Wake-resolved 3D CFD analysis of a J-blade vertical axis wind turbine: startup optimization and tip-speed trade-offs

• Three-dimensional CFD of a J-blade H-Darrieus rotor with URANS SST k–omega. • Gromeka acceleration links vorticity-velocity to turbine performance. • Enhanced self-starting at λ = 0.2 from intensified momentum transport. • At λ≈1.6, cavity lowers Cp yet accelerates wake recovery for arrays. • Agreement with experimental benchmarks; RMSE ≤ 0.045 for C m and C P . This study presents a novel three-dimensional computational fluid dynamics (CFD) investigation of an H-type vertical-axis wind turbine (VAWT) incorporating a trailing-edge cavity “J-blade” modification. The aim is to capture spanwise flow phenomena, vortex evolution, and wake recovery mechanisms that are inaccessible to two-dimensional models. Unsteady Reynolds-averaged Navier-Stokes equations (URANS), closed with the shear stress transport (SST) k -ω turbulence model, are solved on hybrid meshes using sliding mesh interfaces and azimuthal time-stepping. Performance metrics, including cycle-averaged moment and power coefficients, are validated against experimental benchmarks across tip speed ratios ranging from 0.2 to 1.6, with root mean square errors of 0.0318 and 0.0454 for C m and C P , respectively. Flow dynamics are further analyzed using Gromeka acceleration fields, turbulence intensity, and turbulent kinetic energy distributions. Results show that at λ = 0.2, the J-blade significantly improves self-starting by intensifying near-blade angular momentum transport and confining turbulence to the near wake. At higher λ, however, the cavity induces premature separation and reduces tip-vortex coherence, leading to diminished efficiency but faster wake recovery. This work offers the first integration of Gromeka-based diagnostics in a fully 3D framework for VAWTs, providing mechanistic insight into the aerodynamic trade-offs introduced by cavity-based blade modifications.