Under the Sustainability Concept, a joint partnership project was agreed between the NSW Road and Traffic Authority (RTA), the NSW Department of Environment, Climate Change and Water (DECCW) and Boral in order to assess the technical and economic performance of recycled glass as a partial natural sand replacement in concrete. The potential environmental objectives of this project were to:
? Enhance the economic viability and sustainability of domestic kerbside recycling.
? Reduce greenhouse gas emissions.
? Keep in excess of 100,000 tonnes of glass out of NSW landfill.
? Conserve energy and material resources.
Both laboratory and field trials were carried out and the results of this work are presented as follows.
PHYSICAL PROPERTIES OF RECYCLED CRUSHED GLASS
The glass as natural sand replacement in concrete trials was a crushed product with a size distribution of between 3mm ~ 0.3mm. The clear and green glass was very clean with no materials passing 150 and 75 micron fractions.
ACCELERATED TEST OF ALKALI SILICA REACTION OF GLASS SAND
There is a concern that glass aggregate in concrete can be problematic due to the alkali silica reaction between the cement paste and the glass aggregate. Therefore, the crushed glass has been tested for Alkali Silica Reaction (ASR) as per VicRoads? RC376.03 method. Different combinations of glass replacement, with and without fly ash, were considered.
The maximum expansion was 0.03 per cent after testing for 21 days. The classification of the ASR test method in VicRoads considers aggregates as non-reactive if the expansion is less than 0.15 per cent for fine aggregate. Therefore, none of the combinations tested were reactive. However, the positive impact of fly ash on expansion is noted, which is well in line with current knowledge on the ASR and previous work by University of Sheffield, United Kingdom in 2004.
LABORATORY CONCRETE TRIALS
Three concrete trial mixes (in triplicate) were carried out against a control concrete mix using only natural products by replacing 30 per cent, 45 per cent and 60 per cent of the natural coarse sand with recycled glass sand. The mixes contain cement, fly ash, coarse aggregates (20/10mm) and coarse sand.
For a constant slump of 60?10mm, concrete with glass sand had marginally lower W/(C+FA) ratios. The air content and bleeding are slightly higher for all mixes using crushed glass as natural sand replacement. In addition, the setting time becomes much longer and it increases with the amount of glass sand added to the mix.
The properties of hardened concrete vary with the different replacement percentages. A better performance of the 45 per cent natural sand replacement mix is noted. The compressive strength results shows that when 45 per cent of natural sand was replaced, the water/cementitious (W/C) ratio dropped slightly, resulting in about eight per cent higher compressive strength at 28 days when compared with the control mix. However, there are no major differences in any of the mixes, when compared with control for flexural/indirect tensile strength and modulus of elasticity. It is also noted that the abrasion resistance is a major issue with all mixes using crushed glass not performing well when compared with control.
Interestingly, the 56 days drying shrinkage results are slightly lower for all mixes using crushed glass.
With regard to the durability performance, three concrete cylinders from each concrete mix were tested for the chloride diffusion as per NT Built 443 at age of 28 days. All concrete cylinders have been immersed in the salt solution for 35 days. The results, in terms of chloride diffusion co-efficient, show a better performance for all mixes using crushed glass as a natural sand replacement.
FIELD CONCRETE TRIALS
The field trials were carried out in early 2010. The supplied concrete was assessed for the slump of five truck loads before it was placed in-situ. Some concrete was sampled and tested later in a laboratory for the compressive strength and drying shrinkage. Concrete core samples were drilled and tested at 28 days.
Having similar slumps of 60?10mm, concrete with 45 per cent natural sand replacement with glass sand has achieved very similar compressive strength, either tested by cylinder or core samples, in comparison with the control concrete. The glass sand concrete had lower drying shrinkage, which is consistent with laboratory trials.
CONCLUSIONS
It has been established from this study that:
? Crushed glass sand can be used to partially replace the natural sand to produce concrete with at least equivalent mechanical properties of a concrete mix using natural aggregate only. The optimum replacement percentage is 45 per cent.
? Durable concrete is expected when glass sand is used to partially replace natural sand. The drying shrinkage, apparent volume of permeable voids, and chloride diffusion co-efficient decrease with the increment of replacement ratios of glass sand. The alkali silica reaction results show that the glass sand used in this project is acceptable as an aggregate. Data from this project indicates that better durability can be achieved when crushed glass sand is used in concrete mixes.
Ion Dumitru is the technical manager and Tony Song the senior development engineer for Boral Construction Materials NSW/ACT