Utilizing Construction Demolition Waste Aggregates in Cement- Free Concrete for 3D Printing (2025-08)¶

The rising CO₂ emissions released into the atmosphere is a well-known environmental issue, with the construction sector being a major contributor. According to the Circular Economy Action Plan, construction sector accounts for over 35% of the EU's total waste and greenhouse gas (GHG) emissions, largely due to the production of materials such as cement and ceramics. Recognizing this, the European Green Deal emphasizes the need for energy-intensive industries like construction to reduce emissions and transition toward more sustainable practices. 3D concrete printing (3DCP) is an emerging technology that can minimize material use and waste trough optimized geometry. Currently, many 3DCP projects primarily serve as stay-in-place formworks, with load- bearing capacity still provided by traditional reinforced concrete. Most existing 3DPC formulations contains large amounts of cement – a major source of CO₂ emissions. From an environmental perspective, it is crucial to develop mixtures for 3DPC with reduced environmental impact, particularly in applications where lower mechanical properties are sufficient. This can be achieved by reducing cement content or incorporating alternative aggregates. To address this issue, this study developed multiple low-CO₂ mixtures suitable for 3DCP by replacing cement with industrial waste ash as a binder and incorporating construction demolition waste aggregate (CDWA). High-calcium oil shale ash (OSA), which exhibits self-cementing properties, was used as a cement replacement. To enhance its binding properties, OSA was combined with a pozzolanic waste material – metakaolin (MK). Two types of CDWAs were sourced from different construction and demolition waste landfills and compared to commercially available natural coarse aggregate (NCA) of dolomite to assess performance differences. NCA had an apparent density of ~2700 kg/m³ and a relatively low water absorption of ~2.0%. Both RWAs had slightly lower densities of ~2600–2650 kg/m³ and significantly higher water absorption of ~9.0%. Additionally, the RWAs contained a notable amount of fine dust particles, as well as traces of organic compounds and bitumen. All three aggregate types were used in the 2/8 mm fraction. Granulometric analysis showed nearly identical particle size distributions (PSD) between the RWAs, with only slight variations compared to the NCA. This ensured comparable packing densities across all mixtures while maintaining consistent sand and binder proportions, leading to similar workability and water demand. First, three series of cement-containing mixtures were formulated using NCA and both types of RWAs, with a binder composition of 60 wt% OPC, 25 wt% OSA, and 15 wt% MK. Then, for each cement- containing mixture, a corresponding cement-free mixture was formulated, where the binder consisted of 80 wt% OSA and 20 wt% MK. In total, six different series were tested for both cast and 3D-printed specimens. 3DCP was carried out using a gantry-type printer, with each mixture printed into a square- shaped element. After hardening, the printed objects were sawn into prismatic specimens of 40×40×160 mm for flexural and compressive strength tests, which were performed in print direction. Additionally, cast specimens of the same dimensions were prepared from the same batch for comparison. Cement-containing mixtures showed hardened state densities of ~2150 – 2250 kg/m³, while cement-free mixtures had slightly lower densities of ~2050 – 2150 kg/m³. The density results showed negligible differences between cast and 3D-printed specimens. For all series containing cement, the samples achieved flexural strengths ranging ~5.5 – 7.5 MPa, while cement-free mixtures exhibited strengths of ~2.5 – 4.5 MPa. Surprisingly, the 3D-printed samples consistently outperformed their cast versions by at least 20% across all mixtures. Visual observation of the samples revealed that the aggregate distribution in cast specimens was more random, while the extrusion process of 3DCP led to a more structured arrangement of relatively large aggregate particles. This could have contributed to a stronger interlocking effect, improving stress transfer under bending loads. Similar findings have been reported in studies on conventionally cast concrete, which suggest that larger particles enhance load transfer and crack bridging. The compressive strength results showed a significant difference between cement-based and cement-free samples. As expected, the presence of cement led to higher compressive strengths of ~50–65 MPa. Despite the absence of cement, the cement-free mixtures still achieved strengths of ~30 MPa, which falls within the range of normal-strength concrete. This suggests that the combination of OSA and MK can provide sufficient structural integrity even without cement. Interestingly, in all cases, whether with or without cement, the differences in compressive and flexural strength results between formulations using NCA and CDWA were minimal. Although 3DCP samples exhibited ~15% lower compressive strength compared to cast specimens, the cement-free 3DCP mixtures with CDWA still demonstrated unexpectedly high compressive strength, making them suitable for various 3DCP applications, such as stay-in-place formwork, non-load-bearing structures, and even low-rise buildings. Lyfe cycle analysis results showed that the majority of the emission reduction comes from cement replacement, by reducing ~80-90%. The use of CDWA contributed to a smaller reduction when compared to mixtures with NCA.