1. Technical background and results already obtained

The theses showed that various factors were fundamental in the description of the parameters of influence on the trajectory of the boulders isolated or in groups of "average" size.
Trajectory studies were carried out using complex numerical research tools (discrete elements) that require quite specific calibration phases and calculation times that are not always compatible with the duration of a risk study by a design office.
A reduced model allowing the well-controlled study of the shape effects of blocksand substrates has been developed in the 3SR laboratory in recent years. It has allowed the constitution of a small and incomplete database.
The objective was to study the statistics of block positions in the deposition zone. From a photo of this one, the positions of 28 blocks launched 50 times are analyzed.   The objective was to characterize this distribution using simple statistical laws (gamma, Weibull etc.). These statistical laws have the advantage of being implemented very quickly as part of risk management, where the aim is to achieve a certain level of reliability (for example: no more than 3% of blocks ejected beyond a given distance).

This work is part of an "IRGA-UGA" project submitted to the MSTIC pole (SARRA-Statistical Analysis of Run-out for Risk Assessment Project, led by Frédérique
Leblanc, 3SR partnership, INRAE and Univ Bordeaux) which is in charge of the statistical analysis of the data.
The scale effect (passage from the reduced model to the real model) will be the subject of future projects, but it is necessary to identify the scientific obstacles
 

2. Problem to be addressed

. A small experimental campaign with uncensored data is to be established with, as for the previous one, campaigns of 50 throws of 28 blocks by varying the angle of slope (20 to 50°), the height of fall (0 to 20 cm), and the bedrock of thestop zone (sand or wood). Depending on the degree of progress of a parallel project of automatic processing of the run-out images obtained, the processing will be done manually (on a part of the data then only) or automatically. Eventually a model topography of a real site will be examined (3D printout of the Millau site and/or the Antone-C2ROP quarry).

. The effects of scales or similarities ("scale model/real model" effects) will be digitally processed from a simplified two-dimensional DEM code on the fall of a single block, after a prior theoretical similarity study in order to choose there levant parameters to be retained.

. Depending on the time available, a study on the independence criterion of the launched blocks (1 to 125 blocks simultaneously) will be carried out without censorship   as   well.   Indeed,   trajectography codes are well adapted for individualized falls or with block independence. In the case of numerous interactions, the behavior goes out of the trajectographic field (Thesis B.Garcia), for  the  avalanche  field  (Thesis  Cuervos),  or  even the "fluid-typical" field of  the behavior, such as mass landslides.