

Demonstration of Model Assisted Reliability Assessment Protocol on a Proposed Low Frequency Vibration Based Damage Sensing Case
Abstract
This paper describes the progress on demonstrating a methodology for reliability assessment for structural damage sensing (SDS) systems. In particular, it presents the initial results obtained from the application of a protocol designed for utilizing empirical data, models, simulations, and uncertainty analyses for statistically characterizing SDS reliability for a damage detection case. The manufactured test article representing an aircraft structure of medium complexity consists of three plates connected by two lap joints with fasteners. Fatigue crack damage around the fastener holes was simulated by manually created thin cuts at selected locations. The test fixture design provides the capability to vary critical parameters of the system with a focus on force loading boundary conditions, joint fastener torque conditions, and temperature. The initial demonstration on this test article and fixture uses a low frequency vibration based damage detection method. Frequency domain metrics were utilized for studying changes in structural dynamics due to mechanical loading, thermal loading, and actual damage. For the demonstration, key factors that affect the capability to sense the effects of damage were assessed through controlled studies of (a) loading and unloading, (b) fastener torque, (c) boundary condition variation, (d) temperature variation and temperature gradients, (e) sensor bond quality and operation performance, (f) ambient noise, and (g) sensitivity to flaw growth. The design of the full validation study and the current results of the implementation are presented, highlighting general protocol feasibility while identifying remaining challenges for full demonstration and broad use of the methodology and protocol.