Danuta Żyła

2025-01

Test lab measurement chain calibration automation — built Python scripts that retrieve calibration database records from the lab LIMS system, flag sensors approaching calibration expiry, and generate calibration certificates automatically.

Challenge: Lab LIMS system used a proprietary database format with no published API — accessing calibration records required reverse-engineering the LIMS database schema by observing database file changes during manual operations.

Learned: LIMS systems often have no programmatic API — database schema reverse-engineering is sometimes the only practical automation path. Document the reverse-engineered schema version-specifically, and test against a fresh LIMS install after any LIMS software update to detect schema changes before the automation silently breaks.

2024-06

MIL-STD-810H Method 514.8 vibration qualification for an engine-mounted sensor — developed test tailoring rationale based on actual measured vibration environment from engine test campaign, replacing the default Category 24 values with measured spectra.

Challenge: Measured vibration environment on engine nacelle had spectral peaks at blade passing harmonics that were 12 dB above the default MIL-STD-810H Category 24 spectrum. Using default Category 24 would have under-tested the sensor for the actual service environment — required tailored test level development.

Learned: MIL-STD-810H default vibration test levels are conservative in frequency range but can be non-conservative at specific spectral peaks driven by engine harmonics. Always develop tailored test spectra from measured data for engine-mounted equipment — default category levels are designed for ground vehicles, not turbine engine attachment points.

2023-10

Engine component vibration test fixture design — designed a vibration test fixture for a HPT disk test article to transmit shaker excitation to the disk bore without introducing spurious resonances in the 0-2000 Hz test bandwidth.

Challenge: The initial fixture had a first resonance at 820 Hz — within the test bandwidth. NASTRAN modal analysis of the fixture design predicted this resonance, but the initial design was based on a rapid CAD model that differed from the manufactured part. As-manufactured fixture resonance was 740 Hz — even closer to the test input peak.

Learned: Vibration test fixture resonances must be validated on the as-manufactured fixture before the test article is mounted. Manufacturing tolerances shift resonant frequencies by 5-15% from nominal — re-measure the empty fixture frequency response function after manufacture and before mounting the test article.

2023-03

DO-160G Section 8 vibration qualification test for a new avionics LRU — swept sine (5-500 Hz, 0.5g) and random vibration (12.9 Grms) testing in 3 axes on an electrodynamic shaker. Test performed to Equipment Category U.

Challenge: During random vibration in the z-axis, the LRU exhibited a resonance at 87 Hz with amplification factor Q = 28 — far higher than the assumed Q = 10 in the test specification. The amplified response at the critical PCB mounting point exceeded the circuit board deflection limit, causing solder joint cracking detected on post-test X-ray.

Learned: DO-160G random vibration tests must include a pre-test resonance survey to identify high-Q resonances before running the full endurance test. A high-Q resonance that amplifies the test input can cause damage during the endurance test that would not have occurred in service — the resonance survey catches this before the LRU is damaged.