Strain state approximation with Electrical Impedance Tomography using elastoresistive thin-film sensors

Electrical Impedance Tomography (EIT) is a method to reconstruct the conductivity or conductivity change inside an area of interest by systematic voltage measurements at the boundaries. Through the use of an elastoresistive thin-film as EIT object, it is possible to draw conclusions about the skin's spatial strain state. This might be utilized for Structural Health Monitoring (SHM) of thin-walled mechanical structures, which are typical for lightweight design. Thereby, the EIT in combination with an elastoresistive thin-film sensor that covers a mechanical structure allows the monitoring of larger areas. The detection and localization of sensor damages and strains is readily state-of-the-art. However, the identification of the 2-dimensional strain state is still an open topic. The present work proposes a model-based method to approximate a homogeneous 2-dimensional strain state by the evaluation of EIT measurements for, (i) electrode movements to find the the major principal strain directions, and (ii) quantitative conductivity changes to find equivalent strain values. The individual element values of a general strain tensor cannot be separated so far, however, for unidirectional strain states this novel evaluation method enables the full identification of the strain tensor. The proposed method is demonstrated and discussed by numerical forward simulations of strain induced sensor deformation and conductivity changes. Subsequently, these results are validated by an experimental study on flat tensile-loaded coupons that are equipped with differently oriented rectangular elastoresistive thin-film sensors each for EIT strain measurements. The proposed method estimates all experimentally tested strain orientations at highest accuracy. Estimation errors and limits of the proposed method are discussed.