Numerical simulation of Deflagration, Detonation and DDT in large scale inhomogeneous reactive mixtures

Demand for energy and petrochemical products has been on the raise in recent decades leading to huge production, transportation and storage infrastructures all around the world. The industry needs to store and transport large volumes of fuels and petrochemicals which are highly reactive and normally in liquefied form, such as Liquefied Natural Gas (LNG). However an accidental release of these fuels into air can occur either due to human error or machine failure and would result in chemicals evaporation, dispersion and formation of highly reactive vapour clouds. These vapour clouds, under the right conditions, can ignite and generate devastating explosions. Depending on the ignition size and strength as well as the fuel concentration in the cloud, an initial flame or explosion could be generated. The ignition, if strong enough, can result in direct initiation of Detonation waves. In contrast, a mild ignition may generate an initial flame or deflagration wave in the cloud. Several numerical and experimental studies have shown that, under the right condition, a flame can accelerate and eventually undergo transition to Detonation. Compared to the flames, Detonation waves can cause a more substantial damage (due to the pressure impulse) in an extremely short time which leaves practically no chance of containment, quenching or mitigation, therefore it is vital to take every measure to stop a flame from acceleration and reaching a condition where it can transform to a detonation wave. Containing the Flame Acceleration process (FA) and stopping the Deflagration to Detonation Transition (DDT) can save lives and millions of pounds. Therefore FA and DDT are major hazards which influence the safety standards as well as risk assessment and management in the petrochemical and energy industries. Currently there is no reliable modelling approach to predict formation DDT; therefore the only solution is to fully resolve them numerically. Meaning that; the existing numerical approaches, which are based on coarse gird in combination with traditional flame and turbulence models, are of no real value simply because the DDT mechanism is ignored and wiped out by averaging the key parameters through use the of traditional models on coarse grids. This project intends to develop a practical and reliable modelling approach for simulating DDT in large scales through an in depth analysis of the DDT mechanism with the view of obtaining more insight into possible alternatives for describing the transition mechanism through a model instead of fully resolving it. It is also intended to run validation tests to verify the performance of the model.

CFD code development (preferably using OpenFOAM), High Performance Computing, new model development and validations are essential parts of this studentship.