L aurent Ponson

Abstract

Predicting the residual lifetime of structures require the prior knowledge of the material failure properties and the applied loading conditions, two crucial information that are generally missing for in-service structures.

Here, we propose a new methodology of predictive maintenance that provides such information without requiring to destructive testing. Instead, our approach measures both the residual mechanical properties and the applied load amplitude from the sampling and then the test of a (non-destructive) millimetric sample. This innovative approach relies on the determination of the fracture properties of materials through the scan and analysis of their fracture surface by statistical fractography, an engineering technique that can be applied robustly and safely to submilimetric specimens.

In this presentation, I will show how this innovative approach paves the way for the reliable predictions of the residual lifetime of structures for various ranges of applications, including cracked structures under fatigue conditions. I will also finally the applications of this methodology and its implications in maintenance of old structures through industrial use cases.

Session