KnowMesh
← Back to search

Wioletta Urbańska

Turbine Cooling Engineer — Film Cooling · Thermal Engineering

Warsaw, Poland

Turbine Cooling Engineer specializing in film cooling effectiveness analysis and thermal barrier coating (TBC) thermal modeling. Uses 3D CFD to predict adiabatic wall temperature and film cooling coverage patterns on HPT blades and vanes, and develops coupled TBC thermal models for life prediction. Works closely with materials engineers to translate thermal predictions into coating specification requirements.

Expertise

  • film cooling effectiveness CFD and testing
  • thermal barrier coating (TBC) thermal analysis
  • adiabatic wall temperature prediction
  • cooling hole shape and arrangement optimization
  • TBC spallation and life prediction

Technologies

ANSYS CFX ANSYS Fluent MATLAB Python ParaView Linux HPC Git

Work History

2025-02

Film cooling AI surrogate model — provided 500 CFD training points (blowing ratio, hole pitch, injection angle) to ML team for lateral effectiveness surrogate. Validated surrogate accuracy and documented usage limits.

Challenge: Surrogate accuracy degraded at high blowing ratios (M > 1.5) where the film jet lifts off the surface — the physics changes qualitatively (attached vs. detached jet) and the surrogate could not interpolate across this discontinuity without explicit discontinuity treatment.

Learned: Film cooling surrogates must be built separately for attached and detached jet regimes. A single surrogate spanning both regimes will have poor accuracy near the liftoff transition — the physics change is too abrupt for smooth interpolation.

ANSYS Fluent Python pandas

2024-06

Film cooling hole shape optimization — compared cylindrical, fan-shaped (laidback fan), and console-shaped holes for effectiveness and aerodynamic penalty. Ran a 30-case CFD matrix at 3 blowing ratios.

Challenge: Fan-shaped holes produced the highest film effectiveness but also the highest aerodynamic penalty (higher total pressure loss coefficient). Required defining a merit function that traded cooling effectiveness against aerodynamic penalty — the weighting of the two objectives required input from the engine cycle team.

Learned: Film cooling hole shape selection is a multi-disciplinary decision — cooling effectiveness, aerodynamic loss, and manufacturing capability (EDM limitations on hole shape) must all be weighted. Involve the cycle team early to establish the effectiveness-vs-loss trade weighting before running the CFD matrix.

ANSYS CFX Python MATLAB ParaView

2023-11

TBC thermal analysis — computed temperature gradient across TBC layer (yttria-stabilized zirconia, 100µm thickness) and bond coat during T/O and relight thermal transients. Provided thermal inputs for TBC life prediction model.

Challenge: TBC spallation failure is driven by the bond coat oxidation rate, which depends on the time-temperature integral at the TBC/bond coat interface. Computing this over an engine cycle required coupling the steady-state CHT analysis to a transient thermal model that tracked the cyclic temperature history.

Learned: TBC life assessment requires cyclic thermal analysis, not just design-point steady-state temperatures. The hot-section thermal transients during acceleration and relight cycles contribute significantly to the oxidation-induced spallation life — do not size TBC thickness only on cruise steady-state conditions.

ANSYS Fluent Python MATLAB

2023-04

Film cooling effectiveness mapping for HPT stage 1 NGV using ANSYS Fluent — simulated coolant ejection from 42 shaped holes at varying blowing ratios. Predicted adiabatic effectiveness contours and laterally averaged effectiveness distribution.

Challenge: Shaped film cooling holes require a mesh that resolves the internal hole geometry and the external jet — total mesh count of 35 million cells per hole passage made a full-vane model computationally infeasible. Used a periodic sector model with 4 holes, validated against literature for full-vane extrapolation.

Learned: Full-vane film cooling CFD with resolved hole geometry is computationally impractical. Periodic sector models validated against literature data are the industry standard — document the validation carefully as DER reviewers will ask.

ANSYS Fluent Python ParaView Linux HPC