Eddy current measurement system evaluation for corrosion depth determination on cast aluminum aircraft structure

Abstract

An eddy current (EC) technique was developed to determine the corrosion depth on a bare flange face of a cast aluminum A356-T6 aircraft engine structure. The EC response and the corrosion depths determined through metallurgical cross sections were used to develop an empirical relation between EC response and depth. The EC technique and depth determination are used to inspect the engine structures during overhaul to determine if they are fit for continued service. An accurate and reliable Non-Destructive Inspection is required to ensure that structures returned to service are safe for continued operation. NDE system reliability demonstrations of the eddy current technique are traditionally reported in terms of Probability of Detection (POD) data using MIL-HDBK-1823A. However, the calculation of POD data is based on a simple linear predictive model that is valid only if certain criteria are met. These are: 1) NDE system response is measurable (i.e. continuous data), 2) Flaw size is known and measurable (i.e. continuous data), 3) relationship between the NDE system response and flaw size is linear (or linear on a log scale), 4) variation in measured responseresponse around a predicted response for a given flaw size is normally distributed, 5) the variation around the predicted response is constant (i.e. variation does not change with flaw size), and 6) inherent variability in the NDE system is known and fully understood. In this work, a Measurement System Evaluation (MSE) of the Eddy Current System was used to address some of these concerns. This work was completed on two aircraft structures having varying corrosion depths. The data were acquired in a random manner at fifty regions of interests (ROIs). Three operators participated in this study, and each operator measured Eddy Current response three times in each ROI. In total, there were four hundred and fifty data points collected. Following this, the two structures were sectioned for measuring corrosion depth. The obtained EC response and depth data were used for quantifying the EC System inherent variability by determining Analysis Of Variance (ANOVA), Gage R&R, control charts, Gage run charts, and regression curve. Initial results from this study show that the Eddy Current System is capable for determining corrosion depths on the structures as well as for discriminating good parts from bad.