Spatial differences in how people live and move are key to understanding how flu spreads, but many traditional models ignore these details, making it harder to predict real-world outbreaks. In this study, we present a new three-dimensional (3D) model that treats the spread of flu like a flowing process. We connect groups of susceptible people, infected people, recovered people, and virus-carrying particles through a set of equations that describe how they move and interact over space and time. This movement is driven by “pressure” from high-density areas pushing toward lower-density areas, creating what we call an “epidemic flow.” We show mathematically that the model is stable and reliable to work with. To solve the equations, we use the finite volume method with an upwind flux approximation for spatial discretization and Strang operator splitting for time integration. Tests confirm the accuracy of our approach. Our simulations show how an outbreak can grow from a small hotspot into a widespread event, with shifting patterns of movement responding to changes in infection levels. Under the parameter settings considered in our simulations, close-range person-to-person transmission contributed more to infection spread than aerosol-mediated transmission. Overall, this study shows how a 3D modeling approach can help capture the complexity of flu transmission and support more targeted public health responses.