PSI - Issue 37

Elizabeth K. Ervin et al. / Procedia Structural Integrity 37 (2022) 6–16 Ervin and Zeng / Structural Integrity Procedia 00 (2021) 000 – 000

12

7

Fig. 1. Overall procedure of optimized change detection methodology.

4. Numerical example

An ABAQUS ® model of a laboratory-sized two-story frame was built to test the proposed DI combination method. Creating multiple physical states, damage is simulated by reducing Y oung’s modulus of a structural member. The modal properties are then processed with the new GA-based methodology, and output damage detection results are evaluated. Figure 2 shows geometry and designations of the test frame comprising steel Unistrut ® . The frame is 4 feet by 4 feet (1.22 m × 1.22 m) in plane and 5 feet (1.52 m) in height. The footings are 0.5 feet (0.152 m) in height, and the heights for the first and second stories are 2.5 feet (0.762 m) and 2 feet (0.610 m), respectively. The frame members are categorized into beam and column based upon their longitudinal direction. Beam names are represented by circles while column names are represented by squares. Horizontal members are denoted by Beam 1 to 12, and vertical members are denoted by Column 13 to 16. The origin coordinate of the frame is set at the bottom of Column 13. Enough seeds are applied to each frame member to capture the structure’s high-order mode shapes, so 52 nodes are selected and numbered as output. Conditions must be set at the foundation and where beams meet columns. Inadequate constraints at boundaries