A global statistical model based approach for vibration response-only damage detection under various temperatures: A proof-of-concept study

•An approach for damage detection under various temperatures is presented.•The approach is based on functionally pooled statistical time series models.•It can operate even with vibration response signals only.•Its effectiveness is illustrated via laboratory experiments.•Comparisons with alternative methods are also presented.

The statistical damage detection in a composite beam structure operating under different temperatures is considered based on vibration response-only signals. For this purpose a novel stochastic global model approach is introduced based upon statistical hypothesis testing and identified Functionally Pooled models capable of describing the temperature-dependent dynamics. Two versions of the approach that use either modal or discrete-time model parameters are postulated. This is a proof-of-concept study in which the effectiveness of the approach is confirmed via laboratory experiments. Comparisons with alternative methods attempting removal of the temperature effects from the damage-sensitive features are also made.