Study on the engineering properties and prediction models of an alkali-activated mortar material containing recycled waste glass

Abstract

In recent years, the global temperature has risen as greenhouse gases increase. With Taiwan’s limited natural resources and the rise of environmental considerations, energy saving, carbon reduction and waste recycling are universal subjects at the present time. This study combined waste LCD (liquid crystal display) glass with alkali-activated slag mortar as a thermal insulating material and evaluated its engineering properties. Different liquid-solid ratios (L/S=0.45, 0.50, 0.55 and 0.60) were used. The alkali equivalent was fixed at 1%, and the slag was replaced by glass sand (0%, 20% and 40%). Various engineering properties were tested at the ages of 3, 7 and 28days. The results show that with more glass sand replacement and higher liquid-solid ratio values, the workability is better, and the slump and slump flow are higher. The compressive strength decreases as the liquid-solid ratio increases, and the compressive strength increases when the glass sand replacement is 0–20%. However, in contrast, when the replacement is 20–40%, the compressive strength decreases. The ultrasonic pulse velocity decreases as the glass sand replacement increases. The thermal conductivity increases with age, but the thermal conductivity decreases as the liquid-solid ratio initially increases, and the thermal conductivity is at its minimum when L/S=0.55. When the L/S increases continuously, the thermal conductivity increases accordingly. The resistance to sulfate attack increases with the replacement and the liquid-solid ratio. The effect is the best when L/S=0.60 and the glass sand replacement is 20%, meaning that a certain amount of waste LCD glass in the alkali-activated slag mortar can enhance the engineering properties. In addition, the prediction models for the compressive strength, ultrasonic pulse velocity and thermal conductivity of alkali-activated slag mortar material with waste LCD glass were deduced in this study. According to the comparison between the prediction analysis values and the test results, the compressive strength mean absolute percent error (MAPE) values are 4.37–5.52%, the ultrasonic pulse velocity (UPV) MAPE values are 0.01–1.53%, and the thermal conductivity (kt) MAPE values are 0.82–2.68%. All the MAPE values are smaller than 10%, so the analytical models built have good forecasting accuracy.