Concrete structures can be subjected to elevated temperatures in accidental situations. Fire is the most known situation but the nuclear field may also involve some specific situations inducing elevated temperatures (loss of coolant accident, core meltdown severe accident). The concrete designer has to limit the risk of thermal spalling of concretes when potentially exposed at elevated temperature. This input data is usually a portion of many others that not all go in the same way: the high durability requirements can induce high strength, low permeability and porosity, and therefore a high thermal spalling risk. The behaviour of concrete under high temperature is generally evaluated from representative tests (representative scale of the concrete structure, representative thermal conditions). Given that this type of test is complicated and cumbersome, it usually comes at the far end of the concrete design process.

The aim of this paper is to present the study performed in the concrete laboratory of EDF Ceidre TEGG department in order to define a simple test that guides the choice of concrete composition during the design stage. The tests conditions were chosen to promote the development of thermal instability phenomena: concrete samples initially in a water saturated state, heating rate about 10°C per minute and heating up to 550°C. Concrete properties were followed before and after the test on the same samples by non-destructive measurements of ultrasonic pulse velocity and resonance frequency (allowing the determination of the dynamic modulus). Compressive strengths have been also measured after the tests and compared to the results obtained on intact samples.

This paper exhibits particularly the tests that were performed on several types of high performance concretes: high performance mortar, high performance concretes with different proportions of slag by cement replacement, high performance concrete containing some polypropylene fibers,...