Size effect of test specimens on tensile splitting strength of concrete: general relation

A broad experimental program was carried out on cubes, cylinders and prisms (loaded transversally or longitudinally) of sizes usually used in laboratory testing. For casting, concrete mixes of various compositions were used. The 1600 results obtained were evaluated by mathematical statistical methods. The volume of test specimens was substituted either by the size of the corresponding fracture area or by the size of the highly stressed volume (HSV) in their loaded cross-section. The absolute values of the tensile splitting strength were converted to relative values (depending on the basic size of the fracture area or on the highly stressed volume). For mathematical expression of the relationships between these quantities, an exponential function was used. This relation enables to convert the test results obtained on specimens of different sizes and shapes to the strength of the basic size specimen. This is important for the test results obtained on samples taken from finished structures, as these samples are usually of different sizes.