Increasing demand for steel has prompted the increase in production capacities for hot metal and need to improve the efficiency in the blast furnace operation. In the handling of the raw materials used in ironmaking, and during their charging and descent in the blast furnace, undesired segregation and percolation phenomena may occur. In large silos used for intermediate storage, smaller fines of the material may accumulate and purge non-uniformly due to complex particle flow mechanics. In the blast furnace, iron ore percolation into underlying coke layers can reduce the voidage of the coke layer, which affects the pressure loss and distribution of the gas. In extreme cases, this may contribute to process disturbances in the furnace. This thesis studies percolation computationally using the discrete element method (DEM). This simulation approach has the advantage of providing visual and graphical insight into percolation in the system studied, which cannot be obtained from the system due to the opacity of the granular material involved and the hostile environment of the real systems. Over the years, the discrete element method has also evolved, making it possible to study more realistic systems today, partly due to improvement and parallelisation of the code and the growing computational speed. As the application of the technology is still limited by computational resources available, the systems studied must be down-scaled to make the solution feasible. The effect of different material and process parameter were investigated. Size ratio, descent velocity vibration amplitude and frequency were found to enhance percolation of pellets in underlying coke beds. Some ways to reduce the extent of percolation were also briefly discussed, the most important being reducing the size ratio between the large and small particles. In silos, suppressing large changes in the filling and discharge heights may be a remedy.
|Tila||Julkaistu - 2019|
|OKM-julkaisutyyppi||G2 Pro gradu -työ, ammattikorkeakoulun laajennettu opinnäytetyö|