Computational Fluid Dynamics (CFD) Portfolio

Shock-Wave / Boundary-Layer
Interaction at Compression Ramps

A supersonic CFD study of shock-induced flow separation at a 24° compression ramp, validated against high-speed schlieren experiments and resolved with a density-based, convergence-stabilised ANSYS Fluent workflow.

Mach Number2.0Supersonic Regime
Ramp Angle24°Compression Corner
ValidationStrongAgreement with practical experiment
Key PhenomenonSeparationShock-induced loop backflow
01

Aerodynamic Context & Experimental Baseline

Context & Function

Shock-wave/boundary-layer interactions (SWBLI) at compression ramps are a critical phenomenon in supersonic aerodynamics, governing the performance of supersonic engine inlets and control surfaces. The impinging shock creates severe adverse pressure gradients that force the boundary layer to separate, producing aerodynamic drag, flow unsteadiness, and potentially catastrophic inlet unstart.

Research Baseline — Validation Benchmark

Shock-Wave/Boundary-Layer Interactions at Compression Ramps Studied by High-Speed Schlieren

Sun, Z., Gan, T., & Wu, Y. — AIAA Journal, Vol. 58(4), 2020

The CFD model was validated against this benchmark study, which uses 20 kHz high-speed schlieren to capture the unsteady separation shock dynamics across two Reynolds number regimes at Mach 2. The AIAA Journal is a peer-reviewed scientific journal.

View DOI
Baseline Parameters
Freestream Velocity U∞
514m/s
Reynolds Numbers Reδ
18.6 / 35.6×10³
Flow Diagnostics
20 kHzSchlieren
Ramp Height h
20mm
02

Computational Domain & Meshing Strategy

Platform — ANSYS SpaceClaim

The 2D domain was constructed in ANSYS SpaceClaim, defining an upstream flat-plate length sufficient for a fully developed turbulent boundary layer to form before the compression ramp corner, with a downstream extent large enough to capture the reattachment shock and exhaust without reflective boundary contamination.

SpaceClaim Geometry: Domain & Dimensions
ANSYS SPACECLAIM
SpaceClaim CAE domain geometry with 24° compression ramp dimensions
Meshing with biased edge sizing
ANSYS MESHING
CFD mesh with biased edge sizing concentrating elements along the compression ramp
Edge Sizing & Refinement

Edge sizing was used to heavily concentrate elements around the ramp corner and the anticipated shock-foot locations, resolving the sharp gradients of the oblique shock and the separation region without smear.

Boundary Layer Resolution

A biased inflation layer resolved the viscous sublayer to a target wall y+ ≈ 1, accurately capturing flow separation, skin friction and shear stresses within the separated bubble.

03

ANSYS Fluent Solver Methodology

Solver Setup

A density-based solver was used. This is mandatory for capturing the highly compressible, supersonic shock structures that dominate this flow. The coupled pressure-velocity formulation resolves the strong compressibility coupling that a pressure-based solver would smear across the shock front.

Convergence Strategy — 1st to 2nd Order

The simulation was initially run using a 1st-order upwind spatial discretisation scheme to establish the basic supersonic flow field and prevent early divergence from under-resolved shock gradients.

Once the residuals steadied, the solver was transitioned to a 2nd-order upwind scheme to crisply resolve the shock waves and minimise numerical dissipation, recovering the sharp density jumps observed in the schlieren.

04

Kinematics of the Separation Bubble

Pathline Analysis

When the adverse pressure gradient driven by the shock propagates upstream through the subsonic portion of the boundary layer, it forces the flow to separate and curl back on itself, forming a closed recirculation bubble ahead of the ramp corner. This is the cause of the interesting phenomenon.

Full Domain Pathlines
PATHLINES · MP4

Full domain streamlines interacting with the compression ramp.

Close-Up Pathlines
PATHLINES · MP4

Close-up of the shock-induced separation bubble and loop backflow.

05

Transient Dynamics & Shock Oscillation

Low-Frequency Unsteadiness

The SWBLI is inherently transient. The separation shock exhibits low-frequency broadband oscillations driven by the expansion and contraction of the separation bubble, a phenomenon documented in the Sun et al. study and reproduced in the present CFD. However, when running a transient simulation in Ansys, the turbulent eddies couldn't dissipate energy outwards in the Z direction so the separation bubble inflated instead of oscillated. To simulate the oscillation, the research paper mentions Direct Numerical Simulation (DNS) and Large-Eddy Simulation (LES) would be required. While these methods would improve the solution, they were too computationally expensive for me currently..

Transient Mach Contour
TRANSIENT · MP4

Transient Mach contour evolution.

Shock Oscillation Tracking
TRANSIENT · MP4

Unsteady shock-wave growth tracking.

06

Experimental Validation & Reconciliation

Experimental 20 kHz Schlieren
Experimental schlieren snapshot of the 24° ramp shock-wave/boundary-layer interaction
CFD Numerical Density-Gradient
CFD rescaled streamwise schlieren density-gradient contour over the 24° ramp
Numerical vs Experimental Schlieren

The CFD density-gradient field was compared directly against the 20 kHz schlieren snapshots from the Sun et al. paper. The separation shock angle, separation bubble length, and reattachment shock configuration all matched the experimental record closely.

Validation Metrics
CFD Separation Shock Angle
47°
Experimental Shock Angle
47°
Difference between CFD and experimental
0°

Exact agreement

07

Future CFD Extensions

The current RANS baseline establishes a validated foundation. The following iterative steps extend the study toward higher-fidelity turbulence modelling and active flow control.

High-Fidelity Turbulence

Large-Eddy Simulation (LES)

Transitioning from RANS to LES to capture the highly unsteady, three-dimensional turbulent structures inside the separated shear layer and resolve the broadband shock motion directly.

Parametric Study

Reynolds Number Sweep

Running parametric sweeps to match both Re₁ (18,600) and Re₂ (35,600) from the literature to study interaction length scaling and bubble growth with Reynolds number.

Flow Control

Thermal & Separation Mitigation

Introducing active flow control (micro-vortex generators or bleed slots) to minimise the separation bubble size and reduce shock-induced drag and unsteadiness.

William Colohan
About the Engineer

A Sheffield Aerospace Engineering student with extensive CAD, FEA, CFD and MATLAB experience: turning aerospace challenges into validated, optimised designs.

References
  1. [1]

    Sun, Z., Gan, T., and Wu, Y., 'Shock-Wave/Boundary-Layer Interactions at Compression Ramps Studied by High-Speed Schlieren,' AIAA Journal, Vol. 58, No. 4, 2020. https://doi.org/10.2514/1.J058257

SWBLI at Compression Ramps

Supersonic CFD Validation Portfolio

ANSYS SpaceClaim · ANSYS Fluent · Density-Based Solver · Schlieren Validation