Comparing the different test methods for concrete resistance to scaling and the crucial lack of correlation between the laboratory test results and the field results, their representativeness can be questioned. Many researchers tried to clarify most of the mechanisms involved in the damages due to scaling, as well as thermal parameters influencing damage intensity. However, it is remarkable that the prediction of the degradation in relation to mix design characteristics remains unreliable and seems to be disconnected from the performance on site. Moreover, the interpretation of the test results also remains difficult, appearing in contradiction with the first order parameters e.g. minimum freezing temperature or freezing rate. These considerations led us to undertake some measurements campaigns in order to characterize the concrete's temperature response to severe freezing environment, this environment being itself fully characterized by a set of relevant meteorological parameters measured on field site, very closed to the samples.

In this paper, we present the results of three years of meteorological measurements (2013-2016) from two concrete blocks placed in a complete meteorological station located in a severe freezing environment. The temperature measurements were taken on one hand from an XF4 concrete block equipped with thermocouples placed at 2 and 8.5 cm deep and on the other hand from the temperature of the air mass coming from a Stevenson shelter with double shelf in opposite orientation, both compared to the laboratory thermal cycle according to CEN/TC12390-9 (Slab test). If this latter is giving a very symmetrical and periodic temperature signal, the first one is rather characterized by a strong asymmetry and irregularity in profile and full amplitude. We show that the natural conditions are very far from the laboratory conditions in terms of full amplitude, freezing rates and frequency of freezing-thawing cycles. In addition, we show that these field conditions are representative of severe freezing environment according to EN 206 standard, in the French Alps as well as in the Nordic conditions which can be found in Norway, Sweden and Finland.