Experimental Comparison of Compressive Strength, Flexural Strength, and Splitting Tensile Strength of a Single Normal Concrete Mixture with Empirical Predictions from SNI 2847:2019

Abstract

Concrete has three principal mechanical properties: compressive, tensile, and flexural strength. In reinforced-concrete design, compressive strength (f’c) is the primary indicator of concrete quality, while tensile and flexural strengths indicate its ability to resist tensile and bending loads. SNI 2847:2019 provides empirical expressions relating splitting tensile and flexural strengths to compressive strength. This study compares the measured mechanical properties of one normal-concrete mixture from a single batch with values predicted using the applicable SNI expressions. It does not establish a general relationship among these properties because that would require multiple strength levels and independent batches. The concrete mixture was designed according to SNI 03-2834-2000 for a target 28-day compressive strength of 20 MPa. Eighteen specimens were produced: six cylinders measuring 150 × 300 mm for compressive testing, six cylinders of identical dimensions for splitting tensile testing, and six beams measuring 150 × 150 × 500 mm for flexural testing. After recalculating the results from recorded loads and cross-sectional areas, mean compressive strength was 20.65 MPa (SD 0.58 MPa; CV 2.79%), flexural strength was 1.91 MPa (SD 0.05 MPa; CV 2.59%), and splitting tensile strength was 1.77 MPa (SD 0.15 MPa; CV 8.29%). The measured flexural and splitting tensile strengths were approximately 39.8% and 37.3% below the SNI-predicted values of 3.18 MPa and 2.82 MPa, respectively. Inadequate compaction is only one unverified explanation because supporting air-content, density, porosity, ultrasonic, or microscopic evidence was unavailable. Conversion errors, loading rates, fracture locations, and specimen alignment are also plausible factors. Therefore, SNI 2847:2019 overestimated both strengths for the investigated mixture under controlled laboratory conditions. However, this finding cannot be generalized because it represents only one mixture, batch, and testing age. Confirmation requires independent batches, multiple strength levels, repeated testing, and regression analysis.