By exploiting the idea of Hamiltonian paths and strand grids, an unstructured grid solver has been developed that is capable of using high-order stencil-based discretization and line-implicit operators along the loops and strands to solve compressible turbulent flow. The solver is equipped with domain partitioning using METIS and distributed compute capability using MPI. The mesh system is also extended to utilize overset meshes, and the overset interpolation is improved using linear-least squares at the boundary. The solver is used to study a rotor in hover and forward flight. Improvements have been made to the solver by way of a low-Mach preconditioning to improve the convergence rates and solution accuracy for low-Mach flows. Efforts have been carried out to analyze formally the order- of-accuracy of the schemes within the solver for the near-body unstructured grid and the background Cartesian grid. Preliminary results suggest that a finite-difference formulation of the reconstruction can achieve high-order for smoothly varying grids when compared to a finite volume approach. © 2017, American Institute of Aeronautics and Astronautics Inc. All rights reserved.