Re-Resinated Wood Strand Panels: Enhancing Performance Through Waste Recycling

Abstract

The construction sector’s increasing eco-consciousness is driving the need for higher-performance wood-based engineered products from underutilized timber resources, such as small-diameter trees from hazardous fuel treatments of our forests. Strand-based products, including oriented strand board (OSB) and lumber (OSL), are widely used. However, hot-pressing during their manufacturing generates approximately 10% waste, which includes a substantial amount of resinated strands that are landfilled. The huge potential of using strand-based products has led to many studies and growing interest in strand-based three-dimensional sandwich panels that can be used as wall, floor, or roofing panels. As the market grows, understanding the recyclability of these resinated strands becomes crucial. This study investigates the feasibility of using re-resinated waste strands that were collected during lab-scale production of strand-based panels. Results demonstrate significant improvements in dimensional stability, mechanical properties, and fire resistance. Specifically, recycling increased internal bond strength, flexural strength, time to ignition, time to flameout, mass loss, and time to peak heat release rate by 107%, 44%, 58%, 35%, 51%, and 27%, respectively, and helped decrease water absorption and thickness swell by 51% and 58%, respectively.

1. Introduction

The wood industry has strategically capitalized on the abundance of small-diameter timber and forest thinning residues in US and European forests, leading to the development of oriented strand board (OSB) [1] and lumber (OSL) [2,3]. Strand-based composite materials offer lightweight, high-performance alternatives for structural applications [4]. While wood-based composites have a long history [5,6,7], recent industrial advancements have focused on industrial chemicals, biofuels, and nanocellulose [8,9]. Natural fiber composites (NFCs) have seen significant adoption due to their advantages over synthetic fibers, including lower weight and cost, abundance, reduced energy consumption, and safer handling [9]. These advantages and interests have resulted in NFCs successfully penetrating many markets including aerospace [10], medical [11], automotive [12], food [13], electronics [14], and building materials [15]. Due to their sustainability, reduced carbon dioxide (CO₂) emissions, and advantages over synthetic materials, natural fibers, and NFCs have gained significant attention from researchers and industry. Despite wood’s inherent advantages, including superior structural properties, its potential has been underexplored. The current wood product manufacturers produce waste in different forms, which can range from sawdust, wood shavings, to strands, veneers and even lumber pieces. Though most of these have an established recycling market and process, resinated strand waste is generally treated as landfill, forming the primary focus of the current study.

Mass timber panels, such as cross-laminated timber (CLT), have spurred renewed interest in wood composites [16,17]. The success of utilizing low-quality forest resources for high-value building materials has prompted studies on strand-based corrugated core sandwich panels for walls, floors, and roofs with both functional and structural advantages [4,17,18]. A standard OSB mill has a solid waste of about 1.63 kg/m³ [19], even after ensuring the majority of the waste has been utilized. This number would be higher for start-ups, where strand-based technologies would yield about 5–10% of solid waste, which can amount to about 300 kgs/day for a small production facility with an average capacity of about 733 m³ per annum, producing about 100 panels daily, each measuring 1.22 m × 1.44 m × 0.009 m, in a single daylight press at 75% capacity. These numbers have been developed through calculations made by the authors for current industrial partners interested in strand-based products. For comparison, a standard OSB manufacturing plant has a capacity of 7.3 million m³ per year, which is conventionally measured at 75% of the plant’s actual capacity [19]. This waste, having almost 50–70% of usable strands, contains feedstock for almost 10 or more panels, where about 50% are resinated strands produced from blending, mat forming, and pre-forming stages. Wastage could be significantly higher with increased production capacities. Current practices involve using the waste biomass as landfill due to the presence of resin, but the cost is substantial [20]. Addressing the need to investigate the reusability of waste strands, this study explores recycling and sieving to create new panels, potentially increasing carbon sequestration and enabling zero-waste production.

A review of the existing literature shows a lack of data in understanding the reusability and recyclability of waste resinated strands for use in composite panel products. To understand their reusability, it is critical to understand the performance of the panels made from recycled strands, compared to those made from virgin strands (referred to as control strands in this study). Addressing the need to investigate the reusability of waste strands, this study explores recycling and sieving to create new panels, potentially increasing carbon sequestration and enabling zero-waste production. The research compares the performance of panels made from recycled strands to those made from control strands. It focuses only on recycling resinated waste strands collected during blending, mat-forming, and pre-forming stages in making standard boards in lab-scale facilities, which has also been used as the control for the current study, assessing the impact of the initial adhesive application on panel properties. The primary objectives are to:

• Assess the recycling effects of re-resinated strands on panel performance, including bond performance among the strands and mechanical properties.
• Understand the influence of recycling on the dimensional stability of the panels.
• Compare the fire performance of the panels made from control and re-resinated strands.

2. Materials and Methods

2.1. Materials

Low-quality Englemann spruce and lodgepole pine (ESLP) lumber, acquired from Idaho Forest Group, with an average density ranging between 0.35 and 0.38 g/cm³, were used to produce the required wood strands. A CAE Disc Strander (Kadant Int., Surrey, BC, Canada) available at Washington State University’s (WSU’s) Composite Materials and Engineering Center (CMEC) was used for processing the strands. Due to the maximum size limitation of the lab-scale CAE strander, the lumber needed sectioning at 140 mm before being soaked in water for three days to reach a moisture content higher than 50% before stranding. The strander has a rotational speed of 500 revolutions per minute, and the blade gap was adjusted to produce strands that were approximately 140 mm in length, 19 to 38 mm in width, and 0.38 to 0.51 mm in thickness. The strands were initially air-dried and then conditioned to ~10–12% moisture content for three to four weeks before fabrication.

2.2. Fabrication Process

Flat panels with final trimmed dimensions of 1220 mm × 2440 mm × 6.35 mm were produced in a hot-press (Iron Works, Warwick, RI, USA) at WSU’s CMEC. The panels were consolidated to a target density of 640 kg/m³, with strands generally oriented along the major axis or the longer panel dimension. Polymeric methyl diphenyl diisocyanate (pMDI; product code Rubinate1840 supplied by Huntsman Corporation, The Woodlands, TX, USA), at a loading amount of 6 wt.% of oven dry wood weight, was used as the binder and the panels were hot-pressed at a platen temperature of 171 °C. The PressMan software (Version 0.78), integrated with the hot press, was used to develop the press schedule where a consolidation time at the final thickness of 6 min was used. During the process of blending, forming, and hot-pressing panels, there are generally left-over resinated strands that are usually considered waste and disposed of as landfill. For the current study, these waste resinated strands, acquired only post blending, mat-forming, and preforming processes, were accumulated and sieved to ensure no fines and dust particles were present. The sieved strands were then conditioned in a chamber at 65% relative humidity and 21.4 °C for 30+ days to emulate the longest possible interval between initial use and re-use times. A similar panel consolidation process was then followed to hot press 6.35 mm thick flat panels, which were 2440 mm long and 1220 mm wide; strands were again oriented with respect to the major axis or the long dimension of the panel. A 6% pMDI resin content was used to re-resinate the strands. A mat was formed and hot-pressed with the same press schedule and temperature as with virgin strands. The entire process is illustrated in Figure 1. A vein box with a spacing of 76.2 mm between the adjacent veins (Figure 1) was used for orienting the strands. It should be noted that studies [21] have shown that the ideal gap between the vein box and the surface of the mat being formed should be maintained at around 76.2 mm during the entire mat forming process. However, in this study, the vein box was maintained at a fixed distance, thus this gap was not constant. This directly influenced the overall orientation of the strands, with the strands in the lower part of the panel having a more random distribution. The orientation had a profound influence on the mechanical properties of the panels, especially tensile properties. Thus, the study aims to provide a comparison of all the properties, where the forming technique was constant for fabricating both panel types, with virgin and re-resinated strands.

2.3. Experimental Methods

The influence of reprocessing the re-resinated strands was evaluated by testing specimens from hot-pressed panels for their dimensional stability, mechanical performance, and fire performance. All the tests were carried out as per ASTM standard specifications. Specimens were conditioned for about three to four weeks before testing at 20 °C and 65% RH to achieve an equilibrium moisture content of 12% [22].

2.3.1. Dimensional Stability and Moisture Interaction

Dimensional stability is crucial for wood and wood-composite panels due to wood’s inherent hydrophilic nature [23]. This limits the applications of wood-based materials, particularly in high-moisture environments [6]. Understanding the interaction between panel surfaces and water is essential for determining panel wettability and overall performance. Research has shown that resin content directly influences the hydrophilic nature of the final panel, with higher resin content resulting in increased hydrophobicity [24]. To assess surface wettability, the surface contact angle and surface tension using the sessile drop method was calculated, following the ASTM D7490-13 [25] standard. The process involved using a VCA Optima (AST Products, Inc., USA) video system. Employing a Hamilton 81020 syringe (100 µL capacity), deionized water was dispersed uniformly on the specimen surface. Five specimens for each panel type were tested. Eventually, to understand the performance of the entire panel when exposed to a high moisture environment, Method A of ASTM D1037-13 [26] standard was employed to estimate the water absorption (WA) and thickness swell (TS) of the specimens. Again, five specimens of each panel type were tested, and the results were calculated after 2 h and 24 h of complete immersion in water. Five points on each specimen were pre-marked to measure the thickness swell, before and after water soaking. Deionized water was again utilized, and the total moisture uptake after 24 h was calculated according to ASTM D4442-20 [27] standard, after oven drying for 24 h at 105 °C. All the WA, TS, and moisture uptake results were plotted in percentage increase from the conditioned samples.

2.3.2. Mechanical Performance

Vertical density profiles (VDP) of both panel types were analyzed using an X-ray vertical density profiling machine (QMS Density Profiler Model QDP-01X by Quintek Measurement Systems, Inc., Knoxville, TN, USA) at CMEC, WSU. Specimens measuring 50.8 mm by 50.8 mm were tested to examine density variations along the panel thickness. Ten specimens from each panel type were evaluated. Studies have demonstrated that resin content plays a significant role in determining the mechanical performance of wood-composite panels. Therefore, following non-destructive testing for panel density through thickness, the same specimens were subjected to internal bond (IB) strength analysis according to ASTM D1037-13 standards [26]. With a testing speed of 0.508 mm/min, based on the 6.35 mm panel thickness, IB strength was assessed as a critical parameter for ensuring proper curing and bonding of pMDI resin during production. Results were compared between the two panel types to evaluate the impact of strand re-resinating on bond performance.

In addition to IB strength, tensile and bending properties were analyzed to assess the load-bearing capacities of the panels for design purposes. ASTM D1037-13 was used to determine tensile and flexural properties in both longitudinal and transverse directions. Five replicates of each panel type were tested for each property. Dog bone specimens were cut from panels using a cutting jig on a router machine, with dimensions of 254 mm in length and 50.8 mm in width, featuring a reduced cross-section and a gauge length of 50.8 mm with 70 mm grip lengths on each side. A universal testing machine (UTM) from Instron^® was employed at a crosshead speed of 4 mm/min to calculate tensile strength and Young’s modulus using standard equations from ASTM D1037-13. The same standard was applied to determine the modulus of elasticity (MOE) and modulus of rupture (MOR) through midpoint bending tests. Specimen spans were set at 48 times the panel thickness, with a width of 50.8 mm and a loading rate of 12.2 mm/min, utilizing the same UTM for testing.

The nail withdrawal properties of the panels were also examined to evaluate fastener holding capabilities and mounting performance. This analysis followed ASTM D1037-13 standards using an 8d common nail driven completely through panel samples measuring 76.2 mm by 152.4 mm, with seven replicates per panel type. The nail depth corresponded to the panel thickness, and withdrawal testing was conducted at a crosshead speed of 1.5 mm/min using the UTM. To contextualize these results, equivalent specific gravity (ESG) values were calculated for comparison with established engineered wood products. ESG values were derived using empirical equations from the Wood Handbook [28] and National Design Specification [29], correlating specific gravity with mechanical properties to benchmark the new material against traditional timber products. The equation can be given as follows:

$$ES{G}_{nail}={\left({\displaystyle \frac{P_{nail}}{54.12\times D\times L}}\right)}^{{\displaystyle \raisebox{1ex}{$2$}\!\left/ \!\raisebox{-1ex}{$5$}\right.}}$$

(1)

where P is the withdrawal strength; D is the diameter of the nail; and L is the length of nail embedment. The calculated analytical value would also help establish the functionality of the panels, compared to standard strand-based panels in the market, like OSBs. These analyses provide comprehensive insights into the mechanical performance of wood-composite panels, including their bonding quality, load-bearing capacity, and fastener retention properties, enabling practitioners to evaluate their suitability for various applications while comparing them against established materials through ESG calculations.

2.3.3. Fire Performance

The fire reaction properties of the panels were analyzed to evaluate how strand recycling and resin reapplication affected thermal stability and combustion behavior. Thermogravimetric Analysis (TGA) and Derivative Thermogravimetric Analysis (DTGA) were initially conducted following ASTM E1131-20 [30] (using TGA/DSC1 STAR^® System by Mettler Toledo, Columbus, OH, USA). The specimens underwent controlled heating at 10 K/min in an inert atmosphere, with three replicates per panel type tested to track normalized mass loss and decomposition rates as functions of temperature. Results revealed compositional variations between recycled and virgin materials through distinct thermal degradation patterns.

Mass-loss calorimetry tests, following ASTM E2102-17 [31] and ISO 17554:2014 [32] standards and using GBH International Corp equipment at the University of Idaho, were conducted. Five replicates of each panel type were exposed to 50 kW/m² heat flux at 25 mm from a cone heater, with a 3 mm spark gap positioned 13 mm above the samples. Heat release rate (HRR) profiles were recorded for 120 s post-flameout to capture sustained combustion behavior, ensuring stable data for comparison. The tests were conducted using a calibrated mass loss calorimeter, with a calibrated load scale, ensuring the accuracy and repeatability of the results of how recycling influenced flammability characteristics through mass loss changes and HRR trends. The distance between the specimen and the cone heater was maintained at 25 mm, and a heat flux of 50 kW/m² was used. A spark plug was used with a 3 mm spark gap, located 13 mm above the center of the sample.

3. Results and Discussions

The experimental results and discussion are most effectively organized based on the three distinct characterization methods that evaluate the feasibility of recycling waste strands generated during the fabrication of strand-based composite panels. By comparing these findings with the control, valuable insights can be gained into the impact of recycling re-resinated strands on the final properties of the composite panels.

3.1. Dimensional Stability

To investigate the hydrophobicity of the panels and assess the impact of recycling re-resinated strands, water interaction with the panels was analyzed using contact angle measurements. As shown in Figure 2a, panels fabricated from re-resinated strands exhibited a higher average contact angle (100.15° ± 4.1°), representing a 4% increase compared to panels made from control strands (96.28° ± 3.1°). However, a single-factor ANOVA analysis (α = 0.05) revealed that this variation was not statistically significant, with a p-value of 0.18. Similarly, surface tension measurements showed a 12% difference between the panels, which was also statistically insignificant. Panels made from re-resinated strands demonstrated lower surface tension compared to control panels, agreeing with findings from previous studies comparing hydrophilic and hydrophobic surfaces [33]. These results suggest that while recycling may slightly influence surface properties, the observed differences are not substantial enough to be statistically significant.

The WA and TS results after 24 h are presented in

Table 1. Experimental values of water absorption and thickness swell tests.

Panel Type	Water Absorption at 24 h			Thickness Swell at 24 h
	Mean (%)	Standard Deviation (%)	Coefficient of Variation (%)	Mean (%)	Standard Deviation (%)	Coefficient of Variation (%)
Control Panel	44.38	6.4	14.4	15.54	1.96	12.6
Re-Resinated Panel	21.65	4.5	20.6	6.46	1.03	16.02

. These findings demonstrate that recycling the strands significantly improves the dimensional stability of the panels, with approximately a 51% reduction in WA and a 58% reduction in TS. The TS/WA ratio, shown in Figure 2b, further highlights this trend, as a lower ratio indicates improved dimensional stability [34]. Panels made from re-resinated strands exhibited a reduced average TS/WA ratio compared to those made from control strands; however, statistical analysis revealed that this variation was not significant, with a p-value of 0.19, as illustrated by the box plots in Figure 2b.

3.2. Mechanical Properties

The density of the panels had negligible variations, with a slightly higher average value for the panels made from re-resinated strands, as shown in Figure 3a. The use of re-resinated strands also resulted in reduced variability in density values, as shown in the same figure.

Table 2. Experimental values of the through-thickness density and internal bond strengths for the two panel types.

Panel Type	Through Thickness Density			Internal Bond Strength
	Mean (kg/m3)	Standard Deviation (kg/m3)	Coefficient of Variation (%)	Mean (MPa)	Standard Deviation (MPa)	Coefficient of Variation (%)
Control Panel	683.6	123.1	18.01	0.76	0.15	19.4
Re-Resinated Panel	735	93.2	12.68	1.57	0.28	18.2

illustrates the density values of the panels, with the control panels having a density of 683.6 ± 123.1 kg/m³, compared to 734.97 ± 93.24 kg/m³ for the panels made from re-resinated strands. The ANOVA, however, confirmed that the variation between the values is not statistically significant, with a p-value of 0.68. Therefore, it can be concluded that the variations in the following mechanical properties, if any, will not be influenced by the variations in the panel densities. Additionally, the average densities of the panel types were further normalized to prove that the statistically significant variations were not driven by increased density in the panels made from the re-resinated strands. Equation (2) illustrates the formula used to calculate the equivalent mechanical properties of the panels, which can be stated as follows:

$$X_{C_E}={\displaystyle \frac{{\rho }_{RR}}{{\rho }_C}}\times X_C$$

(2)

where $X_{C_E}$ represents the equivalent average value of the concerned mechanical property in the control panels; $X_C$ represents the experimental average value of the same property in the control panels; ${\rho }_{RR}$ is the density of the average density of the re-resinated panels; and ${\rho }_C$ is the average density of the control panels.

The IB study demonstrated significantly improved strength in panels made with re-resinated strands compared to controls, as shown in Figure 3b. This improvement is directly linked to residual pMDI resin retained in the strands even after 3–4 weeks of conditioning. The higher resin content facilitated stronger interfacial bonding during curing, likely due to increased adhesive distribution per unit cross-sectional area of the strands. Further chemical analysis of the strands might help establish the claim. Re-resinated strand panels exhibited a 107% increase in the IB strength relative to the control panels (Table 2). ANOVA confirmed these differences as statistically significant with an extremely low p-value (approximately 0). These results support the value of recycling the waste strands during panel processing by re-resinating and reusing the strands for composite panels. This approach offers a sustainable pathway for optimizing composite panel production without compromising quality. Equivalent IB strength in the control panels, calculated as per Equation (2), was 0.81 ± 0.16 MPa, which was still significantly lower than those obtained from the re-resinated panels.

The influence of tensile strength was analyzed to assess the effect of recycling re-resinated strands on the tensile properties of the panels. Although the random orientation of strands in the lower mat limits the representativeness of absolute tensile strength and stiffness values, the study provides valuable insights. Panels made from re-resinated strands exhibited higher average tensile strength compared to control panels, with the strength increasing approximately by 40% along the major axis and 11% perpendicular to it. Equivalent average tensile strengths of the control panels were also calculated to be 30% and 3% lower than those of the re-resinated panels parallel and perpendicular to the major axis, respectively. Similarly, modulus values were about 12% and 30% higher (5% and 23% higher than the equivalent moduli) for re-resinated strand panels in the respective directions. However, a single-factor ANOVA revealed that these variations were not statistically significant, with p-values of 0.12 and 0.54 for the tests along the major and perpendicular axes, respectively. A cross-comparison between the tensile strength of recycled panels perpendicular to the major axis and control panels along the major axis also showed no statistical significance (p = 0.72). Higher variability in control panel data was attributed to density fluctuations, as illustrated in Figure 4a,b. Importantly, it was assumed that the strand length was not reduced during recycling, resulting in no statistically significant variations in the tensile properties, as the strand length is known to significantly influence panel properties [35]. Overall, these findings indicate that recycling and re-resinating strands had no statistically significant impact on the tensile properties of the panels in any direction.

The flexural comparison yielded similar results when tested in the longitudinal direction along the major axis, as shown in Figure 5. All flexural specimens failed in tension, and panels made from re-resinated strands exhibited average flexural strength and modulus values approximately 8% and 14% higher, respectively, compared to the control panels. These higher properties can be attributed to the residual resin content in the recycled strands and higher IBs. However, a single-factor ANOVA indicated that these variations were not statistically different, with p-values of 0.41 for strength and 0.37 for modulus, both exceeding the alpha threshold of 0.05. Therefore, the slight variations can also be attributed to the density variations in the panels. Considering the equivalent flexural properties of the control panel, as per Equation (2), the variation due to density can be further established because the average strength of the re-resinated panels was less than 1% higher with the average flexural modulus being about 6% higher when compared to the control panels. In contrast, testing in the transverse direction (perpendicular to the major axis) revealed statistically significant enhancements in flexural properties for panels made from re-resinated strands. Flexural strength increased by approximately 44% (34% when equivalent strength was considered), while flexural modulus improved by about 56% (45% for equivalent modulus of the control panels), demonstrating significant improvements in panel performance. This contrasting variation between the two directions can be attributed to possible variation in the fiber direction, where the strands for the re-resinated panels can be oriented more randomly due to human mat-forming variations and errors. Increased IB strength should ideally help enhance the flexural properties, which is again an indication of random strand orientation. Further repetitive studies might be required to establish the actual values. However, the current findings suggest that re-resinating strands can produce similar and even higher-quality panels and offer greater potential for optimizing adhesive usage during recycling to achieve comparable properties to those of control panels, further contributing to cost-efficiency and sustainability.

Nail withdrawal tests were conducted to evaluate the withdrawal strength of the panels, a critical property for applications in the built environment. The results, shown in Figure 6, indicate that panels made from re-resinated strands exhibited a screw withdrawal strength of 35.4 ± 8.6 N/mm, compared to 27.4 ± 6.4 N/mm for the control panels and 29.46 ± 6.9 N/mm for equivalent strength in the control panels using Equation (2). Despite this improvement, a single-factor ANOVA revealed that the variation was not statistically significant (p = 0.11). The ESG values were also calculated for both panel types using Equation (1), with results summarized in

Table 3. Experimental values of the nail withdrawal test and the calculated equivalent specific gravity.

Panel Type	Nail Withdrawal Strength
	Mean (N/mm)	Coefficient of Variation (%)	Equivalent Specific Gravity
Control Panel	27.4 ± 6.4	23.27	0.43 ± 0.04
Re-Resinated Panel	35.4 ± 8.6	24.37	0.48 ± 0.05

. Panels made from recycled strands showed an ESG of 0.48 ± 0.05, approximately 11% higher than the control panels, which had an ESG of 0.44 ± 0.04, with negligible change in the average ESG value when equivalent strength of the control panels was considered in Equation (1). These values are comparable to or slightly higher than those of similar strand-based structural panels; for example, OSBs typically have an ESG of around 0.4 [36], while OSLs exhibit a slightly higher ESG of approximately 0.51 [37]. It must be noted that all the tests were conducted only in the face direction. These findings demonstrate that panels made from recycled strands achieve high-quality performance and are potentially suitable for structural applications in the built environment. Furthermore, the overall mechanical performance comparisons confirm that waste strands can be effectively reused to manufacture panels without negatively impacting their mechanical properties, where, in some cases, properties such as internal bond strength, withdrawal strength, and ESG can even be improved.

3.3. Fire Reaction Properties

The influence of thermal stability due to recycling and re-resinating the strands was established using the TGA study, as illustrated in Figure 7. TGA was used to generate mass loss curves, revealing that the re-resinated samples exhibited higher mass retention (25%) and lower overall mass loss compared to the control panels, which showed a mass retention of 22%. This improved thermal stability can be attributed to the higher pMDI content present in the strands during the panel-forming process. The DTGA study further showed that the reference temperature or peak mass loss temperature (T_R) was 354.17 °C for the control panels compared to 349.67 °C for the panels made from re-resinated strands. However, this variation of about 1.4% was not statistically significant. Therefore, the TGA and DTGA studies helped prove that recycling of the strands had no significant influence on the thermal stability of the final panels. The increase in mass retention, however, would have a direct influence on the fire reaction properties of the samples, which were observed to be significantly better in many aspects tested. The primary reaction that helps the pMDI adhesive bond the strands together to form a panel is the one between the adhesive and moisture [38]. One molecule of isocyanate in pMDI reacts with moisture (H₂O) and forms carbamic acid, carbon dioxide (CO₂), and amine [38]. The amine then reacts with another isocyanate molecule to form urea. Urea is known to have excellent fire resistance, where it can work as a filler and promote char formation. Studies [39] have shown that in wood/pMDI curing systems biuret, polyurethane, and polyurea are commonly formed. All these compounds are known to have flame-retardant properties [38,39,40,41], which directly influence the performance of the panels made from re-resinated strands in the current study.

Figure 8 represents the HRR and mass loss trends of the two panels when tested in the mass loss calorimeter. The HRR trends, shown in Figure 8a, represent those similar to wood and wood-based composites, where two peaks are commonly seen [42]. Wood-based materials have been observed to have two peaks, where generally the initial peak is lower and represents the release of heat when the sample catches fire. The inherent char forming ability of wood then reduces the heat release, which peaks again later once the char is broken [6]. A similar trend was observed in the current study as well, where the higher amount of pMDI in the re-resinated strand-based panels resulted in longer heat suppression, due to a higher amount of Urea, resulting in a later peak heat release rate (PHRR). The longer heat suppression also resulted in lower fuel availability, resulting in lower HRR peaks. Figure 9a shows that the time to PHRR was almost 27% longer for the panels made from recycled strands, compared to the control samples. Figure 9b further illustrates that the PHRR for the control panels was 294.32 ± 10.1 kW/m², compared to 217.88 ± 19.3 kW/m² for the panels made from re-resinated strands, which is about 26% higher. A single-factor ANOVA also proved that the variations were statistically significant for both the PHRR (p-value = 0.004) and time to PHRR (p-value = 0.01) results.

The time to ignition, which helps understand how quickly the sample catches fire under standard spark-enforced conditions, was also significantly higher for the panels made from recycled strands. The presence of a higher amount of urea helps impart greater fire resistance to the samples resulting in taking about 58% longer to ignite, as can be observed from Figure 9a. This demonstrates the improved fire reaction properties of the panels, where the variation was statistically significant, with a p-value of 0.000013. The mass loss rate, as can be observed from Figure 8b, showed higher mass retention for the panels made from recycled strands where the final mass after flameout was observed to be 22.07 ± 0.9%, as shown in Figure 9a. The control panels, however, showed a higher amount of mass loss, as can be observed from the mass loss trends illustrated in Figure 8b, with a lower mass retention of 10.81 ± 1.1%. The variation of about 51% was also observed to be statistically significant, with a p-value of 0.0002, establishing that the panels with recycled strands had higher char formation due to the presence of a higher amount of urea, resulting in an observation that was analogous to the TGA results.

The time to flameout was also observed to be significantly higher for the panels with re-resinated strands, as can be observed from Figure 9a. The control panels had a time to flameout of 306.67 ± 17.8 secs, resulting in a burning time of 289 ± 17.7 secs. The panels from recycled strands had a flameout time of 472.33 ± 18.6 secs, resulting in a burning time of 430 ± 18.2 secs. Higher burning time further proves that the char formation was higher in the panels with re-resinated strands, which suppressed the heat for a longer time and resulted in a comparatively more fire-safe structure. The ability to sustain an induced flame for longer times helps produce panels that can maintain their structural integrity longer, and if applied in the built environment, can be denoted as a safer structure where inhabitants would have more time to evacuate in any kind of fire scenario. Thus, a higher amount of pMDI helps make structures comparatively safer under fire. The variation in flameout time was calculated to be about 35%, which was also established to be statistically significant with a p-value of 0.0004. Total heat evolved (THE), which can also be called the peak total heat released, is known to directly influence the surrounding temperature and conditions of the structure under fire. This factor is critical during evacuation, as higher heat might cause detrimental effects to human evacuation activities. The results showed that the total heat evolved was almost similar for both samples, with a slightly higher value for the panels made from re-resinated strands, as shown in Figure 9b. The variation was, however, not statistically significant, with a p-value of 0.08. Additionally, THE for the re-resinated panels, although higher, took about 33% longer time to reach compared to that of the control panels. In any cone testing standard apparatus, THE is generally reached right at the time of flameout, as it represents the cumulative addition of all the heat evolved during the burning phase. Therefore, a comparison at the same time reveals that THE in the control samples around flameout (305 s) was 53.82 ± 3.2 MJ/m², compared to only 36.71 ± 4.9 MJ/m² for the panels made from recycled strands at 305 s. This variation of about 32% was observed to be statistically significant with a p-value of 0.006. Therefore, it can be safely concluded that the panels made from re-resinated samples had lower THE compared to the control at any time of the burning phase. The overall fire performance analysis revealed that recycling the strands and subsequent re-resination significantly enhanced the panels’ fire reaction properties. This improvement can be directly attributed to the increased availability of urea, resulting in higher, thicker, and more stable char, due to the higher pMDI content in the re-resinated panels.

4. Conclusions

This study comprehensively evaluated the physical, mechanical, and fire performance of panels incorporating recycled pMDI-resinated strands, a material often discarded as waste during strand-based panel production. Notably, even after three to four weeks of conditioning, recycled strands retained residual pMDI resin, and when combined with re-resination, this resulted in panels with elevated adhesive content. This enhancement led to substantial improvements in water absorption (reduced by 51%) and thickness swell (reduced by 58%) compared to control panels. While improvements in other dimensional stability metrics (such as contact angle, surface tension, and TS-to-WA ratios) were observed, these differences were not statistically significant.

Importantly, panels made from re-resinated strands demonstrated markedly superior internal bond strength, indicating improved inter-strand adhesion. Flexural properties in the transverse direction (parallel to the major strand orientation) significantly improved, with flexural strength and modulus increasing by 44% and 56%, respectively. Although the overall mechanical property variations were not statistically significant, the consistently higher average performance suggests that panes made through re-resinating recycled strands hold promise as products that can not only be used in equivalent markets but also in better premium markets.

In terms of fire performance, recycled- and re-resinated-strand panels exhibited notable advantages. Despite no significant differences in thermal stability, re-resinated panels showed approximately 12% lower mass loss. More critically, fire performance metrics—including time to ignition, peak heat release rate, total heat evolved, and time to flame-out—all pointed to superior fire resistance. This indicates the potential for using recycled-strand panels in fire-safe construction applications, adding a valuable safety dimension to their reuse.

Overall, these findings highlight that panels made from recycled strands not only match but, in some respects, exceed the quality and performance of panels made from virgin material. Therefore, not only can these panels be used for the same market, but the study also opens promising opportunities for manufacturers to market these panels as a premium alternative, potentially offsetting the costs of re-resination. To fully realize this potential, future research should focus on structural applications, long-term performance assessments, and detailed techno-economic analyses to validate the industrial viability of recycled-strand panels. Additionally, optimizing the resin content during secondary resination could help balance product quality with cost efficiency—a critical factor for large-scale commercial adoption.