Experimental studies and drying shrinkage prediction model for concrete containing waste foundry sand

Concrete being most extensively used construction material all over the globe has resulted in the over-exploitation of natural resources such as river-sand and gravels. Meanwhile, advancing industries and increasing population have also lead to an increased generation of waste materials. Many of these waste materials have the potential to be used in concrete. This study has investigated the effect of one such waste material referred to as Waste Foundry Sand (WFS) on the properties of concrete. Many researchers have studied the effect of WFS on the mechanical properties of concrete. However, no consensus has been reached yet and very contradictory results have been reported. Moreover, shrinkage of concrete containing WFS has not received much of the researchers' attention and very limited literature is available related to this property. This study explores the workability, compressive strength, split tensile strength and drying shrinkage of such concretes. WFS was used as partial replacement of fine aggregate and the replacement levels were varied from 0 to 50% in equal increments of 10%. It was observed that the compressive strength and split tensile strength of concrete decreased with the addition of WFS. However, the mix WFS3 (30% WFS) showed the strength very similar to that of control concrete. A significant increase of 16.7%, 23.44%, 29.05%, 36.35% and 45.18% with respect to control concrete at the age of 28 days was observed in the magnitude of drying shrinkage when the amount of WFS in concrete was varied from 10 to 50%. Furthermore, it was seen that the available shrinkage prediction models could not be applied to the concrete containing waste foundry sand. Hence, a multivariable regression model inspired by ACI-209 model was developed for prediction of drying shrinkage in concrete. The proposed model included the parameters related to drying age and percentage of WFS in concrete and gave a high value of the coefficient of regression (R-2) indicating the effectiveness of the proposed model.