Skip to content

Digital Twin

Digital Twin + physical testing

A Digital Twin increases the value of physical testing by helping engineers predict expected behaviour, compare measured results, investigate deviations and improve the next engineering baseline.

Proposed approachNever a replacement for physical testing.
Close-up of APEX propulsion development hardware with flanged metal sections, fluid fittings, yellow connectors and cabling.
Fig. 03APEX propulsion development hardware: close-up of cabling and fluid fittings.

Physical ↔ digital

Build it.Model it.Test it.Correlate it.

Physical

APEX propulsion development hardware standing on a workshop floor, with tubing, cabling and a spherical vessel above a large conical section.

Build Instrument Test Measure

Engineering correlation

Digital

Model Predict Compare Update

Physical evidence builds confidence. Digital models help engineers learn more from every test.

The twin loop

Predict. Test. Compare. Learn.

The loop runs once per test. Each pass leaves a prediction, a measurement and a decision on record.

  1. 01PredictBefore a test, the model says what the hardware should do in this configuration.
  2. 02TestThe instrumented hardware is tested to a written objective.
  3. 03CompareMeasured data is set against the prediction, and every deviation is investigated.
  4. 04LearnThe model, the hardware or both are updated — and the next test starts better informed.

The loop in detail

  1. Engineering model
  2. Prediction
  3. Physical test
  4. Measured data
  5. Predicted vs measured
  6. Engineering review
  7. Model update or hardware update
  8. Retest

What the twin does: pre-test prediction · simulation · expected response · scenario exploration · test planning · predicted vs measured comparison · anomaly investigation · model updating · design feedback.

Predicted vs measured
Illustrative engineering data — not APEX test measurements
Predicted (model) Measured (test) Agreement band
0.00.51.00.000.250.500.751.00NORMALISED TIMEDeviation → engineering reviewPredictedMeasured
Show values as a table
Illustrative predicted and measured values, normalised
TimePredictedMeasured
0.000.000.00
0.100.510.44
0.200.760.73
0.300.880.85
0.400.940.94
0.500.971.07
0.600.991.01
0.700.991.00
0.801.000.99
0.901.000.99
1.001.001.01

Digital Twin maturity

Five potential levels

Confidence in a twin is earned the same way as confidence in hardware: by evidence, within defined limits.

  1. Level 1ModelAn engineering representation of the hardware.
  2. Level 2Simulated systemSubsystem interactions represented.
  3. Level 3Correlated modelCompared with test measurements.
  4. Level 4Validated operating regionUseful agreement demonstrated within defined limits.
  5. Level 5Operational engineering twinContinuously updated where appropriate.
APEX level: owner confirmation requiredNo maturity level is claimed for an APEX twin.

Configuration-aware twin

Same configuration, or no comparison.

Predicted-versus-measured comparisons are meaningful only when the digital and physical configurations correspond.

Engine configurationPhysical
APEX-ENG-001Revision C
  • Hardware baselinelinked
  • Firmware baselinelinked
  • Software baselinelinked
  • Sensor packagelinked
  • Calibration setlinked
  • Test configurationlinked
Corresponds ⇄
Twin modelDigital
DT-ENG-001Revision 08
  • Model geometry baselinelinked
  • Controller model versionlinked
  • Sensor model maplinked
  • Calibration importlinked
  • Prediction setlinked
  • Correlation reportlinked
Illustrative record formatIdentifiers are examples of structure, not APEX records.

Instrumentation

If you cannot measure it, you cannot correlate it.

  • 01Pressure
  • 02Temperature
  • 03Force / thrust
  • 04Vibration
  • 05Strain
  • 06Acceleration
  • 07Rotational speed
  • 08Position
  • 09Timing
  • 10Flow-related measurements
  • 11Controller states
  • 12Electrical measurements

Sensor to correlation — and who can contribute at each stage

  1. 01Sensor
  2. 02Signal conditioning
  3. 03DAQ
  4. 04Time synchronisation
  5. 05Test data
  6. 06Engineering analysis
  7. 07Digital Twin correlation

Engineering test data

Every test, a traceable record

Each physical test becomes a digital record tied to its configuration. Through the engineering collaboration, Gamyaa AeroSystems contributes digital-engineering, test-data, model-correlation and engineering-evidence capabilities.

Identity
  • Test ID
  • Test objective
  • Procedure
Configuration
  • Hardware configuration
  • Serial numbers
  • Software version
  • Firmware version
  • Sensor configuration
  • Calibration references
  • Model revision
Results
  • Recorded data
  • Events
  • Observations
  • Anomalies
Decision
  • Analysis
  • Disposition
  • Approval
Record structureFields only. No APEX test data is shown.