Three-dimensional CFD-DEM simulation of raceway transport phenomena in a blast furnace

in Fuel (2023), 334(2),

Improving energy efficiency in a blast furnace (BF) has a significant effect on energy consumption and pollutant emission in a steel plant. In the BF, the blast injection creates a cavity, the so-called ... [more ▼]

Improving energy efficiency in a blast furnace (BF) has a significant effect on energy consumption and pollutant emission in a steel plant. In the BF, the blast injection creates a cavity, the so-called raceway, near the inlet. On the periphery of the raceway, a ring-type zone is formed which is associated with the highest coke combustion rate and temperatures in the raceway. Therefore, predicting the raceway size or in other words, the periphery of the ring-type zone with accuracy is important for estimating the BF’s energy and coke consumption. In the present study, Computational Fluid Dynamics (CFD) is coupled to Discrete Element Method (DEM) to develop a three-dimensional (3D) model featuring a gas–solid reacting flow, to study the transport phenomena inside the raceway. The model is compared to a previously developed two-dimensional (2D) model and it is shown that the assumptions associated with a 2D model, result in an overestimation of the size of the raceway. The 3D model is then used to investigate the coke particles’ combustion and heat generation and distribution in the raceway. It is shown that a higher blast flow rate is associated with a higher reaction rate and a larger raceway. A 10% increase in the inlet velocity (from 200 m/s to 220 m/s) caused the raceway volume to grow by almost 40%. The DEM model considers a radial discretization over the particle, therefore the heat and mass distributions over the particle are analyzed as well. [less ▲]

Prediction of the biomass particles through the physics informed neural network.

in ECCOMAS Congress 2022 - 8th European Congress on Computational Methods in Applied Sciences and Engineering (2022)

Woody biomass energy is a kind of renewable energy that contributes to the reduction of greenhouse gas emissions, the creation of healthier forests, and the reduction of wildfire danger. Simulations of ... [more ▼]

Woody biomass energy is a kind of renewable energy that contributes to the reduction of greenhouse gas emissions, the creation of healthier forests, and the reduction of wildfire danger. Simulations of biomass combustion, in general, are time-consuming simulations with a large number of input particles. We use a deep hidden physics-based neural network model to predict the behavior of particles throughout the simulation based on the equations of motion to achieve an efficient simulation and reduce the processing effort. We replace discrete element methods with inverse methods, which have the advantage of simulating velocity fields without knowing the simulation's boundary and initial conditions. Reconstruction of the velocity fields is done using a recurrent neural network in conjunction with a physics-based loss function. The proposed model is suitable for modeling problems that involve moving particles in a fixed bed. The number of neurons and activation functions in the artificial neural network are optimized, and the effect of the sampling method and the number of outputs are studied. [less ▲]