Underground » Strata Control and Windblasts
This project focuses on long term pillar stability and strata, as well as improvements in understanding of roadway failure mechanisms. Understanding the time dependent behaviour of coal and coal measure rocks from a mechanical viewpoint has always been a significant challenge for underground coal geotechnical engineers.
It has been extensively reported in different research that the large strains required to account for the amount of roof sagging observed in some coal mines, is likely due to creep in brittle coal measure rock and occurs because of micro-cracking along weak bedding planes and weakening of asperities of joints. In most cases the actions against creep behaviour in underground coal mines is reactive rather than being proactive. This is due to the difficulties associated with the characterisation of coal or coal measure rocks under long term testing conditions within the laboratory environment.
The objectives of this project were to investigate the time dependent behaviour of coal and coal measure rocks experimentally and analytically by specifically conducting a series of tasks.
In this work, the creep behaviour of different rock types along with prediction of their long term strength and time to failure were examined. While samples of granite, marble, and coal were tested, the limited quantity of those materials allowed for only few number of tests. The majority of the experimental program focused on Gosford sandstone, for which sufficient samples were available. A systematic laboratory analysis was conducted on this material to investigate its creep behaviour under different stress magnitudes. The results showed that the secondary creep strain rate is highly sensitive to the stress magnitude under creep testing conditions. It was demonstrated that a 10% reduction in the creep stress can lead to about three times decrease in the secondary creep strain rate. The experimental results also illustrated that there is a strong correlation between the time to failure and the secondary creep strain rate.
From the experimental results, a novel time dependent failure model was introduced. Then, the resulting log-linear equation from the experimental data was combined with the proposed model to back calculate the valid creep stress for each tested sample leading to reconstruction of accurate creep stress versus time to failure curve. In such an approach, the significant experimental scatter which is commonly observed during the time to failure test was mitigated resulting in a robust prediction of long-term strength of tested rocks. Finally, it was shown that for a range of rock types, the relationship between creep failure time and secondary creep strain rate is linear, highlighting the versatility of proposed model irrespective of testing condition and intrinsic inhomogeneity in rock materials.
The project also systematically investigated the interactive effects of shape and time on the mechanical properties of white Gosford sandstone. A number of creep tests were performed under single-step and multi-step axial loadings. The creep tests further revealed that the samples with larger slenderness ratios exhibited classical creep behaviour while those with the smaller slenderness ratios showed abrupt strain increases and localized failures. It was found that the apparent secondary creep strain rate decreases as the slenderness ratio increases. The multi-step creep tests endorsed these results in which the greater cumulative creep strains at various creep stress ratios were observed in the samples with smaller slenderness ratios.
The project outcomes concluded that the resulting failure patterns from both quasi-static and creep tests are shape dependent where the samples with smaller ratios exhibit combined shear and tensile failures, while at larger ratios, shearing is the dominant failure mechanism.