Tracking single bubble in Hall-Héroult aluminium cell : an experimental and numerical study

In simulations of the hydrodynamics of the multiphase flow in gas– liquid systems with finite sizes of bubbles, the important thing is to compute explicitly the time evolution of the gas–liquid interface in many engineering applications. The most commonly used methods representing this approach are: the volume of fluid and the phase field methods. The later has gained significant interest because of its capability of performing numerical computations on a fixed Cartesian grid without having to parametrise these objects (Eulerian approach) and at the same time it allows to follow the interface ( for example bubble’s shape) that change the topology. In this paper, both numerical (phase field method) and experimental results for the bubble shapes underneath a downward facing plane is presented. Experiments are carried out to see the bubble sliding motion underneath a horizontal and inclined anode. It is assumed that the bubble formed under the anode surface is deformed (flattened) due to buoyant field before it goes around the anode corner. The bubble elongates to form a tail-like shape. The change in shape of the bubble is almost instantaneous and has a significant effect on the localised hydrodynamics around the bubble, which could influence the dynamics of the flow patterns in the Hall–Héroult cell. This deformation is the main cause of the bubble wake and the induced flow field in the aluminium cell. Various parameters such as bubble size, deformation and its sliding mechanism at different surface tensions are discussed and compared with experimental results.

Hybrid neural network—finite element river flow model

Results obtained from a hybrid neural network—finite element model are reported in this paper. The hybrid model incorporates artificial neural network (ANN) nodes into a numerical scheme, which solves the two-dimensional shallow water equations using finite elements (FE). First, numerical computations are carried out on the entire numerical model, using a larger mesh. The results from this computation are then used to train several preselected ANN nodes. The ANN nodes model the response for a part of the entire numerical model by transferring the system reaction to the location where both models are connected in real time. This allows a smaller mesh to be used in the hybrid ANN-FE model, resulting in savings in computation time. The hybrid model was developed for a river application, using the computational nodes located at the open boundaries to be the ANN nodes for the ANN-FE hybrid model. Real-time coupling between the ANN and FE models was achieved, and a reduction is CPU time of more than 25% was obtained.