In this paper, two efficient coupling approaches for the prediction of multi-phase flow and aeroacoustics are presented. The first is formulated for the spray injection into combustion chambers, based on an Eulerian-Lagrangian two-way coupling method. In this case, the spray particles are tracked by a Lagrangian model, which is coupled to the flow field prediction. Thus, volume data of the velocity field and source terms must be exchanged between the Eulerian and Lagrangian solution schemes. To avoid waiting times during the parallel execution of the solvers, a novel non-blocking interleaved time step execution pattern is implemented for efficient temporal coupling [1]. Adaptive mesh refinement is used to efficiently resolve regions with high flow gradients and large spray particle density. The second approach is related to the prediction of aeroacoustic noise, for which a coupled CFD/CAA method [2] is used. The flow field is predicted by a finite-volume solver for the Navier-Stokes equations, while the acoustic wave generation and propagation are computed by a discontinuous Galerkin method, which solves the acoustic perturbation equations. The acoustic sources of the turbulent flow, i.e., the fluctuating Lamb vector, is transferred at each time step in a predefined source region. Both solvers are based on a hierarchical Cartesian mesh, which share a joint coarse mesh level. The partitioning of the domain is performed on the coarse level, which allows to exchange the acoustics sources without communication and to use efficient dynamic load balancing. The two coupling approaches are described in detail and results are presented for the spray injection into an internal combustion engine and the noise prediction for chevron nozzle jets.