Application de la methode d'analyse spectrale des ondes de surface a l'evaluation des effets du gel sur les infrastructures routieres (French and English text)

The Spectral Analysis of Surface Waves (SASW) test has been used for studying the significant changes in elastic properties of pavements affected by freeze/thaw cycles. However, when the investigated medium includes a stiff layer, such as an asphalt concrete surface layer or a frozen soil layer in depth, higher modes of propagation of surface waves occur simultaneously and must be taken into account in the analysis of the SASW results. The principal purposes of this thesis are to carry out a forward modelling of the propagation of surface waves in a layered medium in order to study the multimode propagation, to adapt the SASW test for pavement evaluation and to assess the SASW possibilities for monitoring the seasonal variation in elastic properties of pavements in cold regions.; A Modified Linearized Stiffness Matrix (MLSM) method has been developed to provide the vertical displacements at the surface of a layered medium under steady-state loading. This simple and fast algorithm compares the higher modes of propagation of surface waves to the eigenfrequencies of a beam. A theoretical study has shown that the use of more than one pair of receivers enables the identification of the simultaneous propagation of several modes of surface waves with a fast and very simple method. These higher modes have been also detected during experimental measurements carried out in a partially frozen pavement. A field methodology and a data reduction algorithm have been also developed to take into account the simultaneous propagation of several modes of surface waves.; SASW tests have been carried out at regular interval form March to September 2001 in an urban pavement in Quebec City. A weakening period of stiffness modulus in the base, subbase and subgrade soils has been observed during the thawing period followed by a fast recovery of the stiffness in summer. Results clearly show that the SASW test is adequate for detecting and measuring the seasonal variations in elastic properties of pavement layers in cold regions.