This study presents a novel strategy for identifying an impact event on a structure from vibration measurements. Compared to triangulation techniques that require at least three sensors on a plate-like structure, only one sensor is used in this work for localizing the impact point. The proposed approach consists in extracting specific modal ponderations as a signature of impact location. The force reconstruction problem is simultaneously addressed by fitting a parametric law. The proposed procedure captures the main load history parameters such as the impact time, its duration and its intensity. An impact test campaign is performed on a metallic plate, equipped with one accelerometer only, to confirm the validity of the proposed single-sensor approach. The methodology is efficient for accurately localizing impacts that are applied anywhere on the plate and for quickly estimating the main load history parameters.

No abstract

This paper focuses on the localization of impacts applied on a composite aircraft by using low sampling frequency accelerometers. A new vibration-based approach using a modal model of the aircraft on ground is described to localize the impact on the cylindrical fuselage. Axial localization is achieved by estimating the contributions of specific vibration modes to the response. Angular localization is achieved by estimating the impact force direction. An impact test campaign is performed on an Airbus A350-900 equipped with only six accelerometers. Fifty percent of the applied impacts are localized within 2 m and more than 90% within 3 m.

Impact identification is a major concern in structural health monitoring. It consists in localizing the impact point and in reconstructing the applied load history from indirect measurements. However, it is well-known that this double inverse problem is ill-posed. In this paper, a single-sensor approach for quickly localizing a Dirac-like impact and for estimating its intensity from vibration measurements is proposed. It is proved that the relative proportions of specific modal ponderations are a signature of impact location. Rules are provided to select the appropriate vibration modes in the analysis and to position the sensor. The identification process is robust if the vibration modes kept in the analysis are weakly damped and if their natural frequencies are well separated. Experiments are conducted on a metallic plate to validate this singlesensor approach on a 2D structure. The success rate for localizing impacted cells is 100% when appropriate method parameters are selected.

No abstract