Overview
Smart structures, which incorporated structural health monitoring (SHM) systems into mechanical components to enable online damage detection, had a strong potential to reduce inspection and maintenance cost of aircraft structures, while simultaneously maintaining or increasing the level of safety. These technologies were still in the R&D phase and significant efforts are required before smart structures will become the industry norm.
One of the research needs concerned planning of inspection and maintenance for such structures. To be able to fully exploit the capabilities of novel monitoring technologies, it is indispensable that tools were developed to plan maintenance strategies for given safety philosophies (fail-safe, safe-life, damage tolerance) taking into account information obtained from SHM.
The goal of this project proposal was to develop a methodology and corresponding software algorithms for this task. Taking basis in the renewal theory as part of the general Bayesian decision analysis framework, tools for the life cycle optimisation of inspection/maintenance activities were developed. Based on the commercial Strurel software code for reliability analysis developed and owned by the consortium leader the software tool shall be able to assess the effect of different strategies on safety and cost and thus enable the identification of optimal maintenance strategies subject to constraints on safety. The methodology was applied to a stiffened panel provided by the ITD.
Funding
Results
Executive Summary:
The optimisation of maintenance schemes for aircraft structures was a major technical and economic issue in the aerospace industry. There was a strong interest to further optimise and customize maintenance activities, aiming at reducing costs directly and indirectly associated with unnecessary inspections and replacement actions. Technical developments in Structural Heath Monitoring (SHM) provided new possibilities for inspecting engineered structures in general. Knowledge of the “damage level” of the structure and forecasting of its development on a probabilistic basis allows the intelligent system to plan removal from service of the aircraft when the requirement for maintenance is specifically justified
The methodology and the corresponding codes developed in the ROSA project facilitated the quantification of the impact of SHM information on maintenance planning and the optimization of the life-cycle costs. The framework developed enabled the quantification of the effect of changes in the design of the structural components as well as the SHM design on the inspection / maintenance cost and consequently the overall life-cycle cost. All relevant models for the problem description have been presented, developed and implemented into a code. Main achievements of the ROSA project are summarised as follows:
- A mechanical meta-model for the description of the damage growth on stiffened composite material due to different impact scenario has been identified and implemented.
- A procedure for the inclusion of the monitoring system has been investigated. Models for the probability of detection and the receiver operating characteristic have shown to be essential in the perspective of the Bayesian damage update, at the basis of this report.
- The method for the evaluation of the capacity loss due to the damage detected by the monitoring system can be extended in order to include an alarm system for decision support.
- Complex structures, as the aircraft wing in final report, can be easily handled by the method, accounting for simplified structural effects on the element capacity.
- The ROSA strategy has been implemented for specific models but it can easily be adapted to more complex models, due to the modularity of the methodology.
Sensitivity analysis has been presented by changing the most relevant parameters. It has been shown that the quality of the monitoring system, expressed by the Receiver Operating Characteristic (ROC) curves is central to the optimization of the optimal interval between consecutive inspections.
The expected life-cycle cost analysis has been presented. The analysis evidenced that the probability of false alarm of the structural health monitoring systems plays an essential role in the total expected maintenance element cost.
The theory has been implemented in standalone a code and details, examples of runs and results have been provided. The possibility of interfacing the code with the software Strurel has been examined and conceptual solution has been given.