Multi-Fidelity SBLI Prediction Simulation - roboflank

Blog
Projects
About

Skip to main content

Multi-Fidelity SBLI Prediction Simulation

Scramjet Intake Design via Learned Compressible Flow Surrogates

Multi-Fidelity SBLI Prediction Simulation

On this page

Key Claim 01 — The Problem: SBLI Prediction at Design Scale 02 — Technical Approach: Discontinuity-Aware Operator Learning 2.1 The Neural Operator Framework 2.2 The RiemannONet Approach 2.3 Extension to Scramjet SBLI (This Work)03 — The Physics: Oblique Shock Theory & SBLI 3.1 θ-β-M Relation 3.2 Rankine-Hugoniot Jump Conditions 3.3 SBLI Separation Prediction 04 — Multi-Fidelity Data Pipeline 4.1 Fidelity Hierarchy 4.2 CST Geometry Parameterisation 4.3 Data Generation Pipeline 05 — Training Strategy 5.1 Multi-Fidelity Transfer 5.2 Physics-Informed Loss 06 — What This Predicts (Outputs)07 — Novelty & Gap Analysis 08 — Technical Stack 09 — Roadmap References

Home Blog About

Twitter Instagram GitHub LinkedIn

© 2026 roboflank

Gradient background

Gradient background