The finding is expected to assist construction industry professionals, government policymakers, and housing developers to overcome issues around depletion of natural resource, reduction in carbon dioxide (CO₂) emission and environmental concern regarding concrete production by embracing the use PWKS and PMKS as substitute for natural coarse aggregate. This is because an increased demand for non-renewable raw aggregates suggests continuous depletion of natural resources, increase energy demand, and increased greenhouse gas emissions, causing a considerable impact on the environment and society. Nonetheless, this research examined the potential of deploying discarded PWKS and PMKS from agricultural waste as alternatives to natural aggregate in making interlocking concrete pavers. This helps to possibly conserve some significant amount of raw aggregate to achieve responsible consumption, in addition to creating sustainable infrastructure, as well as achieving eco-friendly structures from waste materials which aligns with sustainable cities and communities.

The materials utilized in this research for the production of the interlocking concrete paver blocks were;

i. Periwinkle shells (PWKS);

The sand and granite materials were obtained from quarry sites in Abeokuta, Ogun State. The maximum nominal size for the granite used as coarse aggregate in the mixes ranges between 10–13.2 mm. The PWKS and PMKS shown in Figures 1A, B were obtained as waste from agricultural waste collection point located in Badagry area, Lagos state. The shells were first properly washed with warm water to remove dirt and any surface impurities, then the shells were sufficiently openly air-dried in the sun before their utilization in the mixes. The particle size distribution of the sand, granite and shell materials was evaluated in line with BS 812 (1990). Physical parameters such as bulk density, water absorption, specific gravity, moisture content, aggregate impact value (AIV) and aggregate crushing value (ACV) of the sand, granite as well as that of the PWKS and PMKS materials were evaluated based on relevant standards. These tests are important to show the suitability of these PWKS and PMKS as alternative aggregates to conventional aggregates in concrete paver block production. The Portland cement used was of grade 42.5N, bought from a vendor outlet store located in Ota town, Ogun State. The cement conforms to type 1 Portland cement for general-purpose construction (BS EN 197-1, 2000). Portable water was used in this research for mixing. The oxide compounds in the cement, PMKS, and PWKS materials were determined using the x-ray fluorescence (XRF).

The batching proportioning of the various components for making the block samples was performed by volume in three (Bilodeau and Malhotra, 2000) batches adopting a mixing proportion of 1:1:2 (cement: sand: granite) at constant water-cement (w/c) of 0.50. A minimum targeted strength of 30 MPa was expected to be achieved by the produced paver samples after 28-day period of curing. The first batch was the control mix sample, and the second batch involved partially replacing the granite component separately with PWKS and PMKS at percentage replacement of 5% and 10% in the concrete mixes. The third batch involved combining both PMKS and PWKS materials in the same mixes and used to partially replace the granite component in the concrete mixes using the same percentage replacement levels of 5% and 10% at a constant w/c ratio of 0.5. Table 1 shows the adopted mix proportioning of the constituents in the production of the interlocking concrete paver block samples. Each constituent material was weighed, and the materials were carefully and evenly mixed to achieve uniform distribution. The mixing procedure was done manually, and a slump test was used to determine the mix’s workability. The recorded slump values were observed to be in the S3 class (100–150 mm) for a 0.50 w/c ratio, thus implying that the achieved mixes are sufficiently moist and fluid for making concrete paver blocks that can be deployed for structural uses.

An H-type shape mould was used in this research to make the interlocking paver block units, as shown in Figure 2. Each of the interlocking paver block units produced was 215 × 135 × 80 mm in dimensions. The 80 mm thickness follows the guidance of ASTM C936 (2002) on the minimum thickness for paver block units for possible deployment on street roads or light traffic. A total of 108 paver block samples were produced, and casting was carried out in three layers in a well-oiled plastic mould. Compaction of the concrete during casting was done manually with a tapping rod to compact each layer twenty-five times and a vibrating table to remove pore spaces within the concrete. Afterwards, these paver samples were demoulded after 24 h and curing of the samples was done using two different curing approaches that included; total immersion in water using a water curing tank and open-air curing at temperature ranges of 30°C–31°C. The hardened paver samples were cured for 7 and 28 days before testing.

Table 2 depicts the chemical compositions present in the PMKS and PWKS materials. The result showed no presence of toxic compounds in both materials, which is similar to the outcomes of Ogundipe et al. (2021). Both PMKS and PWKS materials contain a sufficient amount of silica (SiO₂), while there is a relatively high amount of calcium oxide (CaO) content in the PWKS. The findings show that the shells material do not contain toxic compounds similar to the outcomes by Ogundipe et al. (2020); Adewuyi and Adegoke (2008). Table 3 show the results on the tested physical properties of the constituent materials. The results show that the specific gravity values recorded for both river sand and granite are within the defined standard of 2.5–3.0 required for aggregate (BS EN ISO 10545, 1997). However, the recorded values for the specific gravity values for PMKS and PWKS were less, indicating that both materials could be regarded as lightweight aggregate. The water absorption tendency simply shows that the PMKS material has a higher tendency to take in more water. Further, the recorded Los Angeles abrasion values and aggregate impact values showed that PWKS failed to meet wearing surface criteria compared to PMKS which shows some degree of resistance to impact and abrasion. The Los Angeles abrasion and aggregate impact values should not exceed 30% for aggregates used for wearing surfaces. Figure 3 shows the particle size distribution curve for natural sand, granite, PMKS and PWKS aggregate materials.

Figures 4A, B depict the influence of incorporating the PMKS and PWKS materials as partial substitutes for coarse aggregate on the compressive strength of the fabricated interlocking concrete pavers. A steady decreasing trend was observed in the compressive strength values of the paver blocks with the increase in the replacement levels of granite with PMKS and PWKS, and a combination of both PMKS and PWKS in the mixes compared to the control. For instance, the recorded outcomes for the 28-day average compressive strength for the control paver samples are 39.85 MPa and 27.70 MPa for immersion and open-air curing, respectively. For paver samples containing 5%PWKS as granite replacement, the recorded strength was 35.72 MPa and 25.07 MPa for curing by immersion and open-air, respectively, implying a 10% strength reduction. Further increase in the PWKS content to 10% (10%PWKS) as granite replacement also showed a 15% reduction in strength with compressive strength values of 34.02 MPa and 23.82 MPa for curing by immersion and open-air, respectively. Similarly, for paver samples containing 5%PMKS content, the recorded compressive strength values were 34.61 MPa and 24.23 MPa, which indicate a 13% strength reduction for curing by immersion and open-air respectively. However, about 26% strength reduction was achieved for paver block samples containing 10%PMKS. The obtained compressive strength values were 29.54 MPa and 20.68 MPa for immersion and open-air curing, respectively. For concrete paver samples produced from mixes containing the combination of both shell materials at varying percentages, a combination of 5%PMKS & 5%PWKS (31.43 MPa and 22.43 MPa), 5%PMKS & 10%PWKS (32.95 MPa and 23.68 MPa), 10%PMKS & 5%PWKS (30.12 MPa and 21.70 MPa), and 10%PMKS & 10%PWKS (28.32 MPa and 19.66 MPa) as coarse aggregate replacement, the achieved 28-day compressive strengths are for immersion and open-air curing are as indicated respectively. The percentage reduction in strength was 20%, 17%, 24% and 29% respectively. The decrease in the compressive strength becomes more pronounced as the replacement levels increase beyond 10%. This can be attributed to the physical properties of the PMKS and PWKS materials compared to granite materials. The shell material for instance, has a lesser weight compared to the granite, and its shape often resulted in a possible increase of pore spaces within the concrete matrix (Osei, 2013). Generally, there was a steady increase in the compressive strength of all samples to the curing age. However, the results depict the impact of the two curing methods (immersion and open-air) adopted on the strength development of the tested hardened samples. The open-air curing is simply openly exposing the paver samples without covering but with occasional wetting with water as the usual practice in most paver blocks production sites. The plots showed that hardened paver samples cured by immersion achieved the highest compressive strengths for the control and pavers containing PMKS and PWKS compared to the paver blocks cured in the open, indicating that water immersion is a more effective means of curing paver block samples. The outcomes also indicated that mixes containing 5%PWKS and 5%PMKS performed relatively better compared to others. These findings corroborate the reported findings of (Prusty and Patro, 2015; Azunna, 2019; Ogundipe et al., 2021; Adewuyi and Adegoke, 2008; Odeyemi et al., 2020) on how effective PWKS and PMKS materials are as possible alternative aggregate. As shown in Figure 4A, the pavers cured by water immersion were able to meet the minimum compressive strength of 30 MPa as recommended by BS 6717 (2001); IS 15658 (2006) for interlocking concrete pavers used for light load carriageway and non-traffic applications, car parking spaces and building premises, while ESS 4382 (2004) recommended for normal and medium duty loads. However, the obtained compressive strength fails to meet the 40–55 MPa for heavy traffic and duty applications (BS 6717, 2001; IS 15658, 2006). Samples cured in open air failed to meet any of these minimum strength requirements, but can be deployed for walkways (Figure 4B).

Figures 5A, B show the results of the tensile strength conducted on the paver blocks samples. The result indicates that utilization of PWKS, PMKS, and a combination of both shell materials as granite substitutes slightly decreases the tensile strength of the concrete pavers as the shells content increases in the mixes. This trend is like the findings seen for compressive strength. However, the tensile strength values for the mixes containing 5% PMKS and PWKS are relatively close compared to the value recorded for the control. The reduction in the tensile strength could be a result of a possible loss of bonding within the concrete that may have resulted in less resistance to splitting (Yerramala and Ramachandrudu, 2012). Moreover, the granite aggregate is denser than both PMKS and PWKS, which could have possibly caused an increase in void spaces in the mixes resulting in a decrease in the tensile strength performance (Azunna, 2019). The plots also depict the effect of curing methods on the tensile strength development of the paver samples. Again, it was noticed that curing by immersion achieved better tensile strength values after 28 days of curing compared to the open-air curing. The recorded outcomes for the 28-day average split tensile strength for the control samples were 2.68 MPa and 2.37 MPa for immersion and open-air curing respectively. For mixes containing only 5%PWKS as granite replacement, the 28-day tensile strength was 2.61 MPa for water immersion and 2.29 MPa for open-air curing, while at 10%PWKS, the results show the tensile strength values to be 2.53 MPa for water immersion and 2.09 MPa for open-air curing. For mixes containing only 5%PMKS as granite replacement, the 28-day split tensile strength was 2.53 MPa for water immersion and 2.28 MPa for open-air curing, while at 10%PMKS, the results were 2.50 MPa for water immersion and 2 MPa for open-air curing. Similar trends were noticed indicating curing with water immersion achieved better tensile strength performance compared to open-air curing for mixes containing the combination of both shell materials as granite replacement. However, it was noticed that the reduction in the tensile strength for some mixes containing the combination of both shell materials is minimal.

Figure 6 shows the correlation regression linear models between the 28-day compressive strength and tensile strength magnitudes obtained for the concrete paver block samples relative to the curing methods used. The model establishes the best fit by adjusting the relationship between compressive strength and tensile strength variables. The coefficient of determination (R) of these models establishes that the dependent variable, compressive strength was predictable at R ² = 0.73 for samples cured by water immersion, while for samples cured by open air, the compressive strength was predictable at R ² = 0.80.

Water absorption is a relevant attribute that influences the long-term performance of concrete pavers. Low water absorption would imply good resistance to damages caused by water infiltration and environmental effects (Abid et al., 2022). Figure 7 depicts the results of the water absorption of the hardened paver block, showing a proportional increase in water absorption rate with an increase in PMKS and PWKS content. A similar increase was noticed for concrete pavers with combined PMKS and PWKS materials compared to the control. The interlocking concrete paver blocks cured by open air recorded the highest water absorption values compared to those cured by water immersion. The figure showed that pavers with combined shells (10%PMKS/10%PWKS) recorded the highest water absorption values of 6.19% (water immersion curing) and 7.7% (open air curing), compared to 5.26% (water immersion curing) and 6.5% (open air curing) for the control. The findings agree with the recent work of (Olofinnade et al., 2023) on increasing water absorption of composite containing PMKS and PWKS, respectively at increasing content. This was attributed to the high-water absorption capacity of the shells and the possible presence of more numbers of pores in the hardened paver samples, especially those cured in the open air. Similarly, ASTM C936 (2002) and IS 15658 (2006) standards recommended that the water absorption limit for concrete paving units should not be more than 5% and 6% respectively, while (BS 6717, 2001) suggested 7%. However the outcomes showed that all concrete paver units containing PMKS and PWKS cured in open air exceeded the recommended limits, whereas those cured by complete water immersion fell within the 5%–7% recommended for interlocking paving units. The relationship between 28-day compressive strength and water absorption values at 28 days for water immersion and open-air curing is shown in Figure 8. The plot depicts low compression strength values to high water absorption recorded for open-air cured samples. The compressive strength of paver samples cured by water immersion exhibited a more moderate water absorption capacity. The coefficient of determination (R ²) of the models showed that the compressive strength was predictable at R ² = 0.9323 for samples cured by water immersion, while for samples cured by open air, the compressive strength was predictable at R ² = 0.7642. It can be said that samples cured by water immersion performed better than openly cured samples.

Selected concrete fragments from the crushed 28-day samples cured by water immersion were further analysed using the scanning electron microscope (SEM) to view the surface morphology at the micro-scale. The selection was limited to control mix samples and mixes containing 5%PWKS and combined 5% PMKS/10%PWKS (these are mixes that recorded maximum strength performance under applied load). Figures 9A–C show the outcome of the SEM analysis. The SEM micrograph for the control mix (Figure 9A) revealed a dense packing interface with the presence of few micro-pores, thus implying a well-compacted concrete mix with somehow good bonding between the aggregates and cement matrix. This explained the high strength exhibited by the control mix. Additionally, Portlandite (Ca(OH)₂) crystals were noticed to be arbitrarily dispersed through the control sample. Figure 9B depicts the SEM image for the mix sample containing 5%PWKS. The SEM revealed the presence of excess micro-pores, with possible weak interfaces due to the existence of pores within the concrete internal structure. The SEM image suggests that the presence of possibly fragmented PWKS in the mixes may have resulted in poor development of the concrete internal structure pores and increased water intake tendency, causing decreased strength. Moreover, the SEM image for concrete mix with the combined shell material (5%PMKS/10%PWKS), as shown in Figure 9C also displayed a similar presence of excess tiny micro-pores on the concrete interface, coupled with the revealing of weak bond and micro-cracks at the interfacial zone.