Search

Reconstituting soil beds to a desired density is essential to geotechnical modeling tests. In this study, we apply and assess the method of air fluidization to prepare sand beds for geotechnical engineering studies. Through it, an automated bed preparation process can be realized. The details of device structure and design are presented as a reference for application of the methodology. By quantifying the average and local post-fluidization density of the bed, the performance of the fluidized bed device is characterized. With the addition of vibration and by changing the defluidization rate, the sand can be prepared with volume-based relative densities ranging from 10% to 92%. Local sand density, measured with a cone penetrometer, is nearly uniform across the bed: density variation is less than 13% (coefficient of variation with respect to relative density) for all protocols except for some beds prepared by defluidization only. The variation of local density and penetration resistance measured across the bed breadth is comparable to results from beds prepared by the commonly used method of pluviation. This suggests that sand beds reconstituted using air fluidization are suitable for geotechnical modeling tests.