Deep learning methods for solving partial differential equations (PDEs) have gained significant attention due to their ability to accelerate simulations. However, many existing studies utilize datasets with simplified setups, which differ considerably from the complexities encountered in real-world simulations. In this work, we evaluate the performance of the Fourier Neural Operator (FNO), a deep learning-based method, in addressing the Reynolds-Averaged Navier-Stokes (RANS) equations for airfoil simulations. Our study employs a more complex dataset that incor- porates a broader range of variables across different simulations. We adapt this dataset to suit the FNO architecture and implement optimizations to enhance prediction accuracy.izations to enhance prediction accuracy.