| Author: | Dai, Baolin |
| Title: | Hydro-mechanical behaviour of unsaturated and structured loess |
| Advisors: | Zhou, Chao (CEE) |
| Degree: | Ph.D. |
| Year: | 2025 |
| Department: | Department of Civil and Environmental Engineering |
| Pages: | xxviii, 259 pages : color illustrations |
| Language: | English |
| Abstract: | Loess is characterised by a metastable and anisotropic structure formed during aeolian depositional history. It is usually unsaturated due to its prevalence in arid and semi-arid regions. Many investigations have demonstrated that soil structure, including that of structured loess, can greatly affect engineering behaviour. Most of the previous studies on structure effects, however, have focused on saturated conditions. The hydro-mechanical behaviour of unsaturated and structured soils remains insufficiently understood. The principal objectives of this research are to investigate the influence of structure and its evolution on unsaturated soil behaviour. To achieve this, three series of tests were conducted using loess specimens with different initial structures: intact specimens, reconstituted specimens, and compacted specimens with varying compaction dry densities and water contents. First of all, isotropic compression tests with shear wave velocity measurements were performed on saturated specimens to primarily examine the evolution of soil microstructure and anisotropic stiffness under mechanical loading. Then, building on the results, cyclic triaxial shear tests were conducted on unsaturated specimens to investigate micro-macro behaviour under combined mechanical and hydraulic loading. Finally, cyclic hollow cylinder torsional tests were performed on unsaturated specimens to investigate the coupled effects of mechanical, hydraulic, and thermal loads, as well as principal stress rotation (PSR), on the cyclic shear behaviour of structured soils. On the other hand, a bounding surface model was developed to capture the behaviour of unsaturated soils under both monotonic and cyclic loading conditions. The effects of soil structure and anisotropy on the hydro-mechanical response of unsaturated soils are newly incorporated into the model. In parallel, the potential of ground granulated blast-furnace slag (GGBS) as a stabilisation material for mitigating loess collapse was explored. Experimental results indicate that, during isotropic compression, the initial structure in the intact specimen is stiffer and more orientated than in the compacted and reconstituted specimens, as evident by a larger G0/f(e) (elastic shear modulus normalised by a void ratio function) and a more evident stiffness anisotropy in the former. The initially stiffer and more orientated structures are also observed in the compacted specimens with a higher dry density and a lower compaction water content. More importantly, it is observed that when the applied stress reaches about 2 to 4 times the yield stress of the initially stiffer specimen, G0/f(e) becomes smaller than that of the initially softer specimen. This reversal is likely due to more pronounced interparticle contact strengthening and greater pore compression in the latter, as supported by the results of microstructural analysis. This new finding highlights the necessity of considering the important role of soil structure evolution in stiffness. Under cyclic shearing, the influence of soil structure becomes increasingly complex. In the first cycle, the intact specimen may exhibit either a smaller permanent strain (∆εpz,1) than the compacted specimen (type-I behaviour), attributed to its inherently stiffer soil structure, or a larger ∆εpz,1, due to structure degradation-induced additional plastic strain (type-II behaviour). As the number of cycles increases, the intact specimen in type I may have a larger strain increment due to structure degradation. In type II, the compacted specimen may experience a larger strain increment as a result of failure. For compacted loess specimens prepared under different compaction conditions and subsequently adjusted to the same suction, permanent strain decreases and resilient modulus increases with a higher dry density and a lower compaction water content. These effects of compaction conditions become more pronounced as suction decreases, possibly because the soil skeletons of specimens compacted at higher water contents and lower densities are more sensitive to suction changes. The above results demonstrate the important coupled effects of structure and suction. Under cyclic shearing with the same change in the principal stress magnitude, specimens subjected to PSR exhibit larger excess PWP and permanent vertical strain, and smaller resilient modulus than those without PSR. These differences are because rotational loading can induce additional deformation in anisotropic soils. More importantly, the extra strain induced by PSR increases with increasing temperature and decreasing suction. When PSR is considered at zero suction, permanent strain increases by 130% and 230% at 5 and 40 ℃, respectively. As suction increases to 10 kPa, these increases reduce to 50% and 80%. These coupled effects are likely attributed to the greater influence of PSR at lower overconsolidation ratios. These new findings highlight that accounting for structure evolution is essential to enhance the reliability of deformation predictions in engineering practice. Specifically, when both compacted and intact soils are involved, it is not always conservative to adopt parameters from the compacted soil, as the strain increment could be smaller in certain cases. Moreover, strict control of compaction conditions is crucial, given their pronounced influence on soil response to cyclic loading, particularly under saturated conditions. Finally, the effects of PSR on resilient modulus should be considered, especially at elevated temperatures and low suctions. The proposed model newly incorporates the effects of soil structure and anisotropy on both the loading-collapse (LC) curve and the soil water retention behaviour. Experimental results from the literature reveal that the inclination of the normalised LC curve for structured soils can be smaller than, larger than, or equal to that of the reference soils, and these variations are captured within a unified modelling framework. Additionally, the model well represents the increased water retention capacity associated with greater anisotropy, an aspect not incorporated in existing models. The predicted results are well matched with experimental data under both monotonic and cyclic loading, demonstrating that the model effectively captures the effects of structure and anisotropy on the hydro-mechanical behaviour of unsaturated soils. For loess improvement, it is found that the collapse index of loess decreases with increasing binder content (defined as the mass ratio of lime and GGBS to dry soil), with more pronounced effects observed at higher GGBS-to-binder ratios. For example, as the lime content increases from 0% to 3%, the collapse index decreases from 15.3% (severe collapse) to 4.7% (moderate collapse) for looser specimens. More importantly, substituting 50% of the lime with GGBS further reduces the collapse index to 0.06%, which falls below the lower limit for slight collapse (i.e., 0.1%). These observations indicate that the partial replacement of lime with GGBS can effectively reduce the collapse of loess and satisfy engineering requirements. The results of this study have enhanced the understanding and modelling of unsaturated loess. These findings may also be applicable, at least qualitatively, to the analysis of other unsaturated and structured soils. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/14218