3. CORA Scorecard Configuration for a Single Load Case

Scope of the Scorecard

A CORA scorecard is not a single number generator; it is a structured evaluation object that binds one test–CAE signal pair to a defined set of metrics, corridors, and weights, and then reduces those metric results into a weighted overall score. For a single load case, the scorecard must be configured so that every channel included in the evaluation is justified by the physical quantity it represents, and so that the weighting reflects the engineering importance of that quantity for the load case under review. The configuration decisions made here directly determine whether the final score is a meaningful pass/fail indicator or an arbitrary average.

This case configures a scorecard for one load case: a frontal 40 km/h full-overlap barrier impact, with test and CAE acceleration and intrusion channels already ingested and preprocessed onto a common time base. The task is to select evaluation channels, define corridor parameters, assign weights, compute the weighted overall score, and interpret it against a defined acceptance threshold.

Selecting Evaluation Channels

Channel selection should follow the load case's physical story rather than the convenience of available data. For a frontal impact, the primary quantities are occupant-compartment deceleration, structural intrusion, and restraint-relevant accelerations. A practical selection for this load case is:

Each channel must have consistent units, sign convention, and sampling rate between test and CAE before it enters the scorecard. Channels that are dominated by measurement noise, or that have no physically meaningful CAE counterpart, should be excluded rather than down-weighted, because a low-weight noisy channel still injects variance into the overall score.

Corridor Definitions

CORA evaluates magnitude and phase agreement through corridors built around the test signal. The corridor width is defined as a fraction of the signal's peak magnitude, and the phase corridor is defined as a time or phase tolerance band. For this load case, a defensible starting configuration is:

• Magnitude corridor: ±15% of the test peak for acceleration channels, ±10% for intrusion channels, since intrusion is typically held to tighter structural tolerance.
• Phase corridor: ±5 ms for acceleration channels, ±8 ms for intrusion channels, reflecting the slower structural response of deformation channels.
• Evaluation window: 0–120 ms, covering the impact pulse and the early structural rebound, excluding post-event settling where both signals are near zero and corridor metrics become numerically unstable.

The corridor parameters are not universal constants; they are engineering choices that must be documented alongside the score so that a later reviewer can reproduce or challenge them.

Weighting Scheme

Weights encode engineering priority. For a frontal impact, structural deceleration and occupant-compartment intrusion carry the most consequence, while seat track acceleration is a secondary input. A transparent weighting for the four channels above is:

Weights must sum to 1.0. If a channel is removed after weighting, the remaining weights must be renormalized, and the renormalization must be recorded, because a silently renormalized scorecard is not comparable to the original.

Computing the Overall Score

Each channel produces a combined CORA score in the range [0, 1], where 1 indicates perfect agreement. The overall score is the weighted sum:

S = Σ (w_i · c_i),  with Σ w_i = 1.0

Using the illustrative channel scores in the figure, the arithmetic is straightforward and should be recorded step by step so that any reviewer can audit the aggregation:

c = [0.86, 0.74, 0.91, 0.82]
w = [0.35, 0.15, 0.25, 0.25]

S = 0.35*0.86 + 0.15*0.74 + 0.25*0.91 + 0.25*0.82
  = 0.3010 + 0.1110 + 0.2275 + 0.2050
  = 0.8445

The channel scores above are illustrative values chosen to demonstrate the aggregation; they are not measured results. In an actual run, each c_i comes from the corridor evaluation of that channel, and the scorecard should store both the per-channel scores and the final weighted value.

Interpreting Against the Acceptance Threshold

With an acceptance threshold of S ≥ 0.80, the example score of 0.845 passes. Interpretation should not stop at the pass/fail verdict. The seat track channel at 0.74 is the weakest contributor and, although its weight is only 0.15, it signals that the occupant-environment acceleration is the least well-correlated quantity in this load case. A reviewer should inspect whether that channel's phase corridor is being violated by a systematic lag, which would point to a timing or stiffness mismatch rather than a magnitude error.

Configuration Checklist

• Every included channel has matching units, sign convention, and sampling rate between test and CAE.
• Corridor widths and phase tolerances are documented per channel and justified by the physical quantity.
• Weights sum to 1.0; any renormalization after channel removal is recorded.
• The evaluation window excludes near-zero settling regions where corridor metrics are unstable.
• The acceptance threshold is stated explicitly and applied consistently across load cases.
• Per-channel scores are retained alongside the weighted overall score for auditability.