Gas Turbine Design Toolkit

A preliminary single-spool turbojet design toolkit, built module by module to mirror the real industry design chronology: mission requirements → 0D cycle design → engine sizing → component mean-line design → blade geometry → CFD verification → off-design performance.

Given a mission (cruise Mach, altitude, required thrust) and a handful of design-point choices (pressure ratio, turbine inlet temperature, stage loadings, blade geometry parameters), the toolkit sizes a complete engine, designs its compressor and turbine stages, generates 3D blade geometry, runs cascade CFD, and reports off-design performance across the flight envelope — all from one editable input file.

Quickstart

git clone --recurse-submodules https://github.com/Sanjanakamboj/Propulsion.git
cd Propulsion
pip install -r requirements.txt
python3 design_engine.py

Blade Geometry Generator/parablade-master is a git submodule pointing to NAnand-TUD/parablade rather than vendored source — if you already cloned without --recurse-submodules, fetch it with:

git submodule update --init

Edit engine_inputs.txt (plain INI format, comments explain every field) and re-run design_engine.py to design a new engine. Each run writes to Results/<run_id>/:

• Cycle T-s/P-v diagrams and per-stage CSVs (Brayton Cycle Analysis)
• Compressor and turbine stage velocity-triangle/h-s/parameter-table diagrams (Compressor/Turbine Calculations)
• Turbine blade passage geometry, plus a full 3D blade (free-vortex twist across hub/mean/tip, exported as CSV + STL, with a 3D visualization) — compressor blade geometry is currently skipped, see Scope below
• Combustor equivalence-ratio gauge and parameter table (Combustor Analysis)
• Nozzle parameter table and thrust breakdown chart (Nozzle Analysis)
• A cascade-domain preview (and, if run_cfd = true, a full GMSH+SU2 CFD run)
• A PASS/FAIL sanity-check report

...and appends one row to design_log.xlsx next to this script (never overwritten). See Results/example_run for a complete sample of everything one run produces (including design_log.example.xlsx — the live design_log.xlsx itself isn't committed, since it accumulates with every local run).

Project layout

Each folder is a self-contained design stage; design_engine.py at the root wires them together into one pipeline.

Folder	Design stage
Brayton Cycle Analysis	Mission/ISA atmosphere, 0D single-spool turbojet cycle (Inlet→Compressor→Combustor→Turbine→Nozzle), engine sizing, T-s/P-v diagrams. Also includes standalone classic Brayton power-cycle variants (ideal, real, regenerative, intercooled, reheated).
Compressor Calculations	Axial compressor mean-line design: velocity triangles, stage loading, de Haller number, diffusion factor, loss models, surge/choke margins.
Turbine Calculations	Axial turbine mean-line design: velocity triangles solved to match required specific work, loss coefficients, stage efficiency, choking, blade cooling requirement.
Utils	Shared mean-line geometry, plotting, and sanity-check machinery used by both Compressor and Turbine Calculations.
Blade Geometry Generator	2D airfoil sections (ParaBlade-based) sized via Zweifel's criterion, extended to full 3D blade geometry: free-vortex radial twist, multi-span-station sections, 3D stacking with lean/sweep, CSV/STL export, 3D visualization.
CFD Analysis	3-blade cascade CFD of the turbine rotor blade: GMSH meshing, SU2 (RANS) automation, and an OpenFOAM (rhoSimpleFoam) automation path via Docker. Mesh quality metrics, residual/streamline/pressure-contour post-processing, CFD-vs-mean-line validation.
Combustor Analysis	Fuel-air ratio, equivalence ratio, pressure loss, residence time.
Nozzle Analysis	Converging and converging-diverging nozzle exit conditions, choking, thrust breakdown (momentum + pressure thrust).
Performance Analysis	Off-design performance sweep across the flight envelope (thrust lapse, TSFC, efficiencies vs. Mach/altitude) for a fixed engine design.
Materials, Data, Documentation, Validation	Scaffolded for future use; currently out of scope (see below).

Root files: design_engine.py (master orchestrator), engine_inputs.txt (user-editable design point), config.py (shared gas properties), requirements.txt.

Scope

This is a preliminary design toolkit — 0D/1D mean-line methods plus verification-level CFD, not a detailed/manufacturing-ready design tool. Deliberately out of scope, and why:

• Mechanical/structural design (rotor stress, disk sizing, FEA) and materials selection — not attempted; Materials/ is scaffolded but empty.
• STEP/CAD export — no CAD kernel (e.g. OpenCASCADE) is bundled or assumed available; blade geometry exports to CSV and STL instead.
• Emissions prediction — NOx/CO estimation needs empirical kinetics correlations that aren't included, to avoid presenting unverified numbers as if they were validated.
• Compressor/turbine performance maps and full off-design component matching — the off-design model here uses the standard fixed-non-dimensional-operating-point approximation (exact given that assumption), not real map data.

Compressor blade 2D section generation is currently skipped in the default pipeline: Zweifel's criterion (used for blade pitch sizing) is an accelerating-cascade (turbine) criterion and produces a self-intersecting section for the compressor's much smaller flow turning. Compressor Calculations/diffusion_factor.py implements the correct compressor-specific criterion (Lieblein diffusion factor) but isn't yet wired into the blade generation step.

Testing

Each folder's test suite is run independently:

cd "Brayton Cycle Analysis" && python3 -m pytest -q

503 tests pass across the toolkit as of this writing. Tests are grounded in either exact analytical relations (verified by hand-derivation) or real solver/tool output (e.g. an actual SU2/OpenFOAM CFD run, an actual GMSH mesh) rather than only mocked data, wherever that's practical.

Optional external tools

• SU2 (CFD Analysis/SU2.py) — set su2_binary in engine_inputs.txt's [cfd] section to your SU2_CFD binary path, and run_cfd = true to enable a full cascade CFD run (a few minutes).
• OpenFOAM (CFD Analysis/openfoam_runner.py) — drives an OpenFOAM installation running in a Docker container (tested against opencfd/openfoam-default) via docker exec; requires Docker running and that image available.

Neither is required for the default design_engine.py run — CFD is opt-in.