Computational investigation of immersion quenching : understanding of boiling conjugate heat transfer

Error on title page – year of award is 2022.Quenching is a metallurgical procedure used to alter material properties of metals and alloys. Although used extensively throughout human history, its design relies heavily on experience and often the trial and error method. Therefore, the availability of computational tools capable of describing the physics during quenching would help design facilities efficiently, lower costs, and even make the process more environmentally friendly. This work aims to develop and validate a numerical procedure for immersion quenching using computational fluid dynamics. The methodology employs the partitioned approach. An energy equation and Eulerian two-fluid model describe the solid and fluid region, respectively. The heat transfer information is exchanged at the regions' interface. In this thesis, various aspects of the methodology are discussed. Yet, the primary attention is given to the wall boiling and the conjugate heat transfer, which are crucial. Furthermore, during an attempt to formulate stability criteria, the numerical Biot number has been developed as a potential candidate accounting for all three phases, solid, liquid and vapour. The code validation is split into three chapters. The first investigates conjugate heat transfer without boiling in a backward-facing step geometry. The next is concentrated on a hot thin horizontal plate submerged in water, and the last describes the quenching system behaviour during immersion quenching of a cylinder in a vertical orientation. The numerical results proved excellent accuracy for conjugate heat transfer problem without boiling. A good agreement with validation data is also achieved when boiling occurs, yet complications are observed at locations where vapour movement is obstructed. Vapour volume fraction tends to be mesh sensitive, affecting the solid temperature _eld in its proximity. Such behaviour can be often prevented using phase change within the inner fluid mesh. However, doing so, the user limits the solver capabilities as vapour obstruction is often desirable.Quenching is a metallurgical procedure used to alter material properties of metals and alloys. Although used extensively throughout human history, its design relies heavily on experience and often the trial and error method. Therefore, the availability of computational tools capable of describing the physics during quenching would help design facilities efficiently, lower costs, and even make the process more environmentally friendly. This work aims to develop and validate a numerical procedure for immersion quenching using computational fluid dynamics. The methodology employs the partitioned approach. An energy equation and Eulerian two-fluid model describe the solid and fluid region, respectively. The heat transfer information is exchanged at the regions' interface. In this thesis, various aspects of the methodology are discussed. Yet, the primary attention is given to the wall boiling and the conjugate heat transfer, which are crucial. Furthermore, during an attempt to formulate stability criteria, the numerical Biot number has been developed as a potential candidate accounting for all three phases, solid, liquid and vapour. The code validation is split into three chapters. The first investigates conjugate heat transfer without boiling in a backward-facing step geometry. The next is concentrated on a hot thin horizontal plate submerged in water, and the last describes the quenching system behaviour during immersion quenching of a cylinder in a vertical orientation. The numerical results proved excellent accuracy for conjugate heat transfer problem without boiling. A good agreement with validation data is also achieved when boiling occurs, yet complications are observed at locations where vapour movement is obstructed. Vapour volume fraction tends to be mesh sensitive, affecting the solid temperature _eld in its proximity. Such behaviour can be often prevented using phase change within the inner fluid mesh. However, doing so, the user limits the solver capabilities as vapour obstruction is often desirable