ELSEVIER Construction and Building Materials, Vol. 10, No. 8, pp. 571-575, 1996 © 1997 Elsevier Science Ltd Printed in Great Britain. All rights reserved 0950~618/96 $15.00+0.00 PII:S0950-0618(96)00025-6 Failure mechanism of concrete, combined effects of coarse aggregates and specimen geometry A. Di Maio, G. Giaccio* and R. Zerbino LEMIT-ClC, 52 el 121 y 122, 1900 La Plata, Argentina Received 8 December 1994; revised 12 August 1996; accepted 28 September 1996 Concrete compressive strength is mostly evaluated by tests performed on cylindrical specimens with slenderness ratio two. Nevertheless, specimens with lower slenderness ratio are also used (drilled cores, cubes, etc.). Tests performed on cubes are affected by the multiaxial stress field induced by the reduced slenderness ratio. Then, the crack propagation is modified depending on the characteristics of the composite material. This paper analyzes some phenomenological aspects of the evaluation of concrete compressive strength related to the effect of coarse aggregates on the initiation and propagation of cracks. The influences of strength level and microcracking are also discussed. © 1997 Elsevier Science Ltd. All rights reserved. Keywords: compressivestrength; failure; cracks Introduction Concrete is a composite material of brittle nature. It consists of strong particles dispersed in a weaker matrix. Under loads substantially lower than ultimate strength, cracks grow and develop from the pores, microcracks, and especially at the matrix-aggregate interfaces (the weakest links of the composite). Many experiences have verified the strong relationship between the concrete stress-strain behavior and crack formation. They particularly justify the non-linear response of concrete near failure. While a unique crack develops through an ideal brittle material, a multiple crack pattern appears in the concrete before failure l"e. In addition, the propagating cracks are arrested by aggregate particles resulting in the meandering and branching of the cracks. Concrete compressive strength is mostly evaluated by tests performed on cylindrical specimens with slenderness ratio (h/d) two. Nevertheless, when it is necessary to evaluate in situ concrete strength (drilled cores), specimens with slenderness ratios lower than two may be used. Added to this fact, other types of specimens (prisms or especially, cubes) are adopted in many countries because of the greater amount of experimental work based on them. It is well known that tests performed on cubes are affected not only by the shape of its section, but also, and specially so, by the multiaxial stress field induced by the reduced slen- derness ratio. As a consequence, crack propagation is modified depending on the characteristics of the composite material. This paper contributes to the discussion on the failure behavior of concrete under static compressive loads, *Correspondence to G. Giaccio regarding the influence of the characteristics of coarse aggregate when different types of specimens are used. Experiences To discuss some phenomenological aspects related to the combined effects of coarse aggregate and specimen geometry on the measurement of concrete strength, three series of tests were performed. They illustrate the influence of the aggregates' size, previous microcracking and concrete strength level on the measured ultimate compres- sive stress. Table 1 shows concrete mix proportions used in Series 1, 2, and 3. Mortar and concretes were prepared in a tilting mixing machine and were properly compacted by vibrating the mix. In the case of high strength concretes (Series 3), a naphthalene based superplasticizer was used. The effect of aggregate size One mortar and six concretes (Series 1), prepared with aggregate maximum sizes 25, 50, and 75 mm, were designed to evaluate the effect of coarse aggregate on the concrete behavior when tests on different types of specimens are performed. Cubes (D), prisms with slen- derness ratio two (C), cylinders with slenderness ratio one (B), and cylinders with slenderness ratio two (A) were cast. Two types of coarse aggregates were selected: a granitic crushed stone (CS) and a fiver gravel composed of rocks of diverse petrography and rough surface texture (G1). Ordinary portland cement (similar to ASTM Type I), and siliceous natural sand were used. A water to cement ratio that equalled 0.60 was adopted. Specimens were stored for 571