Fiberglass sandwich panels are tested to study a vibration-based method for locating damage in composite materials.
This method does not rely on a direct comparison of the natural frequencies, mode shapes, or residues in the forced
vibration response data. Specifically, a nonlinear system identification based method for damage detection is sought that
reduces the sensitivity of damage detection results to changes in vibration measurements due to variations in boundary
conditions, environmental conditions, and material properties of the panel. Damage mechanisms considered include a
disbond between the core and face sheet and a crack within the core. A panel is excited by a skewed piezoelectric
actuator over a broad frequency range while a three-dimensional scanning laser vibrometer measures the surface velocity
of the panel along three orthogonal axes. The forced frequency response data measured using the scanning laser
vibrometer at multiple excitation amplitudes is processed to identify areas of the panel that exhibit significant nonlinear
response characteristics. It is demonstrated that these localized nonlinearities in the panel coincide with the damaged
areas of the composite material. Because changes in the measured frequency response functions due to nonlinear
distortions associated with the damage can be identified without comparing the vibration data to a reference (baseline)
signature of the undamaged material, this vibration technique for damage detection in composite materials exhibits less
sensitivity to variations in the underlying linear characteristics than traditional methods. It is also demonstrated that the
damage at a given location can be classified as either due to a disbond or core crack because these two types of damage
produce difference signatures when comparing the multi-amplitude frequency response functions.