Development and validation of a multi-strand solver for complex aerodynamic flows. The strand grid approach is a flow solution method where a prismatic-like grid using “strands” is grown to a short distance from the body surface to capture the viscous boundary layer and the rest of the domain is covered using an adaptive Cartesian grid. The approach offers several advantages in terms of nearly automatic grid generation and adaptation, ability to implement fast and efficient flow solvers that use structured data in both the strand and Cartesian grids, and the development of efficient and highly scalable domain connectivity algorithm. An improvement to this approach is the multi-strand strategy, where multiple strands are allowed from each surface vertex to enhance grid resolution near sharp corners. This paper introduces a fully parallel and highly efficient flow solver called mStrand that is developed from ground-up to operate on multi-strand meshes. The strand solver is integrated to HPCMP CREATE$^{TM}$-AV Helios framework to simulate complex aerodynamic flows. Detailed validation of the solver is shown on problems with varying degrees of complexity and comparison with experimental data. A performance study shows that the strand solver is nearly as efficient as a structured grid solver.