Damage detection of truss structure based on the variation in axial stress and strain energy predicted from incomplete measurements

This study derives the static equilibrium equation of a damaged system on the basis of stiffness change due to damage as well as the constraint forces at measurements required for obtaining the measured data. Based on the derived equations, this work provides an analytical method to detect damage from the stress and strain energy variations between intact and damaged truss structures. The applicability of the proposed method is evaluated in detecting multiple damages of low rate in the truss structure from measured data contaminated by external noise. It is demonstrated that it is possible to properly detect damage in an isolated substructure by partitioning the damage-expected substructure from an entire structure and using the displacements measured at the boundary of the partitioned subsystem. The partitioning method has the benefits in reducing the computational time and measured data as well as improving the effectiveness of the damage detection process.