Damage in a structure can be defined as a reduction in the structure's load bearing capacity, which can result from deterioration of the structure's components and connections during its service life. The objectives of the proposed research are to develop and experimentally validate: 1) a robust, iterative algorithm that can accurately detect the locations and extent of small to large levels of damage using a minimum number of vibration measurements; and 2) novel stochastic models for the random impact test in modal testing and a novel random impact device. The new iterative method that combines a general-order, multiple-parameter perturbation method and an optimization method can accurately detect the locations and extent of damage in beam structures such as lighting masts using only the first few measured natural frequencies. Other new features include treatment of ill-conditioned system equations, enrichment of the measurement information by modifying the structure in a controlled manner, development of physically-based joint modeling techniques, and joint damage detection. The random impact test can combine the advantages of the single-impact test and the shaker test; it can increase the energy input to the structure, and improve the signal-to-noise ratio and the measured frequency response function. With the stochastic models for the random impact series the input energy to the structure can be distributed in a wide or desired frequency range. The random impact device will be lighter, simpler, and easier to use than a shaker; it can also be designed to excite very large structures. Collaboration
