**1. Introduction**

Tailings ponds are the artificial hazard sources with high potential energy. In the event of an accident, the resulting tailings flow might bring huge losses to the life and property of downstream populations while also polluting the ecological environment severely [1–3]. Tailings are the kinds of solid wastes generated from mineral separation. Tailings are also a special kind of soil, generally structured with single granule and beehive. Due to the poor comprehensive utilization of tailings, the majority of them are stored in the tailings ponds aside from the cases of beneficiation, mine filling and building material recovery [4–6]. With the continuous exploration by mine companies as well as the growing demand for mineral products, the dam height and storage volume of tailings ponds have

been gradually increasing [7]. Meanwhile, the risk of dam failure in the tailings ponds also heightens greatly in the face of extreme conditions such as earthquakes, rainstorms and failure of flood drainage systems. Cases of severe accidents caused by tailings ponds project failure are not uncommon [8–11]. For example, two dam burst accidents occurred in Minas Gerais, Brazil, in January 2019 and November 2015, leading to at least 232 deaths and 19 deaths, respectively [7,12]. In September 2010, a tin tailings pond in China's Guangdonge collapsed, causing at least 18 deaths and direct economic losses of approximately 460 million yuan [13].

Evolution of tailings flow resulting from tailings pond leakage is a complex mechanical process, which involves multiple interdisciplinary fields including fluid mechanics, geological hazards and sediment transport mechanics [14–16]. So far, numerous scholars have estimated the flow characteristics of tailings pond dam failure such as total tailings leakage, dam breach width and maximum tailings flux through theoretical derivation, statistical analysis and other approaches and also formed relevant theoretical formulas [17–21]. With the rise of computer simulation software, plentiful fluid dynamics software has been applied to the evolution pattern research of tailings flow resulting from tailings pond leakage [22–25]. The fluid rheological equation embedded in FLO-2D, a US Federal Emergency Management Agency (FEMA)-approved flood and mudflow hazard simulation software [26,27], can be used for the simulation of the rheological state during evolution of leaked tailings flow.

This study explores the evolution of leaked tailings flow following dam failure and its effects on the downstream structures and personnel during evolution. Using the FLO-2D simulation software (FLO-2D Software Inc, Nutrioso, AZ, USA), the evolution process of tailings flow is studied under different downstream riverbed slopes and debris blocking dam construction parameters. At the same time, the evolution pattern, sedimentary characteristics and energy loss of tailings flow during evolution are analyzed. In addition, the effects of downstream riverbed slope and debris blocking dam construction parameters on the tailings flow characteristics are also investigated.
