The size effect has its origin in fracture mechanics and describes the formation as well as propagation of cracks in brittle and solid materials in dependence of the specimen size. However, the size effect in concrete spalling describes the damage behaviour on a macroscopic scale for different sized specimens in case of fire. Concrete spalling is a very complex and yet not fully understood phenomenon. To reduce the effort of fire tests to analyse the spalling behaviour of concrete mixtures, this study investigates the susceptibility to spalling for six different concrete mixtures and three specimen sizes. The sizes were divided in full scale slabs (1.8 m × 1.2 m × 0.3 m), intermediate scale cuboids (0.6 m × 0.6 m × 0.3 m) and small scale cylinders (Ø0.15 m × 0.3 m). For this purpose, a novel test set-up was built to test six intermediate scale or twelve small scale specimens simultaneously to ensure a similar heating regime for every specimen. All specimens were fire exposed on one side and remained unrestrained. A size effect occurred for four of the six concrete mixtures. Compared to the full scale specimens the spalling was reduced significantly for all smaller specimen sizes. Additionally, spalling did not occur for the small scale specimens. The results show that the specimen size is an essential parameter to investigate the susceptibility to spalling of a concrete mixture. For future investigations the testing conditions must be adjusted for the intermediate scale specimens to recreate the conditions of the slabs.

No abstract