13:30
Session 8: NDI, inspections and maintenance
Chair: Eric Lindgren
13:30
20 mins
|
NASA NDE fracture critical detectable flaw sizes history and methodology
Peter Parker, William Prosser, Ajay Koshti, David Forsyth, Michael Suits, James Walker
Abstract: NASA requires that NDE methods and inspectors demonstrate 90% Probability of Detection (POD) with 95% confidence for critical flaw sizes when inspecting fracture critical metallic components. NASA addresses the known variability of NDE inspector capability in two ways. The first, Special NDE, requires that every inspector demonstrate the required 90/95 POD, which is resource intensive. The second approach is Standard NDE for which conservative flaw sizes for different NDE methods are provided such that it is expected that most properly trained inspectors will exceed the 90/95 POD requirement. As such, individual POD demonstration testing is not required.
The origin of NASA Standard NDE dates to the start of the Space Shuttle Program in the early 1970’s. In the first study to quantitatively assess NDE methods and inspectors, the performance of multiple inspectors was evaluated for different NDE methods using a large set of fatigue cracked specimens. A rudimentary POD analysis was performed to estimate the 90/95 POD flaw size for each inspector for each method. Additionally, the average and standard deviation of the 90/95 POD flaw size across the multiple inspectors was calculated to estimate the flaw size for which 95 percent of inspectors would provide the 90/95 POD capability. These estimated 90/95/95 POD flaw sizes evolved into the NASA Standard NDE flaw sizes still in use for structural analysis five decades later.
The methodology for performing Standard NDE POD studies was never documented in NASA requirements. Furthermore, POD analysis methods have significantly evolved since this seminal study. Likewise, NDE methods have improved and there has been a push to reassess Standard NDE flaw sizes for existing methods, and to develop Standard NDE flaw sizes for new methods such as digital radiography. In this study, a Standard NDE POD methodology was developed and baselined using the historical data. This reanalysis of the historical data identified several deficiencies in the original test plan as well as an overall lack of conservatism in the estimated 90/95/95 POD flaw sizes. The results of this historical review and the new methodology are being incorporated into an update of NASA NDE POD requirements.
|
13:50
20 mins
|
Aviation-based nondestructive evaluation data analytics
Eric Lindgren
Abstract: Recent connectivity and digitization of aviation maintenance equipment has increased the potential of developing an Internet of Things 4.0 approach to enhance aircraft availability. Typically, these systems generate more data which nucleates interest in using analytical methods, such as artificial intelligence and machine learning (AI/ML), to increase the effectiveness of current aviation maintenance practices. It is important to recall that AI/ML methods are based on statistical regression and classification techniques. However, before such algorithms can be applied, considerations must be given to the quantity and quality (precision, accuracy, and noise) of the data to enable AI/ML. Several case studies are presented that explore these questions and indicate a careful assessment of the data is required to understand the accuracy and the distribution of the results from such analysis. The potential for the use of AI/ML is explored further using nondestructive evaluation (NDE) data. A significant challenge for these analytical methods is the limited amount of data captured for the features of interest, such as fatigue cracks and corrosion. Recall that trends in fleets lead to replacement or modification initiatives before an extensive amount of flaws are present. To mitigate this limitation, the Air Force Research Laboratory (AFRL) and collaborators have explored and implemented alternative methods to assist in the analysis of NDE data that integrates at least two of the following: heuristics, model-based, and data-derived analysis techniques. In addition, success has occurred when retaining the expertise of inspectors, i.e. humans-in-the-loop, to ensure the quality of the decision-making process. AFRL calls this approach Intelligence Augmentation (IA). The USAF has a rich history of using IA to analyze large NDE data sets, typically acquired from inspections that use automated scanning to acquire data. Several representative examples that include at least two of the three analysis methods are discussed, including the implementation process. These examples illustrate the benefit of integrating all resources to enable accelerated decisions with data limitations and the value of retaining humans-in-the-loop. Future opportunities include improved integration of models, especially as a function of their maturity through validation.
|
14:10
20 mins
|
A new approach to accidental damage on aircraft metallic structure
Sébastien Amiable, Ben Ogborne, Alain Santgerma
Abstract: A new approach to define the inspection requirements for Accidental Damage (AD) on aircraft metallic structure is described in this paper. Accidental Damage is characterized by the occurrence of a random discrete event (e.g. impact by foreign object, hail, runway debris, human error during manufacturing, operation or maintenance of the aircraft) which may reduce the inherent level of residual strength of the structure.
The Accidental Damage inspection tasks were defined by Airbus through a crack growth calculation from detectable to critical on the structural areas prone to this type of damage. These inspection tasks were part of the Airworthiness Limitation Section.
Some years ago, it was decided to change from a calculation based approach towards a more realistic and pragmatic approach.
To achieve this, test and in-service experience accumulated on AIRBUS aircraft was exploited. It appeared that very few non-detected accidental damages turned into fatigue cracking in service. Test results showed that readily detectable damages such as dents take a long time, providing many detection opportunities, before a fatigue crack initiates and grows. The philosophy of the approach was changed from crack detection to accidental damage detection.
An MSG-3 evaluation is applied to identify the Significant Structural Items (SSIs) susceptible to AD. The likelihood of AD occuring is evaluated and an Accidental Damage Rating (ADR) is determined. The fatigue, crack growth and residual strength performance of the SSI are used to derive a Stress Sensitivity Rating (SSR), assessing the likelihood of AD to turn into growing fatigue crack. The inspection frequency, expressed in calendar time, is determined through the combination of the ADR and the SSR. The inspection level, general or detailed visual, is determined through a standard MSG-3 assessment. Finally, inspection requirements are consolidated with the Zonal and Structure Programs of the Maintenance Review Board (MRB) Report.
This new approach was developed in a joint effort between the Maintenance Engineering and Stress departments in Airbus. While being more realistic than the previous one, it also optimized the overall number of inspections in the scheduled maintenance program.
|
14:30
20 mins
|
Probability of detection of automated tap testing for disbond detection in metallic honeycomb structures
Marc Genest, Shashank Pant, Muzibur Khan, Catalin Mandache, Dmitrii Klishch
Abstract: This paper investigate the reliability of automated tap testing for honeycomb composite panels representative of those found in helicopter structures. A Probability of Detection (PoD) study was carried out for detecting disbond on honeycomb panels using manual and automated tap testing techniques. A process was developed to generate controlled disbond size in the honeycomb panels. The process was confirmed via ultrasonic testing and micro computed tomography inspection. A reference panel was manufactured as well as eight test panels containing a total of 70 different damage sites that included dents, disbonds, as well as combined dent and disbond were inspected by 11 inspectors using both manual and automated tap testing techniques. Overall, the a90/95 value decreased by 0.6 inch equivalent diameter using the automated tap test as compared to manual tap testing (from 1.70 to 1.10 inches). Furthermore, the automated tap testing led reduction in false calls, up to 3 time less.
|
14:50
20 mins
|
A new method for defects detection in CFRP composites using wavelet analysis and non-contact lamb waves propagation
Lea Lecointre, Ryo Higuchi, Tomohiro Yokozeki, Masakatsu Mita, Shota Tonegawa, Naoki Hosoya, Shin-ichi Takeda
Abstract: Ultrasonic Testing (UT) is the most common Non-Destructive Testing method used for Composite materials. One of the most promising methods which have been identified in the recent years in order to improve the efficiency of UT is the Lamb Waves Testing. However, Lamb Waves propagation is a complex phenomenon which requires adapted and efficient Signal Processing method.
In this study, we propose a method based on the Wavelet Analysis to process experimental Lamb Wave signals propagated into Carbon Fibers Reinforced Plastics (CFRP) samples for detecting artificial defects. Moreover, this experiment uses fully non-contact generation and reception of the Lamb Waves which shows very promising applications for maintenance or complex geometries.
The Lamb Waves were generated in several healthy and artificially delaminated CFRP test samples from a Laser Induced Plasma system. The reception of the propagating waves was performed by measurement of the out-of-plane displacement with a Scanning Laser Doppler Vibrometer.
Then, a continuous Wavelet Transform was performed on the measured signals, which allowed to store values of Wavelet Transform coefficients (cwt) in a 3D matrix in spatial, time and frequency domains. From this matrix, spatial-time analysis was performed by extraction of 2D images at each frequency sample. The experimental group velocity dispersion curves were calculated and compared with theoretical values. The results shown that the experimental group velocities had a good correlation with theory. Moreover, for 16-ply samples including delamination at the middle depth, group tend to reach the value of 8-ply structure.
Finally, a spatial-frequency analysis was performed from the 3D cwt data. Several time samples were extracted for healthy and delaminated plates. The spatial-frequency data for healthy samples were used as baseline for detecting signals in samples containing delamination. The results shown clear and accurate localization of wide delamination from the visible increase of cwt amplitude and frequency bandwidth at the delamination zone. The same process has been performed on samples containing several small delamination. The smaller delamination (Ø10-mm) could not be detected, while Ø30-and Ø50-mm delamination could be detected as groups of defects.
|
|
