Dry stone retaining walls (DSRWs) are vernacular structures made of rubble stones without any mortar between the blocks. The have shaped extended territories and landscapes in the whole Europe (including France) allowing to mitigate the risk of slope erosion while the terraces that were created were used for agricultural purposes. Many of them were were also incorporated in transportation facilities such as roads or railways. Decades of neglect have induced damages that made their repair inevitable before total destruction. Because of a lack of standards for this kind of construction technology, project supervisors have some difficulties to process the required intervention on these structures.

The last two decades have gain perspectives in this respect due to emerging researches on the mechanical behavior of DSRWs. At present, the understanding of the static behavior of plane slope DSRWs is almost achieved and different techniques are available to study and design these structures including the yield design approach, a fully discrete element method (DEM) or a mixed DEM-continuum approach. However, the seismic behaviour of DRSWs is an open issue though numerous walls have been erected in mountainous regions where earthquakes can take place.

In a first attempt to bridge the gap in this field, pseudo-static scale-down (scale divided by ten) experiments on retaining walls have been carried out. The pseudo-static approach is a simplified approach recommended by Eurocode 7 for the seismic design of structures. The tests consisted in tilting a mock-up composed of a wall retaining a backfill until the failure of the wall was reached. The walls were built with small clay bricks that were placed in such a way that the state of the art in actual walls's construction was satisfied. The backfill was composed of loose Hostun sand. Then, the critical tilting angle of the mock-up leading to failure was reported and video recordings were done to have a better insight into the mechanisms at stake during failure.

After some calibration and repeatability tests have been performed, different geometric configurations for the retaining wall (length, width, height) have been studied. Depending on the geometry of the wall, two failure mechanisms have been observed:
-Sliding: For walls with a small ratio height/width typically smaller than 1.7.
-Overturning: For walls with a ratio height/width greater than 1.7.
In the case of failure by sliding, the complete wall fell down whereas in the case of failure by overturning, a part of the wall stayed almost perfectly in place. These features are consistent with previous studies on small-scale and full-scale DSRWs. Moreover, and as expected, the tilting angle leading to failure was found decreasing while the slenderness of the wall increased.

An analytic method based on the Coulomb wedge theory has been developed to model the pseudo-static behavior of the tested retaining walls. A very good agreement with the results of the experiments was found with a departure for the tilting angle leading to failure of about one degree. Then, this method seems to be precise enough to be used for the seismic design of DSRWs.