Multiscale characterization of loess

Multiscale characterization of loess in Hesbaye region (Belgium) with a coupled geological and geotechnical approach. 

Objectives

The aims of this PhD research are (i) to evaluate the heterogeneity of the mechanical properties of the loess sequence, (ii) to define stratigraphic markers within the loess sequence based on common geotechnical tests and (iii) to explore some existing mechanical relations between in situ and laboratory analyses.

Summary

A thorough analysis is performed on aeolian silt (loess) deposits. The study is based on an original approach combining geological aspects with physical and mechanical properties of the material. This approach is realized by means of a multiscale experimental analysis in which cone penetration test plays a key role. The approach is applied to the study of loess deposits located in the Hesbaye region (Belgium), in which the site of the Romont quarry (Eben-Emael, Belgium) is the subject of a more comprehensive and a more detailed study.

Figure 1 : Presentation of results from the investigated site of Rocourt (Liege, Belgium).

About the stratigraphic aspects, a continuous geological cross-section performed along a 18 km long section of the high-speed railway between Liège and Waremme, underlines the variability of the thickness of the loess layer which can locally reach 20 m. The loess sequence is further investigated in three Hesbayan sites (Remicourt, Romont and Rocourt) with cone penetration test (CPT) campaign. The analysis of the cone resistance of the CPT reveals it is possible to define four stratigraphic markers within the regional loess sequence: the near surface decarbonation limit, the transition between the Nagelbeek Tongued Horizon and the underlying Hesbayan loess, the Humiferous Complex of Remicourt, and the Whitish horizon of Momalle on top of the Rocourt Pedocomplex. Physical, geochemical and microstructural characteristics determined from laboratory analyses on samples stratigraphically located around the stratigraphic markers, enable a better understanding of the observed variations of the in situ mechanical behaviour. Porosity, volumetric weight and clay content seem to be three major variables in the understanding of the observed variations of cone resistance.

A comparative analysis of geomechanical properties has also been conducted either from different data related to CPT or between CPT results and laboratory measurements. This study includes four items: (i) the comparative analysis of measurements achieved by different CPT cone types (M1, M2 and E), clearly indicates the cone type has negligible influence on the measured cone resistance for silty material. (ii) Soil behaviour type charts based on CPT data confirm the relative heterogeneity of the investigated silty material. (iii) The empirical relation between the oedometric modulus and the cone resistance of the CPT, including the α factor depending on the soil type, is verified for the studied silty material. (iv) A semi-empirical approach based on the bearing capacity theory has been developed to estimate mean apparent cohesion and friction angle of the silt.

Finally, results of this PhD thesis may be used for specific and original applications: (i) CPT testing seems to be an interesting investigation technique for rescue archaeology in loess context. (ii) Thanks to the in-depth investigation, the Romont site (Belgium) may provide an interesting site wherein sensors or investigation tests related to geomechanical, physical or structural properties of the material may be tested or calibrated into a natural silty medium. (iii) The detailed characterisation of aeolian silt leads to numerous properties related to geomechanical, physical and geochemical aspects of the material. These data may be used in studies dealing with loess deposits.

Figure 2. Pedo- and lithostratigraphic description of the observable section of the working face from the CBR’s quarry of Romont (Eben-Emael, Belgium). The location of the cone penetration tests (CPT) performed behind the section is also presented. Abbreviations: HCR = Humiferous Complex of Remicourt; Ki-A = Kincamp A Horizon; MAL = Malplaquet Soil; NTH = Nagelbeek Tongued Horizon; RP = Rocourt Pedocomplex.

Figure 3. Cone penetration (q c ) results from the CPT survey carried out at the site of Romont (Eben-Emael, Belgium) and resulting stratigraphic interpretation.

Figure 4. Reconstitution of cone resistance results from cone penetratoin testing through the loess sequence.

Funding

University of Liège (teaching assistant)