1 Introduction

The concept of sustainable development has become more and more common and recognizable in recent years. For this reason, it covers an increasingly broader spectrum of economic sectors, including the wood industry and the production of wood-based materials as well. Activities in this area focus on the responsible management of wood raw materials and the implementation of sustainable practices in the harvesting, producing, and utilizing the produced boards. Due to this approach, scientists from research centres located all over the world are trying to use various types of wastes and by-products in the production of high-quality products with a perspective for innovative applications and unique properties. So far, several alternative materials have been found suitable for the production of wood-based materials. The most frequently studied include by-products of primary wood processing (Mirski et al. 2020b), waste wood-based materials (Laskowska 2024), the packaging industry (Andrade et al. 2016), wood of fast-growing species such as poplar (He et al. 2016), willow (Warmbier and Wilczynski 2016), sendan (Utsumi et al. 2019) and agricultural biomass (Müller et al. 2012), in particular the straw of various types of cereals, such as rapeseed (Dukarska et al. 2017, 2021; Cosereanu and Cerbu 2019), wheat (Boquillon et al. 2004; Bekhta et al. 2013), white mustard (Dukarska et al. 2015) and rice straw (Ferrandez-Garcia et al. 2017). It has also been shown that tree bark (Medved et al. 2019; Mirski et al. 2022) and sycamore leaves (Pirayesh et al. 2015) can be used as a partial substitute for wood particles in particleboard production. Additionally, walnut shells are an interesting example of waste material that recently attracted major attention.

Walnut processing has been growing worldwide for years. In the 2022/2023 season, global production reached approx. 2.67 million tons, but according to predictions, it may decrease slightly in the 2023/2024 season. The largest producer processing about 1.4 million tons of walnuts is China. The United States, with more than 682.2 thousand tons, ranks second in terms of walnut production worldwide. In the European Union countries, walnut production is about 163 thousand tons (Shahbandeh 2023; Abdulwahid et al. 2023). In the walnut processing industry, after separating the fruit from the remaining part, a significant amount of shells is generated because they constitute over 40–60% of the entire mass of the walnut fruit. Considering global production, it can be stated that the processing of walnuts contributes to the production of more than 1 million tons of shells per year (Açıkbaş 2018). In the European Union alone, a significant amount is also generated, achieving more than 65 thousand tons of shells per year. Thus, as stated by Jahanban-Esfahlan and Amarowicz (2018), their great potential results from the production in large quantities and high availability. The industrial application of walnut shells reduces the generated waste and supports the development of ecological practices. Currently, they are used for energy purposes as solid fuel, as an abrasive material for polishing, e.g., soft metals, stones, glass fibres, plastics, or wood, and as a filter medium to separate crude oil, hazardous materials, and heavy metals from water (Anjum et al. 2017; Stanicka et al. 2022). Ground walnut shells can also be added as an organic fertilizer to improve the soil’s structure and provide nutrients for plant. Moreover, walnut shell-based extracts or oils can be applied in the cosmetic industry. Following the assumptions of the circular economy, new ways to effectively use such wastes are constantly being sought. The research carried out so far in the field of materials engineering showed that, due to their rich chemical composition, walnut shells can be used in the production of a new class of PUR foams, e.g., as fillers or to obtain bio-polyols (Członka et al. 2020, 2021). It was also found that thermoplastic polymer composites filled with these shells demonstrate better mechanical properties than wood-containing ones. According to Sarsari et al. (2016), adding up to 40% of walnut shell flour (WSF) significantly improves the tensile strength, flexural strength, and elastic modulus of composites based on thermoplastic starch. Furthermore, in the case of polypropylene composites, the addition of WSF results in an increase in both bending and tensile modulus (Ayrilmis et al. 2013) and improved thermal stability (Dobrzyńska-Mizera et al. 2019).

Shells of various types of nuts, such as walnuts or almonds, can also be used in the production of wood-based materials as well. As shown by Barbu et al. (2020), walnut shells can be used to produce particleboards with increased dimensional stability and higher Brinell hardness compared to softwood particleboards. However, despite a higher density, this type of board demonstrates significantly lower bending strength and modulus of elasticity, and the reduction is sufficient enough that it does not allow the standard requirements to be met, even for P1 type boards. Therefore, on this basis, it can be concluded that the concept of partial wood substitution seems to be justified. The necessity of finding the right proportion is confirmed by the research conducted by Pirayesh et al. (2012a 2012b, 2013), which showed that as the degree of wood particle substitution with the walnut shells increases, the water resistance improves and the mechanical properties deteriorate. Moreover, it was found that to reach the required level of modulus of elasticity, the share of shells should not exceed 20%. Interestingly, shell-containing boards glued with urea-formaldehyde adhesive were characterized by significantly reduced formaldehyde emission, which is especially important for the boards intended for indoor applications (Pirayesh et al. 2013). Similar conclusions were drawn by Silva et al. (2017), who showed that adding walnut shells to MDF boards contributes to removal of volatile organic compounds (VOC) and formaldehyde from indoor air. What is important, the authors also found that this effect is irreversible in the case of formaldehyde and dodecane.

In summary, it can be concluded that the research works conducted so far, on the one hand, show improved water resistance and, on the other hand, a deterioration in the mechanical properties of manufactured shell-containing boards. Therefore, this research was carried out to explore the possibility of using particleboards containing walnut shells in the production of sandwich boards. It is worth noting that the novelty of our research was the production of particleboard with a relatively higher amount of walnut shells while reducing the density of the panels and using them as the core of sandwich boards. As can be seen from the bibliography, practically no research has been carried out to date on similar subject, which makes our research innovative. Thus, in summary to investigate the possibility of producing walnut shell-containing boards of low density and to use them as a core for sandwich boards.

2 Materials and methods

2.1 Characteristics of raw materials

In the first stage of the research, single-layer particleboards differing in shares of walnut shells with dimensions of 550 × 700 × 19 mm3 and a density ranging from 500 to 550 kg/m3 were produced. Industrial pine (Pinus sylvestris L.) particles with a moisture content of 3 ± 2% were used for the production. The dimensions of wood particles were determined based on sieve analysis, and the results of the fractional composition obtained are presented in Fig. 1a. Walnut (Juglans regia) shells, obtained from a private plantation in Poland, were used as a partial substitute for wood particles. During preliminary tests, it was determined that in order to obtain a particleboard with reduced density, the replacement of wood particles with walnut shells should not exceed 50%. This limitation is due to the high bulk density of walnut shells, significantly exceeding the bulk density of wood particles, as shown in Fig. 1b. For this reason, the study used shell particles ground into particles with relatively large dimensions, i.e., 8-20 mm long, 9-17 mm wide, and 1.7-5 mm thick (Fig. 2). Following the grinding process, the shells were sifted through a sieve to eliminate excessively large (re-ground) and fine fractions.

Fig. 1
figure 1

Selected properties of the raw material used in the production of bio-composite particleboards: a - fractional composition of wood particles, b – bulk density of wood particles and walnut shell particles

Full size image
Fig. 2
figure 2

Wood particles and ground walnut shells used in the manufacture of bio-composite particleboards

Full size image

Two types of binding agents, such as urea-formaldehyde (UF) resin and 4,4′-methylenediphenyl isocyanate (pMDI) adhesive, were used during the process of bio-composite particleboard manufacturing. The UF resin of industrial origin was characterized by the following properties: viscosity of 2450 mPa·s, density 1.319 g/cm3, solids content 64.5%, pH 7.80, and gel time at 100 °C 152 s. A 20% solution of ammonium nitrate (NH4NO3) in the amount of 2% by a solid mass of UF resin was introduced as a curing agent. The pMDI adhesive was characterized by a solids content of 100%, an isocyanate group content of about 32%, an acid number of 0-2000 pp, and a viscosity of 150–300 mPa·s.

2.2 Board manufacturing and testing

The board production process involved two stages. In the first stage, particleboards containing 25%, 50%, and 75% of ground walnut shells were produced using UF resin (gluing degree of 10%). Unfortunately, during the process of cutting the boards with 75% of the shells, there was damage to the structure of the material, expressed as fracture. For this reason, this particular variant was excluded from further study. In the second stage, particleboards with 25% and 50% walnut shells were produced. However, the wood particles were glued with 10% UF resin, and the walnut shell particles were glued with pMDI with a gluing degree of 5%. After gluing, both materials were mixed together manually during the mat formation.

The hot pressing operation was performed using standard parameters, i.e., unit pressure of 2.5 N/mm2, temperature of 200 °C and pressing time of 22 s/mm of board thickness. The produced boards (Fig. 3) were characterized by a lower density than traditional particleboard (in their case, density is typically ranging from 700 to 750 kg/m3) and, as expected, showed lower strength parameters.

Fig. 3
figure 3

Experimental bio-composite particleboards produced with different shares of walnut shells

Full size image

Therefore, in the next step of the research, the produced bio-composite boards were covered on both sides with HDF board with a thickness of 2.95 mm and a density of 850 kg/m3 (Pfleiderer, Poland). A single-component polyurethane adhesive Chemolan B (Interchemol S.A., Poland) was applied on the surfaces of the HDF boards at a spread rate of 120 g/m2. Then, the prepared particleboards were placed as a core layer between the HDF sheets. The arranged sets were pressed at room temperature, under the unit pressure of 0.8 N/mm2 for 24 h. The resulting sandwich boards were characterized by a thickness of 25 mm.

The manufactured particleboards being the core for the sandwich boards were conditioned for 7 days at a relative humidity of 65 ± 5% and a temperature of 20 ± 2 °C prior to testing, and then their strength properties were investigated. The following parameters were determined: bending strength (MOR), modulus of elasticity (MOE) in accordance with EN − 310 (1993), and internal bond (IB) in accordance with EN − 319 (1993). In addition, in the case of boards glued in a hybrid way with UF resin and pMDI adhesive, the thickness swelling (TS) of the boards after 24 h of soaking in water was examined in accordance with EN − 317 (1998). Their water absorption (WA) was also determined based on the difference in weight before and after soaking.

Samples of bio-composite particleboards were also subjected to a fire resistance test in a mass loss calorimeter (MLC) according to the ISO 13927 (2015) procedure described by Clausen et al. (2014). The samples were exposed to thermal radiation (HF) with an intensity of 35 kW/m2 at a distance of 25 mm from the sample’s surface. The area of the samples exposed to heat was 88.4 cm2 (Schartel and Hull 2007; Mazela et al. 2018). The following parameters indicating the fire resistance of boards, such as mass loss (ML), heat release rate (HRR), total heat release rate (THR), time to ignition (TTI), and time to flame extinction (TTF), were determined.

The manufactured sandwich boards were tested in terms of their physical and mechanical properties as well. Their bending strength (MOR) and modulus of elasticity (MOE) were determined according to EN − 310 (1993), thickness swelling (TS) according to EN − 317 (1998), and water absorption (WA) based on the weight change during soaking.

Regardless of the variant tested, each board was produced in two replicates. In the analysis of the results, the first part of the study used a board made of wood particles only as a reference board (REF). It was manufactured with the same dimensions, density and following the same production conditions as the boards containing walnut shells. In the second part of the analysis, the reference board was also a board made of only wood particles, but covered with HDF boards (REF/HDF). The internal bond, swelling, and water absorption tests of boards were conducted in ten repetitions, whereas bending strength and modulus of elasticity were conducted in five repetitions. Flammability tests of the boards were performed in three repetitions.

2.3 Statistical analysis

Two-way analysis of variance (ANOVA) was performed to analyse the collected results. Moreover, the HSD Tukey test at the significance level of α = 0.05 was performed using Statistica 13.3 software to distinguish homogeneous groups and assess the significance of observed changes.

3 Results and discussion

3.1 Strength properties of the produced bio-composite boards and sandwich boards

The mechanical properties of the produced particleboards were assessed based on their internal bond, bending strength, and modulus of elasticity, and the obtained results are presented in Figs. 4 and 5. Based on the outcomes, it was found that replacing wood particles with walnut shell particles reduced the boards’ strength. The results of ANOVA confirmed the statistically significant influence of both variable factors, i.e., the share of shell particles and the gluing method. As shown in Fig. 4, particleboards bonded with UF resin with 25% of wood particles substituted with shell particles demonstrated a considerable decrease in IB by 44%. Furthermore, an increase in the share of shells to 50% caused a deterioration even by up to 65%. Such a significant decrease in these values justifies the application of pMDI adhesive for the gluing operation. Using the hybrid gluing method led to a significant improvement in the bonding quality of the boards and caused an increase in IB by 58% and over 110% in the case of boards containing 25% and 50% shells, respectively. However, it should be noted that despite a significant improvement, the experimental boards still exhibited lower IB than the reference boards (REF). The difference was approx. 10% for boards containing 25% shells and approx. 23% for boards containing 50% shells. On the other hand, these boards met the requirements of EN 312 specified for general-purpose particleboard (type P1) in this regard - value required by the standard 0.24 N/mm2.

Fig. 4
figure 4

Internal bond of particleboards depending on the share of walnut shell particles and the type of binding agent: a – strength values with statistical analysis; b – cross-section of the board after the test

Full size image

Based on the results of other works, it can be stated that the IB is usually influenced by a combination of various factors. To explain this phenomenon, the following factors should be considered: (i) the surface properties of raw material, including the interface between the material and the bond line; (ii) the type of adhesive and (iii) the working conditions and process parameters (Dunky and Pizzi 2002; Pirayesh et al. 2013). Thus, the irregular shape and the dimensions are much greater than in the case of wood particles, which means that the walnut shell particles are characterized by a much less developed specific surface area. For this reason, a smaller amount of binding agent was applied on the surface of the shell particles during the gluing operation, which eventually limited the area of bond lines between them and the wood particles. Consequently, it caused a decrease in the bonding quality, which in turn led to a decrease in IB. Moreover, perhaps the lack of proper UF resin adhesion to the walnut shells resulting from their poor wettability also contributed to observed deterioration, which can also be observed in Fig. 3b (Pirayesh et al. 2013). What is also worth noticing, walnut shells have much higher lignin and extractives content than wood. These substances may adversely affect the curing process of the UF resin, which probably led to the weakening of the particle-to-particle bond strength and the overall reduction in IB (Dunky and Pizzi 2002; Ayrilmis et al. 2009; Pirayesh et al. 2013). Furthermore, walnut shells also contain a large amount of ash (almost ten times more than wood), which probably affected the results of IB as well (Pirayesh et al. 2012b). For this reason, it is justified to use pMDI for gluing walnut shells. This type of binding agent is characterized by outstandingly high bonding strength and good water resistance. It has the ability to create covalent bonds with the surface functional groups of lignocellulosic raw materials. The moisture contained in them accelerates their curing (Kawalerczyk et al. 2023). Walnut shells are rich in cellulose, hemicelluloses, lignin, and pectin, which provide a significant number of hydroxyl groups available for bonding (Pirayesh et al. 2012b; Dobrzyńska-Mizera et al. 2019). Because of that, the implementation of pMDI adhesive resulted in a noticeable improvement in IB.

Similarly, as in the case of IB, the reduction in the density of the boards due to the replacement of wood particles with walnut shells had a significantly negative effect on the boards’ bending strength and modulus of elasticity (Fig. 5). Based on the outcomes, it was found that the boards glued only with UF resin, not covered with HDF boards demonstrated very low values of flexural characteristics. They showed a bending strength of 1.56–4.30 N/mm2 and a modulus of elasticity of 490–1350 N/mm2, depending on the share of shells. The highest values were found for the reference boards and the lowest for the variant containing 50% of shell particles. It was also noted that pMDI has considerably improved the flexural strength of the boards.

Fig. 5
figure 5

Bending strength (a) and modulus of elasticity (b) of particleboards depending on the share of walnut shell particles and the type of binding agent

Full size image

According to Cheng et al. (2016), the increase in the internal bonding of materials can also lead to an increase in the bending strength and modulus of elasticity. In the case of using pMDI, the bending strength results were increased by 120% and 170% for the boards containing 25% and 50% shells, respectively. A similar tendency was noted for the modulus of elasticity, where the obtained values were higher by 310% and 210%, respectively. Overall, although gluing the walnut shells with pMDI contributed to the major improvement in investigated parameters, they were still significantly lower than the reference boards. Therefore, it indicates the need for further work to optimize the bending strength and stiffness of particleboards containing walnut shells. The reason for such a significant decrease in MOR and MOE values is the reduced density of the boards (ranging from 500 to 550 kg/m3), their relatively high thickness (19 mm), and the single-layer structure. It is known that the decrease in the board’s density while maintaining a relatively high thickness usually leads to the deterioration in flexural strength. Moreover, partial replacement of wood particles with walnut shell particles also decreased these parameters. This is undoubtedly related to the surface structure, significant size, and high density of the shells. Wood particles characterized by lower density are more susceptible to plastic deformation during mat compression, in turn, increases the contact area between the particles. The flexural strength was probably affected also by the considerably greater thickness of walnut shells, which means they have a lower compression ratio. The negative impact of high-density lignocellulosic materials (including walnut particles) was previously confirmed by the works of Malony (1993) and Pirayesh et al. (2012b). Furthermore, the resultant boards were characterized by a less homogeneous structure with numerous voids visible in the cross-section (Fig. 6). The number and dimensions of these voids can affect both tensile and flexural strength, as shown by the research by Mirski et al. (2020a). All of these factors probably contributed to the observed deterioration. These observations are consistent with the findings of other authors who also investigated the effect of replacing wood particles with different types of nut shells in the production of particleboards. Similar trends were observed for the use of almond shells (Pirayesh et al. 2012b) and a mixture of almond and walnut shells (Pirayesh et al. 2011), hazelnut shells (Barbu et al. 2020) and peanut shells (Güler and Büyüksarı 2011).

Fig. 6
figure 6

Cross-sections of the produced sandwich boards, depending on the proportion of walnut shells in the bio-composite particleboards

Full size image

Such low bending strength and modulus of elasticity of the manufactured boards practically exclude their use as general-purpose boards (type P1) or interior fitments (including furniture, type P2). According to EN 312, the bending strength for P1 boards should be at 10 N/mm². However, for P2 boards, the MOR and MOE values should reach values of 11 N/mm² and 1600 N/mm², respectively. It should be noted that the reference board did not reach the required strength values. This is due to a significant reduction in the density of the boards, as well as the lack of outer layers, which are largely responsible for these parameters. For this reason, it is justified to add the external layers that will significantly increase the strength of boards with this type of structure. In this work, isotropic HDF boards with a high density were used as recommended by Labans et al. (2019). It is worth emphasizing that this type of board is commonly used in the industry for the production of sandwich boards intended for, among others, the furniture industry and structural applications in the construction of partition walls or doors. Figure 7 shows the effect of covering the bio-composite particleboards with HDF boards. As can be seen, the implementation of external layers led to a significant increase in both bending strength and modulus of elasticity of the newly developed boards. After covering, the reference board showed a significant increase in bending strength and modulus of elasticity by approx. 47% compared to the parameters before covering. Similar tendencies were also observed in the case of boards with 25% and 50% shell content, where covering contributed to a considerable improvement in both parameters. Due to the use of HDF boards, bending strength increased more than four times, regardless of the share of the shells. Moreover, in the case of boards containing 25% shells, the modulus of elasticity increased from 922 N/mm2 to 2986 N/mm2 due to the applied external layers. Even better results were noted for boards with 50% substitution, for which MOE reached the value of 2415 N/mm2 compared to 486 N/mm2 before covering.

Fig. 7
figure 7

Bending strength (a) and modulus of elasticity (b) of sandwich boards depending on the share of walnut shell particles and the type of binding agent

Full size image

These results indicate that using HDF boards effectively contributed to a significant increase in bending strength and modulus of elasticity of walnut shell-containing boards, which may be crucial for improving their overall performance in various structural applications. These boards, placed on both sides of the core, increased the strength by making the low-strength bio-composite particleboard an integral part of a much more durable structure. In this case, the external layers absorb and distribute the load and therefore, the implementation of such a solution allow for more effective and sustainable use of the materials with lower strength (Smardzewski et al. 2021). It is also worth noting that although the covering led to a significant improvement in strength, the boards containing shells still demonstrated lower values when compared to the reference one. This suggests that the use of HDF boards can eliminate the negative impact of the shell addition, but at the same time, there is a need for further research on optimizing the composition and structure of experimental boards to achieve an even better strength level. Furthermore, covering the core seems to be a crucial operation in applying particleboards with walnut shells. This approach makes it possible to actually use these materials, which is an important step towards improving the particleboards containing these shells and adapting them to various requirements regarding strength and functionality.

3.2 Water resistance of bio-composite boards and sandwich boards

Considering the significant improvement in the strength parameters of the boards by the addition of pMDI adhesive and their potential use in indoor applications, the thickness swelling (TS) and water absorption (WA) were determined after 24 h of soaking in water. These tests allow for the assessment of the dimensional stability under long-term exposure to water and constitute a crucial element for the analysis aimed at fully understanding the behaviour of these boards in the conditions of variable relative humidity. Based on the outcomes summarized in Fig. 8, it can be noted that the use of 50% walnut shells improved the water resistance of the resultant particleboards. However, based on the ANOVA results, it was found that the swelling of the board containing 25% shells was similar to the case of the reference board. Increasing the share of shells to 50% resulted in a reduction in TS by approx. 17%. A slightly different effect was noted in terms of water absorption because the WA also decreased as the share of shells increased. The addition of 25% and 50% shells led to a significant decrease in WA by approx. 10% and 20%, respectively.

Fig. 8
figure 8

Thickness swelling and water absorption of: a – particleboards; b – sandwich boards

Full size image

The observed decrease in TS and WA values resulted from the shells’ high density and very low porosity. Moreover, walnut shells contain more extractives than pine wood. Among them, there are substances like tannins, pectins, fats, waxes, gums, essential oils, and volatile substances that do not absorb water (Pirayesh et al. 2012b). Moreover, the high content of lignin in shells, estimated at approx. 49% compared to 35% in coniferous wood probably contributed to the improved water resistance as well (Pirayesh et al. 2012b; Sarsari et al. 2016; Barbu et al. 2020). It is worth noting that similar results were also observed by other researchers who used different species of walnut shells in the production of wood-based panels and other composite materials. For example, similar results were achieved by Gürü et al. (2008) in the case of particleboard with walnut shells and fly ash. Pirayesh et al. (2011, 2012a 2012b) and Kowaluk and Kądziela (2014) used walnut shells both alone and in a mixture with almond shells. Barbu et al. (2020) conducted studies using hazelnut shells, while Khanjanzadeh et al. (2014) focused on MDF fiberboard with walnut shells. In addition, Członka et al. (2020, 2021) studied composite polyurethane foams with walnut shells. Markedly, the next step of the research showed that covering the boards contributed to the reduction in their WA and TS. For boards with the maximum share of shells, a significant decrease in both TS and WA by approx. 30% was found. Improved dimensional stability of these boards resulted from the use of HDF boards as external layers. They effectively delay and limit the access of water to the core layer. Therefore, it can be stated that all the variables considered in this research (covering and the addition of shells) work synergistically, creating a protective barrier that minimizes water penetration and improves dimensional stability. Overall, the developed method of producing these bio-composite particleboards, consisting in the substitution of wood particles with shells, application of pMDI adhesive and use of external layers, allowed to achieve very low swelling and water absorption values in the conditions of long-term water exposure.

3.3 Flammability of bio-composite boards

Investigation of fire resistance of manufactured particleboards is an additional aspect of the research. Flammability tests were carried out for the reference particleboard and the board containing 50% shells. Their purpose is to provide basic information regarding the resistance of these boards to fire by analysing the results, i.e., time to ignition, time to flameout, percentage mass loss, heat release rate (HRR), and total heat released (THR). The outcomes presented in Table 1 and Fig. 10 may be useful in assessing the suitability of these boards for the application in the production of, e.g., doors and in the context of overall fire safety, which is crucial for designers, manufacturers, and users of this type of product. Based on the results presented in Table 1, it was found that the reference board showed a longer time to ignition by 26% and a longer time to flameout by 23% compared to the board containing 50% shells. This is due to the differences in the chemical composition of wood particles and walnut shells and the surface structure of bio-composite particleboards. Pine wood consists approx. 42% of cellulose, 23% of hemicellulose and 24% of lignin (Tarasov et al. 2018). Walnut shells consist approx. 24% of cellulose, 22% of hemicellulose and 48% of lignin (Nishide et al. 2021). It is also known that lignin is more stable in thermal degradation. The first step of degradation of walnut shells corresponds to heating and evaporation of water and extractives and occurs in temperatures to 160 °C; the second step (160–280 °C) corresponds to pyrolysis of cellulose, hemicellulose and a part of lignin. Lignin is mainly pyrolyzed at temperatures above 350 °C (Shanfeng et al. 2020). The pyrolysis products of cellulose and hemicellulose mainly consist of gases; thus, ignition occurs rapidly. During the tests, ignition time of REF boards was longer because the range of thermal decomposition of wood is a little higher than in walnut shells and starts at 200 °C for hemicelluloses, 250–380 °C for cellulose and 450 °C for lignin (Dietenberger and Hasburgh 2016). The surface of the shell-containing boards was characterized by greater irregularity compared to the reference one (Fig. 9). Although both variants demonstrated similar density, the experimental boards showed numerous voids in their structure. Their presence means that more air in the material can lead to the acceleration of ignition and the increase in the THR and HRR. The values of THR and percentage mass loss in the case of the reference variant were comparable but slightly lower by about 3% and 6%, respectively (Fig. 10a).

Table 1 Results of fire resistance test
Full size table
Fig. 9
figure 9

Comparison of the surface irregularities of the reference board with the board with 50% walnut shells share

Full size image
Fig. 10
figure 10

The course of THR (a) and HRR curves (b) obtained for tested particleboards

Full size image

Based on the analysis of the HRR curves and the maximum values of HRR peaks (Fig. 10b), it was found that the board containing shells showed a higher peak after ignition and the maximum HRR peak. After ignition, there was a rapid increase in the HRR to approx. 161.5 kW/m2 and 144.5 kW/m2 in the case of experimental and reference boards, respectively. Furthermore, the maximum HRR peaks showed the same tendency, and they were 245.1 kW/m2 and 215.7 kW/m2. In the case of the reference particleboard, the time to reach these peaks was longer, and it was 170 s and 848 s, while for the experimental board, it was 105 s and 670 s. Nevertheless, there is a need for further research and analysis to understand better the mechanisms influencing these changes and to assess the potential of these boards in the context of fire protection.

Research by Mancel et al. (2022) on the use of rubber waste (tires and insulators) as fillers in the production of particleboards showed that the ignition times of boards with additives of 10%, 15% and 20% of rubber were similar to those of particleboards alone. The average time to ignition was from 298 s to 309 s. Similarly, the weight loss for the board containing 10% rubber waste additives was similar to that of the particleboard. It was found that boards with rubber waste have similar flammability parameters to traditional particleboards.

Experiments on producing boards based on coconut fiber with modified starch as a binder showed increased fire resistance of these materials. The results obtained using the Limited Oxygen Index (LOI) method indicate that the fire does not spread and that the material is classified as self-extinguishing. The LOI values obtained in the tests ranged from 29.3 to 31.3%, indicating good fire-retardant properties, mainly influenced by the coconut fiber content (Owodunni et al. 2020).

Research provided by Gürü et al. (2015) regarding the use of rice husks and straw in the production of particleboards showed that adding these lignocellulosic wastes increases the fire resistance of the board. The experiments were conducted at different rice husk/urea-formaldehyde ratios from 1.3 to 2. LOI tests of the produced boards were performed. Adding rice husks and straw as a filler to the particleboards resulted in obtaining a non-flammable particleboard. From the combustion test results, it can be concluded that rice straw has high fire resistance, and the LOI increased by increasing the proportion of filler material and polymer binder. Rice straw contains inorganic compounds, mainly silica, which have the greatest impact on the fire properties of the boards.

Ferrandez-Villena et al. (2020) researched using vine branches to produce particleboards. The urea-formaldehyde resin was used as a binding agent of 9% by weight. Eight variants of boards were made, differing in the proportion of particles of different sizes. Reaction to fire tests (flame spread Fs) were performed to classify the obtained materials. Particleboards have been classified by applicable European regulations. Similar Fs results were obtained for all tested boards. The rules state that the boards are classified as B when Fs < 150 mm in 60 s. Therefore, particleboards for vine pruning are classified as B-s2 d0. Particleboards without additives have a Dd0 class, which means that their reaction to fire is worse than those made from vine shoots produced in this work. An explanation may be the silica content of the material.

4 Conclusion

The research concerned the possibility of manufacturing the bio-composite particleboard with a reduced density ranging from 500 to 550 kg/m3, differing in the share of walnut shells (0%, 25%, 50%). Moreover, it investigated the possibility of using this type of board as a core layer of three-layer sandwich boards. As expected, the partial replacement of pine wood particles with walnut shells causes a significant decrease in the strength of UF resin-bonded particleboards. However, a significant improvement in strength parameters can be achieved by implementation of a hybrid gluing system which involves gluing of wood particles with UF resin and shells with pMDI adhesive. Implementation of this approach leads to significant increase in the board’s internal bond by 58% and 110% for variants containing 25% and 50% shells, respectively. There was also a major improvement in bending strength and modulus of elasticity. Compared to the boards bonded only with UF resin, the increase in MOR and MOE was 120% and 310% for boards containing 25% shells and 170% and 230% for boards containing 50% shells. Despite such a noticeable improvement in flexural characteristics, the strength of produced boards still does not reach the level required by the standard for furniture boards. However, it has also been shown that these boards can be used as the core of sandwich boards covered with thin HDF boards, and the resultant structure shows good strength and meets the standard’s requirements. Furthermore, it was found that as the shell content increases, the dimensional stability, determined based on the results of thickness swelling and water absorption, improves. In addition, the use of walnut shells (50%) accelerates the board’s times to ignition and flameout, increases the total heat released and heat release rate, but does not affect the percentage of weight loss in a significant way.