The effect of penetration reducing materials on concrete permeability and strength with "cylindrical chamber" and "torsion" tests

Document Type : Research Article


1 Ph.D, Student of Civil Engineering, Imam Khomeini International University, Qazvin, Iran

2 Ph.D Student, Department of Civil Engineering, Imam Khomeini International University, Qazvin, Iran

3 Department of civil engineering, Faculty of engineering, International Imam Khomeini university, Qazvin, Iran


Harmful materials penetrate the concrete and reduce its durability. Therefore, knowing the permeability of concrete is essential. Today, concrete penetration reducing additives are widely used to construct various concrete structures such as water storage tanks. This paper discusses the effect of factors such as the amount of cement, water to cement ratio, penetration reducing materials, concrete age, and the relation between surface strength, compressive strength, and surface water penetration into the concrete. Concrete cube samples are prepared with the strength of 25, 30, 35, and 40 MPa and ages of 7, 28, and 90 days. Permeation reducers such as waterproof, micro silica, and mezocret have been used in the samples. Using the torsion method with a cylindrical chamber device and a concrete breaker jack, the surface strength, permeability, and compressive strength of concrete specimens have been measured, and their relation with each other has been investigated. Also, the volume percentage of permeable pores was calculated according to ASTM C642-06. This standard was used as a criterion for measuring permeability. The results show that the highest permeability reduction is for waterproof, micro silica, mezocret, and without additive concrete samples, respectively, and its amount varies from 5 to 20 ml. Despite the complex structure of penetration reducing materials, it is possible to predict the water penetrating volume into the concrete specimens with appropriate accuracy by obtaining compressive and surface strength of concrete specimens and using the proposed regression equations.