Analysis of the Properties of Upcycled Wood Waste for Sustainable Furniture Production

Abstract

Although linear overproduction and overconsumption have benefited businesses, they have created an unsustainable society. Converting wood waste into construction material can support the transition to a circular economy. The mechanical properties of beams constructed from wood waste were measured. Squares with 50, 60, and 70 mm side lengths were glued to create beams, to which the three-point test method was applied parallel to the fibres. The stiffness and moduli of elasticity and rupture were analysed with standard industrial statistical techniques. Specifically, a two-stage analysis was performed using the normal distribution and Shewhart control charts. Changes of 100 mm in width and height and 200 mm in length caused a change of 200–400 N/mm² in elasticity and 500–1300 MNmm² in stiffness. Modulus of rupture values were relatively comparable, as they were determined by the properties of oak wood, from which the beams were made. The observed differences in the tested mechanical parameters will be useful in the optimisation of furniture construction, with our research suggesting that it is possible to predict mechanical properties from the dimensions of the waste-wood pieces. Ultimately, this should help to design sustainable furniture that is aesthetic, functional, and safe.

1. Introduction

Although businesses have benefited from linear overproduction and overconsumption, these practices have created an unsustainable economy. The linear economy refers to a traditional economic model in which raw materials are extracted and used to manufacture products, which are then consumed and discarded. This approach, which converts raw materials into waste over a short life cycle, has proven, after many years of experience, to be unsustainable for all involved parties. The linear economy has stripped the world of natural resources, produced large amounts of waste, harmed the environment, and contributed to a variety of social problems.

In contrast, a circular economy is based on three design-driven principles that result in a more sustainable economic model: eliminate waste and pollution, circulate products and materials (at their highest value), and regenerate nature [1]. A circular economy facilitates the maintenance and circulation of goods by reusing, repairing, recycling, and upcycling materials and products in a cascading use model for as long as possible. A circular economy and circular business models can provide sustainable economic growth, benefitting all parties (the economy, the society, and the environment) in the long term [2,3]. The conversion of waste into valuable material affects the investment characteristics of a circular economy and can support the transition from a linear business model to a circular one [4,5,6,7,8,9]. As reported by Deloitte [10], “700 [million] USD is the annual material cost savings in the fast-moving consumer goods industry with the implementation of a new circular economy”. Integrating circular economy (CE) principles into the wood industry is increasingly seen as a way to promote sustainability, reduce waste, and optimise resource use. Typical cases focus on reusing wood waste, recycling materials, extending product life, and creating closed production cycles. Many companies transform wood waste into new products, such as wood panels, bioenergy, or construction materials. This not only reduces landfill waste but also creates new revenue streams and stimulates further innovation in the sector [11].

Product development and material choices are the primary avenues for improving sustainability in production. Introducing sustainable materials can enhance business growth and reduce production costs while providing environmental and social benefits. Sustainability has become a primary decision-making factor in the global economy. According to the United Nations, sustainability aims to meet the “needs of the present without compromising the ability of future generations to meet their own needs”. Innovative, sustainably designed goods can contribute to circular business and economic models [12,13]. Manufacturing new sustainable products from wood waste may be a solution to deforestation, which is a major cause of global warming; the preservation of forests is thus essential for a sustainable future [14,15]. It is encouraging to note that wood recovery methods such as recycling into particleboard, deconstruction, organosolv treatment, and energy recovery are widely practiced, with recovery rates ranging from 30% to 46% depending on the method and context. However, it is thought that the greatest sustainability benefits are offered by the upcycling methodology, which uses advanced eco-design and easy technologies. It is estimated that wood waste can be reduced by up to 100% through the use of these technologies [16]. Figure 1 shows the quantities and treatment methods in the 28 EU countries in 2010 [17].

A review of the extant literature reveals that the most prevalent utilisation practice for wood waste among engineered wood products is that of manufacturing particleboard. As is evident from a survey of the literature, a large number of studies have been carried out on the environmental aspects of the production of particleboard [18]. The wood panel industry is already implementing circular economy practices in its waste management strategies, in addition to potential practices that could enhance its circularity. The changes under discussion are oriented towards sustainable manufacturing and responsible consumption [5].

Implementing circular economy principles in the wood industry, such as upcycling wood waste or residues through the design of high-value sustainable products, can help save forests and develop sustainable business practices. A product has environmental impacts throughout its life cycle, from raw material extraction to manufacturing byproducts to end-of-life waste. Products that are sustainable in all these aspects are required to transition economies from linear to circular, maintaining natural resources, protecting forests, maintaining a healthy environment, and ultimately, promoting human survival [19,20,21,22].

Of primary importance is the sustainability level in materials development. Transforming waste into new, high-value products is an innovative material solution that can help achieve production sustainability in business [23]. As waste materials possess variable characteristics (e.g., size), one of the key tasks when upcycling materials is adjustment. Solid wood from a sawmill has large dimensions, allowing it to be processed into semi-finished products that can be used to manufacture numerous other goods. Due to the various nonstandard dimensions and the presence of waste and residues, wood waste is less versatile. Attempts can be made to glue small off-cuts together to produce a larger semi-finished product; this customisation is one principle underlying smart design [24].

This study aims to investigate the influence of the geometry of waste on the mechanical properties of semi-finished products. This includes the determination of bending modulus, bending strength, and stiffness. A ready-to-make design is presented for a new material composed of wood waste, which should ensure higher resource efficiency and material productivity. This material was developed to demonstrate a new way of designing and implementing sustainably designed semi-finished products according to upcycling principles, transforming wood waste into high-value goods. The adoption of the new material could help conserve natural resources, reduce energy consumption, limit pollution, and ensure zero-waste production while offering a high-value, sustainable interior architecture solution with a ready-to-use design. The properties of such waste materials vary depending on the extent of transformation, whether from a material’s transition from product to waste or from waste to a new, upcycled product.

2. Materials and Methods

Unlike the linear economy, which uses raw material that has never been processed, the circular economy requires waste material [25,26,27]. This study concerns the upcycling of waste from the production of flooring materials from solid oak (Quercus robur L.) wood from southeastern Poland. Beams were selected as base furniture elements to be tested, representing the most common shape in furniture production. The geometry of the beams was selected according to aesthetic features and the experience gained during furniture prototyping (Figure 2). However, there is a lack of information about preliminary mechanical values to guide further development, especially since the beams were made so that bending, compressive, and stretching forces act perpendicular to the fibres. Similar studies have yielded little information, particularly concerning beams with different transverse dimensions [17,28,29,30,31,32,33,34,35,36,37].

2.1. Experimental Materials and Samples

Test samples were made from flooring material post-production waste (Figure 3α) of the highest wood quality class A, following the EN 1927-2:2008 standard [38]. The samples were made of oak wood (Quercus L.) with an average density of 720 ± 40 kg/m³. The samples were stored for a period of one to three months in a roofed room. Therefore, during collection they had air-dry humidity, i.e., from 12 to 15%. The samples comprised rectangular cuboid shapes of 20 mm thickness. Due to the production characteristics of flooring materials made from solid wood, the elements had to be glued together with broad surfaces in a cross pattern (plywood-like) to obtain beams (Figure 3β). In the drawing, the fiber directions in adjacent layers are marked with arrows.

Two groups of samples were prepared using different methods. The first group was pressed with a bar clamp (BC) and the second with a simple press (SP). The method of sample preparation was informed by the large share of micro (94%) and small (4%) enterprises in the furniture and interior furnishings industry [39]. This is particularly true in the solid wood flooring materials industry. Short, medium, and long beams were constructed with dimensions 50 × 50 × 300 mm, 60 × 60 × 500 mm, and 70 × 70 × 700 mm, respectively. Eleven samples of each size were made.

A one-component adhesive, based on a modified polyvinyl acetate dispersion in water-resistance class D3, was chosen. The conditions of sample preparation were as follows:

• Adhesive application 90–180 g/m²;
• Bonding process temperature 15–25 °C;
• Wood moisture 8–12%;
• Air relative humidity 40–70%;
• Cold press time 15–30 min;
• Pressure 1 N/mm².

After glueing, the samples were seasoned at 21 °C and 35% relative humidity for 60 days. These parameters were chosen on the basis of estimated mean values in homes in central Poland [40]. At the end of seasoning, samples had reached a moisture content of 8% [41]. The samples were then measured and weighed. Length was measured with an accuracy of 1 mm. In comparison, width and height were measured with an accuracy of 0.05 mm (the calliper accuracy was 0.01 mm, but the measurements depended on the amount of manual pressure). Density was determined according to EN 323:1993 [42]. The mean sample density was 710 ± 20 kg/m³. No existing standards are relevant to the tested samples because they were designed so that forces act in a direction perpendicular to the fibres rather than parallel, which is the more common situation when using new wood in construction.

2.2. Experimental Methods

Test methods were based on international standards regarding the modulus of elasticity (MOE), static bending strength (MOR), and stiffness, denoted by E_m, f_g [43], and k [44], respectively. The methods followed those established in the literature [45,46,47]. However, the standard test methods had to be adapted to the geometry of the samples. The devices used are also described in these works. The primary difference between our study and previous studies was the distance between support centres. In our study, these distances were 250 mm, 430 mm, and 610 mm for short, medium, and long beams, respectively. All the results were analysed statistically by calculating the arithmetic mean, median, standard deviation, and variation coefficient.

A quantitative analysis of the samples was carried out regarding the feasibility of industrial production. For this purpose, an analysis of the normal distribution and kurtosis was carried out. This stage of the study allowed for the preliminary determination of the stability of the sample production process and the sigma distribution of the obtained parameters. The analyses were summarised using Shewhart charts, which allowed us to decide whether deviations from the mean observed at a given moment could be treated as random or required another explanation. The exact procedure of the conducted analyses was described in detail in previous publications [47,48].

The last step was an analysis to identify statistically significant differences between pairs of variables, comparing the observed values of MOE, MOR and k for all sample groups.

3. Results and Discussion

The results were recorded when about 75% of the wood’s surface area displayed cracks. Samples with a cracked area in the joint exceeding approximately 35% were disregarded. The surface area was estimated by superimposing a grid on the cross-section of the samples after testing. These situations were likely due to the use of handmade samples (BC). The length of the samples relative to their width was large, making it difficult to position each element manually. The first attempts checked whether standard methods of producing furniture elements would meet the requirements, following the principle of starting with simple, well-known production methods and only looking for new and better solutions if these failed. These solutions were determined more by the waste material, due to its defined shape, than by the product being manufactured, as is typical in standard production.

3.1. Modulus of Elasticity

The MOE measurements for individual samples are presented in

Table 1. Modulus of elasticity (MOE) measurements [N/mm²].

Sample Number	Short Beam	Medium Beam	Long Beam
1	1048	1545	1586
2	1078	1208	1318
3	1020	1399	1248
4	612	1266	1375
5	639	1121	1284
6	1342	997	1279
7	1352	1518	–
8	1123	1136	–
Mean	1027	1274	1348

, together with the mean values for the short, medium, and long beams, which are plotted in Figure 4.

The MOE value increased as the samples’ cross-sectional area increased. However, the samples with larger cross-sections were longer and had wider support spacings. Samples with larger cross-sectional dimensions could not be made to the same length as samples with a smaller cross-sectional size, as further increasing the MOE value for furniture, especially skeleton furniture, would not be aesthetic. The MOE was directly proportional to the dimension, making it possible to predict and control the MOE when designing furniture. The results of the calculations necessary to analyse the normal distribution and prepare the Shewhart control charts are presented in

Table 2. Values for the distribution and Shewhart control chart of MOE values [N/mm²].

Variable	Value
Arithmetic mean	1204
Standard deviation	249
Lower control limit (LCL)	457
Lower specification limit (LSL)	706
Upper specification limit (USL)	1702
Upper control limit (UCL)	1951

.

Figure 5 shows the distribution of the MOE results obtained for samples of all dimensions. As can be seen, all results are within the control limits. However, 2 of the 22 measurements (9%) are on the lower specification line. To perform a statistical analysis using a Shewhart control chart, approximately 67% of all results must fall between the control limits. Therefore, it was possible to use a Shewhart chart to analyse the MOE results of the tested samples.

By analysing the distribution of the MOE results, a kurtosis coefficient of 1.07 was calculated. A positive kurtosis indicates a leptokurtic distribution of the obtained results. This means that the MOE values are concentrated around the average but with a greater chance of outliers. However, a result not exceeding 2.0 allows us to safely assume that the obtained distribution is close to the normal distribution in kurtosis.

A detailed analysis of the sample measurement results is presented in Figure 6. The Shewhart control chart allows us to determine which samples differ in values from the others and thus disturb the stability of the production process.

According to the Shewhart control chart, the MOE values of only two of the tested samples are under the lower specification limit of 706 N/mm². The MOE values of samples 4 and 5 from the short beam group are 612 and 639 N/mm², respectively. However, none of the analysed samples were outside the designated control limits (457–1951 N/mm²). Therefore, it was concluded that all the samples produced met the MOE strength criteria set for products intended for sale. Since 100% of the tested samples were between the control limits, it can be concluded that the production process of the components is currently stable and does not require any corrections.

However, the literature sources indicate that the minimum MOE value across the grain for glued laminated timber used in building structures is 300 MPa [49,50], and its minimum value for solid timber structures is 630 MPa [51]. The obtained sample values exceed this in 91% of cases despite being intended for use at much lower loads.

3.2. Modulus of Rupture

The results of the modulus of rupture (MOR) measurements for individual samples are presented in

Table 3. Modulus of rupture (MOR) measurements [N/mm²].

Sample Number	Short Beam	Medium Beam	Long Beam
1	11.14	4.92	4.60
2	9.41	5.54	6.70
3	9.00	6.03	7.17
4	10.52	3.21	6.35
5	12.52	5.62	5.64
6	9.32	7.47	6.07
7	9.48	7.24	–
8	9.01	7.07	–
Mean	10.05	5.89	6.09

, along with the mean values for the groups of samples, which are plotted in Figure 7.

The MOR values decreased with increasing dimensions and lengths of the samples. This is mainly due to the increase in length. MOR is directly proportional to the maximal force. Optimisation of the ratio of cross-sectional area to length will be possible after testing furniture joints and whole furniture models. The proportionality of the results is imperfect, but it should be remembered that the tested elements were made of wood. Each solid wood part is different in structure, leading to differences in physical and mechanical properties. Our results demonstrate the possibility of predicting trends and values. The results of the calculations necessary to analyse the normal distribution and prepare the Shewhart control charts for MOR results are presented in

Table 4. Values for the distribution and Shewhart control chart of MOR values [N/mm²].

Variable	Value
Arithmetic mean	7.46
Standard deviation	2.33
LCL	0.47
LSL	2.80
USL	12.12
UCL	14.45

. Figure 8 shows the distribution of the MOR results obtained for samples of all dimensions.

According to the distribution of MOR values, all results are within the control limits. However, 1 out of 22 results is on the lower specification line, and 1 is above the upper specification line. These represent about 9% of all the measured samples. As discussed in the previous analysis, to perform a study using a Shewhart control chart, approximately 67% of all results must fall between the specification lines. Therefore, it was possible to use a Shewhart chart to analyse the MOR results of the tested samples.

The kurtosis analysis for the MOR values in Figure 8 yielded a negative value of −0.32. This negative but relatively small value could be consistent with either a mesokurtic or platykurtic distribution. The mass of the distribution is spread out; the values are scattered around the mean, and there is a lower probability of extreme values. In practice, this means there is also a lower probability of obtaining values close to the nominal value. A detailed analysis of the MOR measurement results is presented in Figure 9 using a Shewhart control chart.

According to the Shewhart control chart, the MOR value of only one of the tested samples is below the lower specification limit of 2.80 N/mm². However, none of the analysed samples were outside the designated control limits (0.47–14.45 N/mm²). The highest MOR value was found in the short beam group, in which all eight samples had MOR values higher than the target of 7.46 N/mm². The samples from the other two groups (medium and long beam) had MOR values that were mostly close to but below the target.

Therefore, it was concluded that all the samples met the MOR strength criteria set for products intended for sale. Since 100% of the tested samples were between the control limits, it can be concluded that the production process of the components is currently stable and does not require any corrections.

3.3. Stiffness (k)

The results of the stiffness measurements for individual samples are presented in

Table 5. Stiffness measurements [MNmm²].

Sample Number	Short Beam	Medium Beam	Long Beam
1	244	885	2170
2	231	702	2682
3	244	817	2573
4	316	1433	2877
5	334	1265	2702
6	722	1109	2669
7	1745	1858	–
8	1444	1407	–
Mean	660	1184	2612

, together with the mean value for each group of samples. These are plotted in Figure 10.

The calculation of stiffness was based on the MOE, so similar changes were noted. Stiffness indicates how beams resist external loads. There was a large difference in stiffness depending on the overall dimensions of the beams. This suggests several possibilities for the selection of dimensions depending on where in the structure a given element is located and what forces and moments act on it and its joints.

The results of the calculations necessary to analyse the distribution and prepare the Shewhart control charts for stiffness results are presented in

Table 6. Values for the distribution and Shewhart control chart of stiffness values [MNmm²].

Variable	Value
Arithmetic mean	1383
Standard deviation	911
LCL	0
LSL	0
USL	3205
UCL	4116

. The widely scattered stiffness values resulted in a large value of the total standard deviation of 911 MNm². The adoption of this value would result in the determined limits having large or negative values, which is inappropriate and unphysical in the case of stiffness. Therefore, the stiffness analysis assumed a common value of zero for the lower control limit and the lower specification limit.

Figure 11 shows the distribution of the stiffness results obtained for samples of all dimensions. The negative stiffness values are only shown to illustrate the full shape of the distribution.

According to the distribution of stiffness values, all results were within the control and specification limits. To perform a study using a Shewhart control chart, approximately 67% of all results must fall between the specification limits. Therefore, it was possible to use a Shewhart chart to analyse the stiffness results of the tested samples.

The kurtosis analysis for the stiffness values in Figure 11 yielded a negative value of −1.25. This indicates a platykurtic distribution for the obtained stiffness values. The mass of the distribution is spread out. A negative kurtosis value indicates a lower intensity of extreme values and a visually flat distribution. The obtained negative excess indicates that the tails of the stiffness value distribution are narrower than those of the normal distribution. A detailed analysis of the stiffness results is presented in Figure 12 using a Shewhart control chart.

According to the Shewhart control chart, the stiffness values of all tested samples were between the lower and upper specification limits (0–3205 MNmm²). The highest stiffness values were found in the long beam group. Six of the eight samples from this group had k values higher than 2000 MNmm², and all of them had values higher than the target value. The samples from the short beam group had the lowest k values, but they were also the most stable.

3.4. Statistical Analyses

The next element of the analysis was to identify statistically significant differences for pairs of variables, comparing the observed values of MOE, MOR, and k between the different types of studied beams. This approach requires several laboratory analyses, the results of which may not directly reflect typical behaviour. Hence, a common solution is to perform a statistical analysis of the results obtained from research. Accordingly, the values of arithmetic means, medians, minimum and maximum values, standard deviations, and coefficients of variation were calculated for the MOE, MOR, and k values, taking into account the type of material for which these values were examined.

Table 7. Descriptive statistics for the studied material types.

Type of Material	Variable	Arithmetic Mean	Median	Minimum	Maximum	Standard Deviation	Variation Coefficient
Short beam	MOE	983	1047	596	1352	276	28
	MOR	9.13	9.32	5.43	12.52	1.82	19.91
	k	579	282	195	1745	547	94
Medium beam	MOE	1310	1332	987	1603	220	17
	MOR	5.35	5.58	2.91	7.47	1.54	28.77
	k	1164	1187	473	1858	447	38
Long beam	MOE	1491	1375	1248	1796	235	16
	MOR	5.60	5.64	4.28	7.17	1.03	18.44
	k	2556	2573	2170	2877	206	8

presents descriptive statistics characterising the spread of values in the distribution of each variable.

For short beams, the MOE and MOR values were characterised by low variability, as evidenced by the similar arithmetic mean and median values. Additionally, the coefficients of variation for these parameters indicated low variability. For the short beams, the highest variation was observed for k, as indicated by the large differences in the arithmetic mean and median values and the high coefficient of variation (94.43%). For medium and long beams, the values of the arithmetic mean and median for individual toughness parameters were similar, and the coefficients of variation indicated low variability.

4. Conclusions

The mechanical properties of a material made from solid wood waste were tested. Changes of 100 mm in width and height and 200 mm in length caused a change of 200–400 N/mm² in the MOE and 500–1300 MNmm² in the stiffness. The MOR values were relatively comparable, as they were determined by the properties of the individual elements of oak wood from which the beams were made. The differences in the tested mechanical parameters suggest directions to optimise furniture construction; in particular, it should be possible to predict mechanical properties from the dimensions of elements in furniture design, allowing aesthetics, functionality, and safety to be maximised. These new materials can be adopted to reduce wood waste while increasing the output of semi-finished products, yielding higher resource efficiency, productivity, and material utilisation ratios. Using waste as an innovative material to design new, high-value market products provides a pathway to achieve sustainable production in business.

Importantly, since we are dealing with a waste material whose properties are variable and difficult to define, the methods and technologies used to create furniture components must be carefully analysed. Research and quantitative analysis have shown that it is possible to create high-quality, ready-to-use, and sustainable material from wood waste. This type of upcycling can help resolve problems in maintaining forest resources and achieving sustainable growth in a circular economy. Few studies have successfully predicted the values of the investigated mechanical quantities for wood-waste constructions. There is a need for future research exploring construction methods that account for the geometry of wood off-cuts to yield the most efficient use of waste and enable upcycling. Each type of waste or production residue must be approached individually.