Scale-resolving simulations of shock-induced turbulent boundary-layer separation remain a challenging problem for computational fluid dynamics due to the simultaneous presence of compressibility effects, unsteady shock motion, transition to turbulence, and large-scale separation. Classical experimental configurations, such as the Bachalo–Johnson bump, together with subsequent numerical and experimental revisions, have provided valuable reference data. However, these configurations often inherit constraints from their experimental origin that limit numerical reproducibility and systematic comparisons between different scale-resolving approaches. In this work, a new fully computational flow configuration for shock-induced turbulent boundary-layer separation is investigated using high-order scale-resolving simulations. The configuration has been designed within the framework of the Horizon Europe project ROSAS and is inspired by classical transonic bump cases, but redesigned from the outset for numerical simulations. It adopts a planar geometry, avoids artificial turbulence tripping, and promotes transition through a physically induced adverse pressure gradient generated by a small rounded step. The computational domain is sufficiently extended to avoid confinement and blockage effects typically associated with wind-tunnel walls, facilitating comparisons across numerical methods. The compressible Navier–Stokes equations are solved using a high-order Discontinuous Galerkin discretization in an implicit large-eddy simulation framework. Time integration relies on high-order linearly implicit Rosenbrock schemes with adaptive time stepping, allowing accurate and robust time advancement without Courant-type stability restrictions. The results focus on instantaneous flow features and selected flow quantities, including numerical schlieren visualizations, skin-friction distributions, velocity profiles, and Reynolds stress components at multiple streamwise locations. These results highlight the impact of spatial accuracy on the representation of shock–boundary layer interaction, separation, and reattachment, and provide a basis for the assessment of scale-resolving numerical methods.