DBPapers
DOI: 10.5593/sgem2017/12/S02.032

DETERMINISTIC AND GEOSPATIAL APPROACH TO ESTIMATE THE COLLAPSE BEHAVIOR OF QUATERNARY LOESS DEPOSITS

C. F. Dobrescu, E. A. Calarasu, A. E. Huzui-Stoiculescu
Sunday 10 September 2017 by Libadmin2017

References: 17th International Multidisciplinary Scientific GeoConference SGEM 2017, www.sgem.org, SGEM2017 Conference Proceedings, ISBN 978-619-7105-99-5 / ISSN 1314-2704, 29 June - 5 July, 2017, Vol. 17, Issue 12, 247-254 pp, DOI: 10.5593/sgem2017/12/S02.032

ABSTRACT

Due to large widespread of loess deposits in many parts of the world and reported geotechnical problems arising from collapse susceptibility due to soil sensitivity to wetting changes associated with subsidence and damages, advanced and detailed analyses for quantification of specific characteristics are necessary in order to select the optimal technical solutions in terms of structure safety and stability. This paper presents the analysis of various experimental testing results, with particular emphasis on the collapsible parameters of the loess soils collected from different sites located in Dobrogea region, Romania. The determined geotechnical data (grain size distribution, water content, soil density in natural and dry states and physical indices as porosity, void ratio, saturation degree, plasticity limits, specific settlement at wetting, deformability and shear strength parameters) were input into a GIS framework to provide spatial variability maps of collapsible behavior and to identify the most risk prone sites using raster analysis. Considering the ability to manage a specific database and to develop thematic maps, application of GIS tools in preliminary geotechnical site evaluations and urban planning can provide indicators on spatial variation of soil properties in order to achieve better land-use planning, necessary remedial measures or ground improvement solutions for civil engineering structures founded on unstable soils.

Keywords: collapsibility, geospatial analysis, geotechnical data, GIS maps