[home] [Personal Program] [Help]

---

Session: Session 17: Probabilistic modelling and risk analysis 
Session starts: Thursday 29 June, 11:20
Presentation starts: 12:00
Room: Theatre room: plenary

---

Viola Ferrari ()
Min Liao ()
Michea Ferrari ()
Michel Guillaume ()

Abstract:
Structural risk assessments are highly relevant to assure the structural integrity over the whole life-cycle of an aircraft. The Aircraft Structural Integrity Program (ASIP) of the US Air Force introduces design guidance based on deterministic crack growth prediction for safety critical components. Today, with the increase of computational power, probabilistic methods instead of deterministic analysis can be deployed. Probabilistic Risk Assessment (PRA) offers the benefit of taking uncertainty into account and leads to less conservative estimates while meeting safety requirements. Research organizations such as the National Research Council Canada (NRC) or the Defense Science and Technology Organization in Australia (DSTO) are already assessing safety critical elements with PRA. RUAG AG is interested in the application of this method for current and future military aviation systems. Based on the research performed by these organizations, the application of PRA is assessed. The probabilistic calculations are performed with the software tool SMART|DT. Structural engineering and fatigue data is needed for such an assessment, which includes equivalent crack growth data, pre-crack size distributions (EPS), loading distributions and probability of detection (POD). The possibilities of the tool are assessed with four case studies with different data sets (amount of data, quality of data, etc.). Based on the case studies, existing methods are explored and implemented for the input data in order to perform calculations with the software SMART|DT acquisition. For better understanding PRAs and their application for the system, evaluations are performed including comparisons between various spectra, different material data and the influence of different distributions for the probability of failure. With the applied methodologies and the use of SMART|DT tool, the case studies show very conservative results compared to current crack finding during fleet inspections. The application of the tool must be assessed on a case-by-case basis depending on the available input data. The paper shows a possible application of PRAs using the available data with limited effort in the data gathering process and includes simple ways for the data preparation and application in SMART|DT based on known PRA methodologies.