Adequate mechanical ventilation serves as the first and most important life-saving appliance during a tunnel fire. The most severe threat during a tunnel fire is the smoke. Most victims get incapacitated by the smoke, after which they decease from intoxication and/or suffocation. By means of mechanical ventilation, smoke-free escape routes via the central egress corridor are created. Two stages are considered in mechanical ventilation system design. The first stage involves the identification and positioning of jet fans to create a longitudinal ventilation system in the tunnel. The dangerous undesired reverse flow of smoke in the tunnel, back-layering, can be prevented by achieving a minimum critical air velocity. The second stage is a positive pressure establishment in the central egress corridor with respect to the tunnel. This is essential to prevent smoke flow from the tunnel into the central egress corridor through open escape doors. The tunnel ventilation performance is highly influenced by the position and heat release rate of the fire. Given the uncertainty concerning the fire, it is imperative that longitudinal ventilation is designed to consistently meet the back-layering constraint across all possible fire scenarios. For enhanced longitudinal ventilation reliability, jet fan placement at the tunnel entrance is preferable, while positive pressure ventilation benefits from a scattered jet fan layout.
A design approach based on the fundamentals of topology optimization is used to construct a systematic design method. To streamline the design of ventilation within the central egress corridor, optimizing longitudinal tunnel ventilation design while minimizing pressure downstream of the fire in the tunnel is advantageous. To prevent energy dissipation, a secondary objective specified as a penalization objective was introduced to promote the required distance between jet fans. This objective determines the placement of jet fans by considering the positioning of nearby jet fans within a specified distance, influenced by a penalization exponent. To address the fire related uncertainty, a scenario-based approach is applied. This method has the capacity to accommodate multiple fire scenarios simultaneously, where the designer can select the desired quantity.