Characterisation of collapsing loess by seismic dilatometer

Abstract

Structural collapse and sudden volume changes represent a major geotechnical issue, particularly in loess soils. There are different criteria for assessing collapse potential based on laboratory test results, which require the collection of undisturbed samples from test pits. This can be a complicated and costly procedure, often financially unjustified for smaller projects. This paper presents the results of seismic dilatometer tests (SDMT) performed at a single location in a loess soil in Belgrade, in addition to the results of single oedometer collapse tests performed on high quality samples from test pits.After comparing laboratory test results it was possible to determine that in samples collected from boreholes, dry unit weight is approximately 20% higher than in block samples taken from test pits, while moisture content is between 1 and 4% higher. Unit weight estimated from DMT is on average 15% higher than the unit weight obtained from block samples.The constrained modulus (MDMT) determined by DMT are highly compatible with oedometer modulus (Eoed) obtained from samples at natural moisture content. This paper presents two possible ways of identifying collapsing loess based on intermediate DMT parameters and the ratio of G0/MDMT. The tests results indicate that if the ratio between the material index (ID) and the horizontal stress index (KD) is greater than 5, the danger of collapse is imminent. Simultaneous observation of the changes of ID and KD with depth on a semi-log graph in the same scale is recommended. In this way the relative distance between them may be clearly noticed; the larger the distance becomes, the greater the risk of collapse. In the collapsing loess ratio G0/MDMT is found to be higher than 21 for KD less than 0.6, while in non-collapsing loess G0/MDMT is less than 21 for higher values of KD. In terms of DMT, loess is considered as underconsolidated soil, which is one of the most common “definitions” of loess.