This work presents the development of a medium-fidelity aerodynamic solver based on the Full Potential (FP) equation with an Immersed Boundary (IB) ghost-cell method, targeting aero-icing simulations. While body-conforming FP methods were used towards aero-icing in the early 2000, recent Euler-IB work showed the advantages of inviscid-viscous interaction schemes towards aero-icing applications: 1) Inviscid solvers have lower computational cost and better-conditioned meshes compared to Reynolds-Averaged Navier-Stokes (RANS) approaches; 2) The IB formulation eliminates the need for body-fitted meshes, enabling simulations over complex and evolving geometries—a critical capability for ice accretion problems where surface geometry changes continuously during the icing process. The Full Potential approach offers even greater computational speedups, reducing the number of flow variables by a factor 5 while retaining engineering accuracy at subsonic and low-transonic speeds. The methodology builds upon a finite-volume-based approach, inspired by the work of Lyu et al. and Neel, and is implemented into Polytechnique Montreal’s Chapel MultiPhysics Software CHAMPS. Validation is performed on the NACA0012 airfoil using Vassberg and Jameson’s body-fitted mesh series and a corresponding series of Cartesian IB meshes, for subsonic and transonic conditions with and without lift. The pressure coefficient distributions and aerodynamic coefficients show good agreement between IB Full Potential, body-fitted Full Potential, and body-fitted Euler solutions, with both Full Potential methods converging significantly faster than the Euler solver for subsonic cases typical in aero-icing simulations. Application to Case 241 of the 1st AIAA Ice Prediction Workshop demonstrates the solver’s capability to handle iced airfoil geometries, with results comparing favorably against Euler solutions and experimental data for both single-layer and multi-layer ice shapes.