Study on Mechanical Characteristics of Composite Alkali-activated Materials-stabilized Gold Tailings Under Direct Shear

: In this study, the traditional industrial waste residue and some alkaline activators were mixed to prepare 14 a new composite alkali-activated materials (CAAMs), which was used to stabilize gold mine tailings (GMTs). 15 Due to emissions of greenhouse gases and solid dust, alkali-activated materials have been widely used to 16 replace Portland cement to solidify geotechnical materials to enhance their mechanical properties. Different 17 admixture of CAAMs (i.e., 0, 3, 5, 8% ) and gold mine tailings were prepared, and the samples were cured in 18 saturated water and under no air conditions. In order to investigate the mechanical characteristics of CAAMs- 19 stabilized GMTs, laboratory direct shear tests were carried out on samples after curing them for 3, 7, 14 and 20 28 days, respectively. The test results showed that as the curing periods increased, the brittleness of the samples 21 increased, and the stress-displacement curves for all the cured specimens changed from plateau-type to peak- 22 type curves. The curing periods and the content of CAAMs are both beneficial for enhancing the shear strength 23 of CAAMs-stabilized GMTs samples, but the increase rate decreased as the vertical confining pressure 24 increased. Furthermore, the influence of CAAMs content on shear strength increment was larger than that of 25 curing periods. An exponential growth model could be well used to describe the change of shear strength with 26 the curing periods at different vertical stresses.


Introduction
Portland cement has been widely used to stabilize geotechnical materials and to increase their strength physical and mechanical properties. In addition, their hydration fast hardness, corrosion resistance, freeze-46 thaw resistance and thermal stability are also excellent. Therefore, alkali-activated materials instead of 47 Portland cement have been widely applied to solidification in the field of geotechnical engineering. 48 Mining industries around the world generate huge amounts of residues and tailings. Mine tailings are 49 finely ground geotechnical materials after the valuable minerals and metals extraction operations (Zhang et al. 50 2020). Due to high cost for migration, the main treatment method of tailings is in-situ storage behind tailings 51 dams, which requires adequate site selection . With economic development for ore and metal 52 demand and land use restrictions, it is necessary to build more large tailings dams to store more tailings. 53 Tailings contain complex mineral components. If the tailings dam breaks, it will cause serious pollution to the 54 environment and may cause grave loss of people's lives and property (Kossoff et al. 2014). On April 30th, 55 2006, a destruction occurred in a gold mine tailings reservoir in Shanxi province, northwest China, causing 56 17 deaths, 2 people missing and 5 injuries, as well as 76 houses destroyed and flooded (Ju et al. 2012). On 57 January 25th 2019, the structure damming a pond containing mine tailings break down at Brumadinho City, 58 Brazil. About 11.7 million m 3 of a tailings-mud mixture was released from the dam, resulting in damage 300 59 km along the Paraopeba River toward the São Francisco River (Owen et al. 2020). 60 Many studies have been conducted on the impact of different curing agents on enhancement of the stability 61 and mechanical properties of geotechnical materials. Curing agents mainly include cement (Qiao et al. 2007), 62 lime (Boardman et al. 2001), and polymers ) and so on. However, these have certain limitations 63 for improving the properties of tailings. In most cases, the tailings materials are substituted as river sand to be 64 stabilized by curing agents, which are used as the subbase. That is construction component for engineering 65 purposes of mining filling (Chu et al. 2018), subgrade (Yin et al. 2012), wall (Ahn et al. 2011). Few researches 66 focus on the influence of curing agents on mechanical properties of tailings with respect to tailings dam 67 stability. Thus, it is urgent to study the mechanical properties of CAAMs-stabilized tailings, which is not only 68 friendly to the environment, but also can consider the feasibility of replacing cement-based curing agents to 69 improve the safety and stability of tailings dams from the perspective of economy.

70
The safety of tailings dams and the economic requirements of mining enterprises are both considerable. The gold mine tailings used in this study were obtained from Longnan Gold Mine in Gansu province,

81
China. The dry beach surface of the tailings dam is shown in Fig. 1. The color of the raw tailings is charcoal 82 grey with a little particle aggregation in water and mainly fine particles. Uniformity coefficient Cu and 83 Curvature coefficient Cc are 10.01 and 1.23 respectively. The GMTs have a poor gradation according to the 84 Unified Soil Classification System (USCS) (Stevens 1982), with a grain-size distribution curve shown in Fig.   85 2. The average grain size is 0.015mm. The specific gravity of the gold mine tailings is 2.77. Its initial water 86 content is 15.9%. An X-ray diffractometer (XRD) (Panakot Aeris, Netherlands) was used to analyze the mineralogy of the 92 gold mine tailings using a quantitative Rietveld XRD method (Scrivener et al. 2004). The highest characteristic 93 peak in the mineral composition of gold mine tailings is quartz, and the main mineral components are quartz 94 and illite, with a combined content of nearly 90%, which plays a leading role in the strength of tailings. The   According to particle size and the Chinese standard for soil direct shear test method (GB/T50123, 1999), 120 samples for the direct shear tests are in the shape of a circular disc with 61.8 mm in diameter and 20 mm in 121 thickness. The procedures for the stabilized sample preparation are as follows:

122
(1) Preparation of raw materials, gold mine tailings and CAAMs are dry, as shown in Fig. 6 (1). So as to 123 mix the raw materials uniformly and prevent the moisture from overflowing during the compression of the 124 sample, samples were prepared with a moisture content above the plastic limit and below the liquid limit of 125 gold tailings, and the water content of the sample is set to 20 %.

126
(2) Preparation of mixture. According to the targeted CAAMs content, the corresponding gold mine 127 tailings, CAAMs and needed water were weighed. The raw materials were thoroughly stirred and mixed to a 128 uniform sample, and then the required water was added for further stirring together to prevent excessive local 129 CAAMs in the sample from causing faster hydration resulting in uneven strength of the sample. The obtained 130 mixture prepared for the experiment is shown in Fig. 6 (2).

131
(3) Sanples maintenance. In accordance with the compaction state of the gold mine tailings, the initial dry 132 density of samples is set to 1.5g/cm 3 . The mixture is carefully added to a standard direct shear sampler to ensure that each sample has the same quality. The compressed samples with the cutting ring were removed 134 from the sampler, and each sample is separated by a glass sheet, as shown in Fig. 6 (3).

135
(4) Sample curing. Finally, in order to prevent the sample from breaking down easily in the water, the 136 samples were cured at room temperature (25°C) and 90% humidity for one day. Then, all the samples were 137 cured by immersion in water to simulate the curing environment in saturated state for tailings, as shown in    According to the ratio of curing agent and soil designed by Chen (2006) and Sukontasukkul (2012), the 158 proportion of the CAAMs and gold mine tailings is expressed by the mixture for weight ratio of dry tailings 159 (S), CAAMs (C) , and water (W). CAAMs content (Cw ) is defined as the ratio of the mass of CAAMs to the 160 that of tailings (mC/mS). Moisture content (Aw) is defined as the ratio of the mass of water to the sum of the 161 mass of CAAMs and gold mine tailings (mW/(mC+mS)). For the convenience of narrative, the samples are 162 numbered in GTAAx-y format, wherein, GTAA represents gold tailings stabilized by composite alkali-163 activated materials. x is the content of CAAMs added (e.g. 5%); y is the curing period (e.g. 7 days).

164
According to the shear force measured in the shearing tests, the shear stress on the shear surface can be 165 calculated considering the continuous reduction of the shear area. Generally speaking, the direct shear test is 166 on the basis of the hypothesis that the stress distribution in the sample is uniform when a soil sample with a 167 certain thickness is subjected to shear at plane strain. Therefore, the shear stress along the shear failure surface 168 can be calculated using the following Eq .(1), which is obtained from the displacement parameters recorded 169 by the movement of the shear box.
Where T is the shear force applied to the shear box; d is the diameter of the sample, d = 61.8mm; s is 172 the shear displacement; when s=0, the shear surface area is 30cm 2 , it is the initial cross-sectional area of the  Table 1. 187         (kPa) C w =3%, 7days C w =3%, 28days C w =5%, 7days C w =5%, 28days C w =8%, 7days C w =8%, 28days  Table 1. The effect of 304 curing period on cohesive force and internal friction angle of the stabilized-GMTs is shown in Fig. 13(a). The 305 cohesive strength and internal friction angle increase rapidly at first and then slowly with the curing period.

306
During the curing period from the beginning to 14 days, the cohesion strength increases from 6 kPa to 21 kPa, 307 and the internal friction angle increases from 25° to 32°, whereas the cohesive strength is around 21-22 kPa 308 and the internal friction angles are in the range of 32° and 33° between 14 and 28 days of curing. It indicates 309 that the influence of curing period on internal friction angle is significant in the early stage of curing, and its 310 influence is gradually negligible in the later stage. It can be predicted that the curing is completed around 28 311 days.  Where p I is the relative increase rate; GTAA C is the curing period of GATT3-28 or the CAAMs content of GATT8-7; C is the curing period or CAAMs content.

344
It can be seen from Fig. 14  probably due to the stronger compression properties for the sample with a higher content of CAAMs.

349
Furthermore, the increase of cohesive force caused by CAAMs content is also greater than that caused by 350 curing period.