Development of quick and efficient numerical tools is key to the design of industrial machines. While Computational Fluid Dynamics (CFD) techniques based on Navier Stokes (N-S) and Lattice Boltzman methods are becoming popular, predicting aeroacoustic behavior for complex geometries remains computationally intensive for design process and iteration. The goal of this paper is to evaluate application Navier-Stokes approach coupled with Ffowcs Williams and Hawkings (FW-H), and Broad-band Noise Model (BNS) to evaluate noise levels and predict design direction for industrial applications.Steady-state RANS based approaches are used to evaluate under-hood cooling performance and fan power demand. At each design iteration, noise levels and strength of noise source are evaluated using Gutin’s and broad-band noise models, respectively along with cooling performance. Each design feature selected for the final design has lower fan power and noise level with improved cooling. Acoustics simulation was run on baseline and final (optimized) design using transient LES technique to validate predictions calculated using steady-state approaches. Both LES (transient) and Broad-band noise models (steady-state) predicted that the final design will be quieter than the baseline design. It can be concluded from the study that steady-state RANS based approaches can be employed to evaluate and optimize acoustic and under-hood cooling system performance simultaneously. Transient LES technique can be used to predict realistic magnitudes of the noise levels at the end of the design cycle. The final design recommended in this study has significantly lower fan power, improved cooling efficiency, and lower noise levels. There is good agreement between experimental and simulation results (from broad-band noise) on industrial excavator.