The numerical prediction of the runout and spread of liquefied-tailings flows is a complex problem that depends on many factors, including the rheological properties of the liquefied tailings. However, published… Click to show full abstract
The numerical prediction of the runout and spread of liquefied-tailings flows is a complex problem that depends on many factors, including the rheological properties of the liquefied tailings. However, published benchmark problems specific to tailings flows, useful for validation and calibration of numerical models, are virtually nonexistent. This paper presents a laboratory-scale benchmark problem of liquefied-tailings flow. Gold filtered tailings were characterized via rheological measurements, geotechnical index tests, and toxicity chemical analysis. Physical flow experiments of the liquefied-tailings paste, at 70% solids concentration, were carried out in an instrumented laboratory flume with high-speed video and direct measurements of the at-rest “footprint” (lobe) dimensions. Subsequently, using the measured physical parameters, computational fluid dynamics (CFD) tools were used to solve the three-dimensional, rheology-dependent Navier–Stokes equations via the finite-volume method and a multiphase volume-of-fluid (VOF) technique. Thus, the at-rest lobe of the spilled tailings was numerically reproduced. Results show that the liquefied tailings bear nearly zero-yield stress and low viscosity, thereby practically behaving as a Newtonian fluid despite their high solid concentration. In addition, good agreement (within 14% of the main dimensions) was found between the physical and numerically simulated at-rest lobes. Hence, the use of a Navier–Stokes approach, supported on a finite-volume/VOF technique, and a Newtonian-fluid constitutive rheological model, simulates well the at-rest shape of liquefied tailings at laboratory scale. This benchmark problem will aid numerical research specific to tailings flows.
               
Click one of the above tabs to view related content.