Despite its great potential, the reliability of the VAM-method regarding the detection and monitoring of local fatigue damage in metals cannot be taken for granted yet. This is due to the fact that it has not been clarified how various parameters, such as the applied ultrasonic frequency f_c, the sensor positioning, the geometry, locally varying material properties, connections (such as welds or bolts), among others, contribute to the modulation, which is used for damage quantification. Secondly, it is physically not yet well understood how local defects in materials cause VAM in the first place. It has been our primary research target so far to overcome these two limitations. 
Our experimental findings show that a local fatigue crack in aluminum plates only generates significant modulation if a suitable carrier frequency fc is excited. Secondly, it was demonstrated that a high initial (non-damage induced) modulation at the very beginning of fatigue lifetime can camouflage the non-linear contribution of local defects which develop later. Also the non-damage induced modulation was found to be dependent on the excited ultrasonic frequency fc (see Figure 1). Both results are jeopardizing the reliability of the VAM-method for structural health monitoring purposes. Based on these findings an alternative evaluation procedure for VAM-monitoring was developed: Multiple carrier frequencies with very low initial modulation are determined at the beginning of a structure’s fatigue lifetime. Those are evaluated independently over the entire lifetime. As soon as single modulation developments show an exponential increase (indicating fatigue damage initiation), only the correspondent carrier frequencies are taken into account for damage evolution monitoring. We demonstrated experimentally that this approach has the potential to improve the method’s reliability dramatically. Regarding the second target—explaining the occurrence of VAM—we formulated a physics-based analytic approach, that explains why a local defect is causing vibro-acoustic (amplitude and phase) modulation in materials. The approach takes into account the steady-state vibration response resulting from the excitation with the ultrasonic frequency f_c. The approach is capable of explaining why a fatigue crack only leads to measurable modulation if certain carrier frequencies are evaluated. The reason lies in the natural vibration modes of the structure. If the defect or the sensors lie in the vicinity of a node of the excited vibration mode (amplitude = 0) the non-linear characteristics of the fatigue damage do not affect the measured system’s response.
The explanation is currently employed to find ways to localize fatigue damage within the structure. Besides the goal of localization, work is carried out on finding evaluation parameters that might be more sensitive to fatigue damage than the Modulation Index (MI) and, hence, allow the detection of fatigue damage initiation earlier in life.