15. Sepliarsky, F.; Acuña, L.; Balmori, J.-A.; Martínez, R.D.; Spavento, E.; Keil, G.; Casado, M.; Martín-Ramos, P. Modeling of Falling Ball Impact Test Response on Solid, Veneer, and Traditional Engineered Wood Floorings of Several Hardwoods. Forests; 2022; 13, 167. [DOI: https://dx.doi.org/10.3390/f13020167]

16. Grubîi, V.; Johansson, J. Performance of multi-layered wood flooring elements produced with sliced and sawn lamellas. Pro Ligno; 2019; 15, pp. 166-172.

17. Neyses, B.; Sandberg, D. A new methodology to select hardwood species for wooden products. Wood Mater. Sci. Eng.; 2015; 10, pp. 344-352. [DOI: https://dx.doi.org/10.1080/17480272.2015.1046919]

18. Hague, J. Utilisation of Plantation Eucalypts in Engineered Wood Products; Forest and Wood Products Australia: Melbourne, Australia, 2013.

19. Blacklow, P. An Exploration of Pre and Post-Stressed Timber Forms Utilising Plantation-Grown Eucalypt Timber. PhD Thesis; University of Tasmania: Tasmania, Australia, 2016.

20. Blackburn, D.P. Improving Eucalyptus nitens for Sawn-Board, Veneer and Paper Products. PhD Thesis; University of Tasmania: Tasmania, Australia, 2012.

21. Derikvand, M.; Kotlarewski, N.; Lee, M.; Jiao, H.; Nolan, G. Characterisation of physical and mechanical properties of unthinned and unpruned plantation-grown Eucalyptus nitens H. Deane & Maiden lumber. Forests; 2019; 10, 194.

22. Harwood, C. Sawn timber from native forests and plantations in Tasmania. Coop. Res. Cent. For. Bull.; 2010; 13, Available online: https://www.crcforestry.com.au/publications/downloads/Bulletin-13-Sawn-timber-properties.pdf (accessed on 10 August 2022).

23. Millaniyage, K.; Kotlarewski, N.; Wallis, L.; Taoum, A.; Nolan, G. Janka Hardness Evaluation of Plantation-Grown Eucalyptus nitens for Engineered Flooring Applications. Buildings; 2022; 12, 1862. [DOI: https://dx.doi.org/10.3390/buildings12111862]

24. Washusen, R.; Harwood, C.; Morrow, A.; Northway, R.; Valencia, J.C.; Volker, P.; Wood, M.; Farrell, R. Pruned plantation-grown Eucalyptus nitens: Effect of thinning and conventional processing practices on sawn board quality and recovery. New Zealand J. For. Sci.; 2009; 39, pp. 39-55.

25. AS 2082 Timber–Hardwood–Visually Stress Graded for Structural Purposes; Standards Australia: Sydney, Australia, 2007.

26. AS 2796.2 Timber–Hardwood–Sawn and Milled Products, Part 2: Grade Description; Standards Australia: Sydney, Australia, 1999.

27. Derikvand, M.; Nolan, G.; Jiao, H.; Kotlarewski, N. What to do with structurally low-grade wood from Australia’s plantation eucalyptus; building application?. BioResources; 2017; 12, pp. 4-7. [DOI: https://dx.doi.org/10.15376/biores.12.1.4-7]

28. Nolan, G.; Greaves, B.; Washusen, R.; Parsons, M.; Jennings, S. Eucalypt Plantations for Solid Wood Products in Australia—A Review ‘If you Don’t Prune it, we Can’t Use it’; Forest & Wood Products Research & Development Corporation: Melbourne, Australia, 2005; Available online: https://eprints.utas.edu.au/7045/ (accessed on 10 August 2022).

29. Kotlarewski, N.; Taylor, L.; Booth, P. Embracing natural timber features of plantation hardwood: Material-aware digital workflows in product design and development. Proceedings of the 52nd International Conference of the Architectural Science Association; Melbourne, Australia, 28 November–1 December 2018.

30. Riddle, T. A white timber floor. Build. Connect.; 2015; pp. 44-45. Available online: https://search.informit.org/doi/epdf/10.3316/informit.256207485215981 (accessed on 6 January 2022).

31. ATFA. The Final Appearance; Australian Timber Flooring Association: Shelly Beach, Australia, 2009.

32. Vörös, Á.; Németh, R. The History of Wood Hardness Tests. IOP Conference Series: Earth and Environmental Science; IOP Publishing: Bristol, UK, 2020; Volume 505, No. 1 012020.

33. ASTM D 143 Standard Test Methods for Small Clear Specimens of Timber; American Society for Testing and Materials (ASTM) International: Conshohocken, PA, USA, 2000.

34. AS 1720.2 Timber Structures Timber Properties; Standards Australia: Sydney, Australia, 2006.

35. Grześkiewicz, M.; Kozakiewicz, P.; Borysiuk, P.; Romanovski, V.; Cichy, A. Influence of top layer density and thickness on hardness of two-layer floor elements. Drew. Pr. Nauk. Doniesienia Komun.; 2020; 62, pp. 69-80.

36. EN 14354 Wood-Based Panels—Wood Veneer Floor Coverings 2017; European Committee for Standardisation (CEN): Brussels, Belgium, 2017.

37. AS/NZS 1080.1:2012 Timber—Methods of Test—Moisture Content; Standards Australia: Sydney, Australia, 2012.

38. AS/NZS 1080.3:2000 Timber—Methods of test, Method 3: Density; Standards Australia: Sydney, Australia, 2000.

39. Ohtani, T.; Kamasaki, K. Effect of microscopic tissue on three-body abrasion in cell structure of wood. Wear; 2007; 262, pp. 453-460. [DOI: https://dx.doi.org/10.1016/j.wear.2006.06.001]

40. Rede, V.; Essert, S.; Šokčević, Z. Effects of microstructural orientation on the abrasive wear resistance of subfossil elm wood in three orthogonal planes. Wear; 2017; 380, pp. 1-5. [DOI: https://dx.doi.org/10.1016/j.wear.2017.03.001]

41. Mohd-Jamil, A. Janka hardness rating of Malaysian timbers. Timber Technol. Bull.; 2016; 58, 6.

42. Drerup, M.J.; Erdem, I.; Anthony, R.W. Trouble Underfoot—In Situ and Laboratory Investigation of Engineered Wood Flooring. Proceedings of the Forensic Engineering 2012: Gateway to a Safer Tomorrow; San Francisco, CA, USA, 31 October–3 November 2012; pp. 621-630.