Computational analysis of vertical and horizontal jet fires

Experimental Research on the Combustion Characteristics of Transformer Oil Jet Fires in Oil-Filled Equipment under Heat

Under the action of heat, the jet fire of oil-filled equipment in substations produces intense heat radiation and flame impact, causing serious harm to the personnel and equipment. Therefore, this paper has carried out a series of oil-filled equipment transformer oil jet fire experiments with different nozzle diameters to study the characteristics of transformer oil jet flame combustion. The experiment measured the mass loss rate, axial centerline temperature of the fire plume, and radiant heat flux and processed the flame visualization image. The results show that the combustion process of the transformer oil jet fire can be divided into three stages. The functional relationship between the flame height and mass flow and the functional relationship between the flame temperature and axial height are derived. A prediction model of the radiant heat flux of the transformer oil jet fire for the oil-filled equipment of the substation was established. The research results are of great significance for improving the fire protection design of substations and improving fire management and control capabilities.

A comparative study of radiation models on propane jet fires based on experimental and computational studies

Radiation as a consequence of jet fires is one of the significant parameters in process industry events. In the present work, the open field vertical propane jet fire was studied via experimental and computational fluid dynamics (CFD). The predicted values of radiation were verified at three locations in the horizontal direction from the jet fire. In the simulation section, four radiation models of Monte Carlo (MC), P-1, Discrete Transfer (DT), and Rosseland were applied to find the fine model for simulating the jet fire. Shear Stress Transport (SST) and Eddy Dissipation Concept (EDC) models are employed for combustion and turbulence, respectively. The estimated data by the simulation demonstrated that the MC radiation is better than the other models with an average error of 5% for predicted incident radiation from the jet flame axis. Also, the P-1 radiation model had an above 65% error at around the jet fire, but due to the error of less than 15% estimated by MC and DT models, these radiation models could simulate the jet flame radiation. The simulation outcomes proved that the Rosseland radiation model is not applicable owing to a lack of accurate temperature prediction.

Analysis of deep learning neural network combined with experiments to develop predictive models for a propane vertical jet fire

Fires are important responsible factors to cause catastrophic events in the process industries, whose consequences usually initiate domino effects. The artificial neural network has been shown to be one of the rapid methods to simulate processes in the risk analysis field. In the present work, experimental data points on jet fire shape ratios, defined by the 800 K isotherm, have been applied for ANN development. The mass flow rates and the nozzle diameters of these jet flames have been considered as input dataset; while, the jet flame lengths and widths have been collected as output dataset by the ANN models. A Bayesian Regularization algorithm has been chosen as the three-layer backpropagation training from Multi-layer perceptron algorithm. Then it has been compared with a Radial based functions algorithm, based on single hidden layer. The optimized number of neurons in the first and second hidden layers of the MLP algorithm, and in the single hidden layer of the RBF algorithm has been found to be twenty and fifteen, respectively. The best MSE validation performance of MLP and RBF networks has been found to be 0.00286 and 0.00426 at 100 and 20 epochs, respectively.