The WPF-RPs required for TPBs were produced and provided by a manufacturing company in Korea, which has facilities for visual sorting, magnetic sorting, high-rate shredding, melting, and compression molding. The company produces products such as street tree guards and vegetation mats (Ministry of Environment (MOE), 2020b). The dominant type of WPF from household wastes was polyethylene. TPB prototypes, with dimensions 0.3 (W) × 0.3 (L) × 0.125 (D) m, were produced using a newly designed mold. A picture of the TPB unit is provided in Supplementary Figure S1.

Test specimens were prepared from the WPF-RPs, depending on the standard size. Mechanical properties of WPF-RPs as construction materials, were measured and evaluated, including compressive and tensile strengths, compressive and tensile elasticities, Poisson's ratio, density, and coefficient of thermal expansion. Seven specimens each were prepared for testing the compressive strength, compressive elastic modulus, tensile strength, tensile elastic modulus, and Poisson's ratio, while five specimens each were prepared for density and thermal expansion measurements. The tests were conducted according to the procedures given by the Korea Standard Methods (Korean Industrial Standards, KS M ISO 11359-2, 2017; Korean Industrial Standards, KS M ISO 527, 2017; Korean Industrial Standards, KS M ISO 604, 2018; Korean Industrial Standards, KS M ISO 1183-1, 2019) and are summarized in Supplementary Table S1; Supplementary Figure S2.

Accelerated weathering tests were performed to investigate the weathering effect of the WPF-RPs and the deterioration of physical characteristics of WPF-RP TPBs during long-term application in an external environment. Three specimens with dimensions of 7 (W) × 15 (L) × 1 (D) cm were prepared by cutting WPF-RPs and commercial high-density polyethylene (HDPE) samples purchased from Polypenco Korea to compare the weathering performance between new and recycled materials. Xenon lamps (Ci4000, ATLAS) were used with an S-borosilicate (S-boro) filter, black panel temperature of 63 ± 3°C, relative humidity of 50 ± 5%, and irradiance condition of 0.51 W/m² at 340 nm for the accelerated weathering test following the instructions given in Korean Industrial Standards, KS F 2274: 2018 (2018). The simulation cycle was irradiated for 102 min, with exposure to water for 18 min in a weathering chamber. Accelerated weathering tests were conducted for up to 1,500 h at an increment of 500 h. The conversion of irradiation time in the weathering chamber to accelerated weathering time was calculated using the 10-year average value of horizontal solar insolation in Seoul by assuming approximately 5% of ultraviolet irradiation and excluding humidity and temperature; that is, the weathering time in the accelerated weathering test chamber was calculated to be ~1,500 h (540 d). These calculations were conducted according to the methods given by Gewert et al. (2018), and further details are provided in Supplementary Table S2. Later, surface cracking and color variation (CE7000A, x-rite) were tested following Korean Industrial Standards, KS A 0063: 2015 (2015). Tensile strength, and elongation at break were measured following Korean Industrial Standards, KS M ISO 527-2 (2018) (RTF series Tensilon, A&D). Color difference was measured and calculated using the following equation from a previous study (Mokrzycki and Tatol, 2011):

Here, L^*, a^*, and b^* indicate the lightness (black/white), red and green between +127 and −128, and blue and yellow between +127 and −128, respectively, and the subscripts 1 and 2 represent the values of the starting point at time zero and at the irradiation times (500, 1,000, and 1,500 h) during the accelerated weathering test.

Dynamic stability of construction materials from the WPF-RP samples was assessed by the wheel-tracking test (AI-1100-3, Iwata, Japan) following Korean Industrial Standards, KS F 2374: 2017 (2017). The wheel-tracking test is a representative method for evaluating the resistance of a specimen to plastic deformation, mostly in summer, by mimicking indoor sites and passing actual vehicle loads under high-temperature of pavement conditions. Specimens with a size of 300 × 300 × 50 mm were prepared and tested with 686 N (or 628 kPa, 70 kg) of wheel load at a rate of 42 cycles/min (i.e., 2,520 cycles/ 60 min) at 60°C. Deformations at 15, 30, 45, and 60 min for the six test specimens were measured, and dynamic stabilities were calculated using the following equation:

Here, 42 represents the number of repeated cycles per 1 min, d₁ and d₂ are the displacements of deformations (mm) at t₁ = 45 min and t₂ = 60 min, respectively, and C is the correction factor of the test machine type, where the value of C is 1.0 because the crank connecting link drives the test specimens.

To apply the TPB prototype using WPF-RPs, field tests were performed at a site in Seoul, which experienced frequent traffic from heavy vehicles, thereby simulating a real road site. The size of the field was 1.2 m (W) × 6 m (L) and could accommodate 80 WPF-RP prototype units, as shown in Figure 1. Further, the total 6 m length of the test site was divided into 4.5 m and 1.5 m sections with and without installing fastening structures, which were designed to improve the load transferability and equalization between blocks (Supplementary Figures S1, S3) to evaluate the scalability of the TPBs in road pavement construction sites. Before and after one month of the operation of the WPF-RP TPBs, profiles, elevation, and slip (or skid) resistance were measured at the installed test site. The profiles of the four rows (1st−4th) were measured three times each along the longitudinal direction of the installed TPBs (Figure 1) using a Surf-Pro walking profiler (CS8850, Surface Systems & Instruments, Inc. USA). The profiles were not used to measure absolute displacement, but relative displacement based on the starting point of the TPBs, which was considered to be 0 mm. The elevations at certain locations in the middle of each TPB (Figure 1) were measured using a bar ruler and a 30 cm-stainless steel ruler. To mimic the sliding of a vehicle, a standard slip resistance test was conducted with a British pendulum number (BPN) tester (HM-602W, Gilson, USA) following the method of Korean Industrial Standards, KS F 2375 (2016), similar to ASTM E 303.

The contents and leachate characteristics of hazardous substances, including heavy metals, from the WPFs, prototype samples were analyzed following the IEC 62321 and Korean Official Wastes Test Method, respectively (Ministry of Environment (MOE), 2017). Among the contents of 10 hazardous substances regulated by the Act on Resource Circulation of Electrical and Electronic Equipment and Vehicles (Ministry of Environment (MOE), 2019), which was recently updated following the Restriction of Hazardous Substances II directive (Directive 2011/65/EU, 2011), Pb and Cd were measured using inductively coupled plasma-optical emission spectrometry (Optima 8300DV, PerkinElmer, UK), Hg was measured using a mercury analyzer (FIMS 400, PerkinElmer, UK), Cr⁶⁺ was measured using ultraviolet/ visible spectrophotometer (V530, Jasco, JP), polybrominated biphenyls and polybrominated diphenyl ethers were measured using gas chromatography-mass spectrometry (7975, Agilent, US), and bis (2-ethylhexyl) phthalate, benzyl butyl phthalate, dibutyl phthalate, and diisobutyl phthalate were measured using gas chromatography–mass spectrometry (5977B, Agilent, US). Samples for the leachate tests of 11 hazardous substances listed in the Korean Official Wastes Test Method were pretreated with an acid solution. Cr⁶⁺, Cu, Cd, Pb, and As were measured using inductively coupled plasma-optical emission spectrometry (Optima 5300DV, PerkinElmer, UK), Hg was measured by atomic absorption spectrometry (FIMS 100, PerkinElmer, UK), cyanide was measured using an ultraviolet/ visible spectrophotometer (V530, Jasco, JP), organophosphorus compounds were measured by gas chromatography-nitrogen phosphorous detector (6890N, Agilent, US), tetrachloroethylene and trichloroethylene were measured by gas chromatography with a microelectron capture detector (7890A, Agilent, US), and oil was measured by gravimetry after extraction with hexane.

The mechanical characteristics of the WPF-RPs are shown in Table 1. The standard deviations of all the measured parameters were small, indicating the homogeneity of the materials. The compressive strength of the WPF-RP samples was approximately 38 MPa, which is close to that of cement concrete (Han et al., 2003; Lam et al., 2018; Aghaeipour and Madhkhan, 2020), indicating that rigidity was comparatively excellent. As the compressive strength was ~2.3 times greater than the tensile strength, using the WPF-RPs as a compression member rather than a tensile member is advantageous in terms of the construction material performance. However, in the compression test, the elasticity was approximately 32–38 times smaller than that of concrete, which can cause significant deformation during load application, thus making it inappropriate for use as a supporting structure during major loads. In addition, the coefficient of thermal expansion was ~12 times more than that of concrete; therefore, replacing existing construction materials would be difficult in poor environments, such as those with a large annual temperature difference, depending on the weather conditions, exposure to the external environment, and load conditions.

The changes in appearance after the accelerated weathering test for the WPF-RPs and commercial HDPE samples are shown in Table 2. No significant change was observed in the appearance of the two plastic specimens until 1,500 h, which corresponded to 540 d; additionally, no cracks were observed. The original color of the HDPE was opaque white, and no visible color change was observed. However, for the WPF-RP specimens, the light-exposed areas became brighter with increasing exposure times. The color difference was measured quantitatively before and after the accelerated weathering test, with marginally increased values after 500 h in both samples. Although a standard for changes in elevation (ΔE) does not exist, a color difference was observed when its value exceeded 1, and the color difference increased with the increase in its value. The values for HDPE samples were close to 1, whereas those for WPF-RP samples were approximately five times higher (5.3).

After the accelerated weathering tests, we assessed the tensile properties of strength and elongation at breaks (Figure 2). A significant decrease in the values for the HDPE samples was observed over time compared with the samples that were not weathered. Conversely, the values for the WPF-RP samples were constant throughout the testing time up to 1,500 h. Although the tensile characteristics of the WPF-RPs are lower than those of pure plastic before weathering and photo-aging, the tensile strength and elongation of the WPF-RPs could be compared with those of HDPE samples after 500–1,000 h, possibly because of the complex structure of the composite polymers in the WPF-RPs (Shah et al., 2008; Iñiguez et al., 2018).

In addition to the mechanical properties, WPF-RPs should ideally show good stability under vehicle loads in road pavement. Thus, wheel-tracking tests were performed to evaluate the stability of vehicle loads on road pavements. Figure 3 shows that the rut depth values for six samples varied due to material heterogeneity, and the dynamic stabilities ranged from 21,000 to 63,000 cycles/mm with an average of 33,250 ± 15,456 cycles/mm. This indicates that the dynamic stability of the WPF-RPs was much higher than that of the asphalt quality standards generally used for road pavement (>3,000 for flow resistance, 2,000–3,000 for stone mastic asphalt, >2,500 for drainage, and >750 for emergency repair at room temperature) (Ministry of Land, Infrastructure and Transport (MOLIT), 2017). These high stability values acquired through the wheel-tracking tests for the WPF-RPs, compared with those of general asphalt pavements, and verified the dynamic stability of the vehicle load while using WPF-RPs as road pavement materials. Therefore, this study provided a better alternative measure to effectively apply WPF-RPs in construction.

The application of wastes, including WPFs, as recycling resources, require environmental assessment, because the wastes may contain various hazardous substances. Synthetic resin products, such as plastic films, comprise petroleum-based synthetic resins and additives, such as plasticizers, heat stabilizers, and antioxidants that are added in the manufacturing process (Rahimi and García, 2017; Dahlbo et al., 2018). Thus, while using WPF-RPs, these substances may be released to the environment, subsequently affecting environmental quality and human health. Environmental safety criteria differ depending on the usage of products from waste. The 10 restricted materials in the WPF-RPs given by the Restriction of Hazardous Substances II were analyzed. Among these, seven were not detected, and three were within the permissible limits, that is, 19.3 mg/kg of Pb (<1,000 mg/kg), 1.2 mg/kg of Cd (<100 mg/kg), and 732 mg/kg of bis (2-ethylhexyl) phthalate (<1,000 mg/kg). Additionally, only Cu (0.013 mg/L) was detected among the 11 hazardous substances by testing the leachates of WPF-RPs. The concentration of Cu was two times lower than the permissible limit (<3 mg/L), which is used to classify “Designated Waste” as defined in the Korean Official Wastes Test Method (Ministry of Environment (MOE), 2017).

Figure 4 shows the process of using WPF-RP TPBs at the test site, which started with excavation for the designated size of the field test, with dimensions of 1.2 m (W) × 6 m (L) for 80 WPF-RP prototype units, to simulate real road construction.

We performed field tests of TPBs at the site, with frequent heavy vehicular traffic for one month, and inspected for any appearance of damage, such as breakage or cracking. Figure 5 shows the longitudinal profile measurement results of the WPF-RP TPB installation section. The profile change after one month of the test showed no significant difference in the section installed with the fastening structures; however, each column was displaced individually, and a change in the slope was observed in the four columns of the section lacking the fastening structures. However, the profile changes, regardless of the installation of the fastening structures, showed increments in slope of 30 cm, which is the TPB unit size. Therefore, the TPB flatness after compaction and before TPB installation would have probably had a dominant effect on the surface profile rather than the presence or absence of the fastening structures. Further, elevations were measured at the test site (Figure 1) to assess whether the fastening structures that tightly held the individual TPB blocks positively affected the minimization of settlement or subsidence. Figure 6 shows the difference in the elevation before and after one month of operation. The Δ elevation values for individual locations showed mostly settlement and some heaving, probably due to unexpected traffic movement. The extent of the difference was greater without the fastening structure, thus indicating the advantage of the fastening structure. In addition, the values of slip resistance measured as BPN were 34 (± 5) and 33 (± 2) for the WPF-RP TPB samples before and after one month of operation, respectively. These values satisfy the minimum slip resistance BPN values for road pavement materials under average conditions (S3) and for sections, where friction is not important (S4) at risk level 1. This level is defined as when there are few accidents due to road slipping or when accidents are not yet recorded (Ministry of Land, Infrastructure and Transport (MOLIT), 2016). The detailed definitions of minimum friction coefficient and risk are provided in Supplementary Tables S3, S4. This indicates that TPBs could be potentially applied to roads with low-speed driving, such as side roads, and those with minimal friction needs, such as on flat ground.