In HVAC ventilation systems, noise generated by built-in fans represents a dominant source of sound emission. The presence of heat exchangers upstream or downstream of the fan induces disturbed flow fields, which further increase fan noise through enhanced turbulence and non-uniform velocity distributions. Since heat exchangers are among the primary contributors to aerodynamic sound radiation in ventilation systems, a detailed investigation of their influence on fan inflow conditions is essential for noise reduction strategies. However, the wide variety of heat exchanger geometries makes systematic experimental and numerical investigations challenging. To address this issue, heat exchangers are replaced by an active turbulence grid (ATG), which enables controlled generation of turbulence by individually actuating multiple turbulence-generating elements. This approach allows systematic variation of turbulence intensity, length scales, and velocity distributions at high Reynolds numbers. Previous experimental studies have demonstrated that active grids can successfully replicate flow fields downstream of heat exchangers. The present study extends this work by investigating whether the complex flow generated by an active turbulence grid can also be accurately predicted using large-eddy simulation (LES). Due to the high computational cost associated with simulating the full active grid geometry—consisting of 10 horizontal and 10 vertical rotating rods—the simulation procedure is divided into two phases. First, a reduced unit-cell grid with a smaller cross-section is simulated. The resulting turbulence characteristics, including mean velocity, turbulence intensity, and turbulent length scales, are extracted and remapped onto a larger computational domain using an anisotropic linear forcing method. Stationary grid configurations (fully open, half-open, and fully closed) are simulated and validated against experimental measurements, showing deviations in turbulence intensity between 2% and 11%. In the second phase, rotational modes of the active grid are simulated using overset mesh interfaces to account for rotating turbulence generators. Rotational speeds ranging from 50 to 800 rpm are analyzed and compared with experiments. After mapping the unit-cell results to the full grid, deviations in turbulence intensity between 8% and 20% are observed. Nevertheless, the simulations successfully reproduce the spatial distribution of turbulence quantities and capture the