This paper describes the use of decision fusion strategy in damage detection. These techniques fuse multiple individual
damage detection measures to form a detector with higher probabilities of correct detection than that attainable with any
of the individual measures. In this paper, these technique is applied to vibration-based damage detection methods. As a
demonstration of the methodology, the first step was to fabricate an experimental fixture which the vibration properties
of damaged and undamaged structures can be measured. The experimental results with the undamaged structural model
provided information for producing an improved theoretical and numerical model of the mechanics via model updating
techniques. Three common vibration-based damage detection methods using a varied multi-resolution on the
experimental results were implemented to identify the damage that occurred in the structure. Finally, the strategy to join
the information from the three methods with multi-resolution decision fusion rules is introduced. The fused results are
shown to be superior to that from only one method.
Car manufactures are turning to high pressure hydrogen storage for on-broad power applications. Fiber-composite-wrapped
high pressure hydrogen tanks are becoming widely used in onboard vehicle storage applications because of
light weight and high strength. It is widely accepted that the worst case of the equipment at operating pressure should be
only leakage without the risk of explosion. In order to ensure the safety during the operation course, the damage
detection and leakage alert of fiber composite wrapped tank for high pressure hydrogen storage should be investigated.
The aim of this paper is to find an effective nondestructive damage detection method for the identification of fatigue
cracks on composite wrapped tank. First, a three-dimensional finite-element model is developed as the baseline model.
Then fatigue crack in inner aluminum alloy, as the typical damage form, is simulated with the position, length, and
direction of the crack as investigation parameters. Two nondestructive damage detection methods are applied to identify
whether the damage has occurred based on the natural frequency and mode shapes of the fiber composite wrapped tank.
The damage detection capability of each method is studied, and the influence of the vehicle vibration caused by road
surface roughness and environment noise on damage detection are discussed. Finally, feasible strategy to alert the
leakage of the hydrogen of fiber composite wrapped tanks is suggested.