Abstract
If one is to deal with experimental analysis of soil-foundation interactions, the first problem to consider is that of reconstituting soil models. This paper presents the results obtained by using a sand-spreader which was designed and built by the author for this purpose. The paper also looks at design criteria and the calibration of the assembly. The dimensions of the sand sample, which is sprinkled into and contained inside an iron caisson, are 1.60 meters in width, 2.50 meters in length, and 1.70 meters in height, giving a total volume of 6.80 cubic meters. The preparation of sand beds can be accomplished with relative densities ranging from 25% to 70% and can be repeated within an error margin of ±1% ; the relative density inside the bed itself is fairly constant, having a mean standard deviation equal to 1.47%.
