Melih YıldızTürkçe

Conference talk · 16th EASN International Conference · 27–30 October 2026

Experimental Validation of an Aircraft Digital Twin: From Concept to Validation

Ayşe Nur Dişlitaş, Melih Yıldız (presenter)

Department of Aeronautical Engineering, Faculty of Aeronautics and Astronautics, Erciyes University, Kayseri, Türkiye · ISAE-SUPAERO, Toulouse, France

✓ Presented: 29 October 2026, 18:20 · room 05-114 · Digital Engineering for Aerospace Systems IV session · by Melih Yıldız

Before building a test, check that it could actually prove your model wrong.

One-pager (PDF, A4)One-pager (HTML)Full paper: forthcoming

Summary

Digital twins are usually validated by whatever data can be found. A comparison validates nothing, however, if the correction applied to the measurement is larger than the model error it is meant to reveal.

We express this as a validation-resolution ratio, Rv = u / (δ · y): the uncertainty of the corrected measurement divided by the model discrepancy the test must detect. A comparison is informative only if Rv < 1. We apply it to the three evidence sources available for an ATR 72-500 digital twin built in Simcenter Amesim that claims 5–10 % accuracy.

Horizontal bar chart of the validation-resolution ratio on a log scale. HAPS wing section 0.33, L1 component data 0.40, L2 and L3 0.67, L4 ADS-B operation 1.07, and the 1:113 ATR 72 half-model tunnel test 16.6, far beyond the R_v = 1 line.
The validation ladder: each evidence source placed against the Rv = 1 line. The acceptance band at each level is what Rv ≈ 1 gives for the best source available there.

Three findings

  1. The tunnel campaign was cancelled after a one-afternoon audit. The largest model that fits gives Re = 4.0×10⁴ against 1.4×10⁷ in cruise (×348). The scale correction is 122 % of the drag coefficient itself, giving Rv = 16.6, and still 1.96 under an optimistic flat 10 % correction error.
  2. Operational data change the question. Across 428 ADS-B profiles the median stage is 371 km, while the design mission is 750 km, a distance only 6 % of flights reach. The unknown mass matters less than expected: ±21.5 % in take-off mass gives only ±3.4 % in block fuel.
  3. Match the facility to the aircraft class. The same tunnel reproduces a 20 km HAPS wing's Reynolds number (1.9×10⁵) exactly with a 97 mm chord, giving Rv = 0.33.

Three evidence sources, one test

SourceKey numberRv
ATR 72 half model, 1:113, 400×400 mm tunnelRe gap ×348; correction 122 % of CD16.6
ADS-B (OpenSky), 428 complete ATR 72 flightsresolves ≈ 8 %; mass → fuel attenuated ×6.41.07
HAPS wing section, 97 mm chord, same tunnelRe matched exactly; M 0.087 vs 0.100.33

The ratio is not tied to tunnels: flight tests, one-count drag increments and loads models are audited in the same way. When δ is small the criterion is harder to meet.

Full paper: Engineering Proceedings (MDPI), EASN 2026 volume · forthcoming

Companion talk at this conference: Machine Learning-Assisted Multidisciplinary Design Optimization in Aircraft Conceptual Design Using Digital Twin Technology (Elif Zeyneb Tekin, Melih Yıldız)

All talks at this conference →