Assessment of sensor debonding failure in system identification of smart composite laminates

- State-space representation of a piezo-bonded composite laminate is identified from input-output data.
- The identified model is assessed for the case when the output is spurious due to partial sensor debonding.
- The state-space model identified with partially debonded sensor misapprehended the true dynamic behavior of the system.
- Principle component analysis successfully recovered both the presence and severity of partial sensor debonding.

System identification is an inverse algorithm of developing/improving mathematical representation of a physical system from input-output responses. This paper assessed a mathematical model of smart composite laminate identified with spurious output data due to sensor debonding failure. Improved layerwise theory and higher-order electric potential field were incorporated to develop electromechanically coupled finite element based mathematical model of the smart composite laminate with or without sensor debonding failure. The input-output data of the developed model were fed into the direct system identification algorithm to identify the state-space model of the smart structure. The developed theory was numerically implemented on a 16-layer cross-ply laminate with surface bonded piezoelectric sensor and actuator. Results showed that system identification algorithm misapprehended the true dynamic behavior of the smart structure in the presence of sensor debonding failure. In addition, principal component analysis was used to detect the presence and severity of partial sensor debonding in the identified state-space model.

230