Marcin Fijałkowski
Engine Performance Engineer · Propulsion Engineering
Warsaw, Poland
Engine Performance Engineer responsible for linking propulsion cycle models to aircraft-level performance analysis. Builds and validates engine performance decks used by aircraft performance teams for payload-range, fuel burn certification, and mission analysis. Has deep experience in post-test data reduction and model-to-test correlation at sea level and altitude test conditions.
Expertise
- engine performance deck modeling
- aircraft-engine performance integration
- fuel burn and payload-range analysis
- engine test data reduction
- performance model validation
Technologies
Work History
2025-02
Developed Python-based performance deck automation pipeline — generates, validates, and delivers engine performance tables to aircraft performance team on demand. Replaced a semi-manual process that took 3 days per deck revision.
Challenge: Automating the GSP deck generation required parsing GSP output files (custom format) and handling failed GSP runs gracefully — some operating points at extreme altitude/Mach fail to converge and must be flagged and interpolated.
Learned: Performance deck automation must handle solver failures gracefully. A pipeline that crashes on any failed operating point is not production-ready — failed points should be flagged, reported, and filled via interpolation with documented uncertainty.
2024-06
Model-to-test correlation for the new turbofan development engine. Correlated GSP performance model predictions against 150 test points across 4 power settings and 3 ambient conditions. Updated component maps for HPC and HPT.
Challenge: HPT component map required significant reshaping to match test data — the map from design CFD was over-optimistic by 1.2% in total-to-total efficiency at the design point. Correcting the map without distorting off-design behavior required constrained map scaling.
Learned: CFD-derived component maps should never be used as-is in system performance models — they need test-based correction. Build in a map correction step from the start of the test campaign rather than trying to retrofit it at the end.
2023-11
Sea level test data reduction for engine acceptance testing — processed raw thermocouple, pressure, and flow measurements to compute corrected thrust, SFC, and component performance from 200+ test runs.
Challenge: Measurement uncertainty propagation through the data reduction chain was not documented. Built a formal uncertainty budget following GUM (Guide to the Expression of Uncertainty in Measurement) methodology — revealed that SFC measurement uncertainty was ±0.3%, larger than the expected improvement we were trying to measure.
Learned: Engine test data reduction must include formal uncertainty analysis. Without a GUM uncertainty budget, it is easy to draw conclusions from data that are within the measurement noise floor — a costly mistake in a development program.
2023-05
Payload-range performance analysis for two engine thrust rating variants — quantified the fuel burn penalty and MTOW impact. Used GSP engine deck integrated with Piano-X aircraft performance model.
Challenge: Piano-X uses tabular engine decks with specific formatting requirements. Generating compliant deck tables from GSP output required a custom Python converter that handled unit conversions, table interpolation limits, and extrapolation guards.
Learned: Engine deck format converters are brittle if written ad-hoc. Build a proper converter with explicit unit handling and format validation from the start — format mismatches propagate silently into aircraft performance errors that are hard to detect.