Paper presented at the 8th ECNDT, Barcelona, June 2002

Abstract

Work has been focused on practical aspects of infrared thermographic inspection of composite aircraft structural components as performed in maintenance facilities and repair stations, which relate to test sensitivity, identification of true indications and evaluation of damages. IR thermographic inspection of a specific reference sample has been carried out and then correlated with the results after alternative NDT methods. The most significant results will be now presented to support some conclusions in an attempt to further improve suitable inspection reliability.

Introduction

IR thermographic inspection of in service composite aircraft structural components is becoming, day after day, a more extended NDT practice in both aircraft repair stations and maintenance facilities. Not surprisingly, major users’ concerns relate again with its impact on inspection reliability and cost effectiveness and, accordingly, a review of actual practices and inspection results, together with further extensive tests are needed to highlight some related key factors.

Present work concentrates on some of the factors affecting delectability and evaluation of water ingress related indications in honeycomb panels (aerospace structures) when applying conventional active IR thermography. Particular attention has only been paid here to required inspection procedures; the applicability of other IR techniques, though beyond the scope of this presentation, must be considered regarding their potentially outstanding capabilities, impact on reliability and affordability, as a key point.

For the detection of water ingress into aircraft honeycomb panels with IR thermography, reliable interpretation of obtained indications not only requires the complete understanding of the basics, scope and limitations of the particular inspection technique and both suitable familiarisation with the products and processes but, in addition, the use of both a suitable reference sample and an accurate enough information on the actual configuration of the part under test, including all existing repairs and relevant modifications, are required.

Experimental Results

In order to properly identify and evaluate some of the key variables affecting the reliability of inspections performed on actual aircraft components, a series of tests were performed on.a reference sample containing artificial discontinuities, including different contents of water in cells. As a complement and a reference for comparison, suitable radioscopic tests were also been performed.

For this first stage, the raw surfaces of the reference sample (CFRP faces) were just cleaned for IR inspection and effect of differences in emissivity due to paints, etc. has not been considered.

Test sample

The reference sample is a 1000 x 420 mm sandwich panel with two 3.36 mm thick CFRP solid laminate (16 prepeg plies each, 0/45º) face skins. These were adhesive bonded to a pair of nomex honeycomb cores (hexagonal and rectangular cells). Each core has two areas 39 and 60 mm thick and they have been bonded together with a foam line. The core machining, polymerization and autoclave curing was performed as required in usual manufacturing procedures (aerospace composite components).

Into the reference sample (Fig. 1), a series of artificial discontinuities has been included to simulate typical in process and in service defects such as disbonding of adhesive foam lines, excess of foam in bond line, cracks in foam reinforcements, disbonding of honeycomb core nodes, breakage of honeycomb core, water ingress, etc.

Figure 1. Reference sample: general description with position of discontinuities, details (sketches) of the areas with different water contents and picture of an intermediate step in the manufacturing of the sample.

Radiographic inspection

As a reference for comparison, detailed RT inspection was also performed with an automatic radioscopic system equipped with a Rich.Seifert X-ray tube Isovolt HS160 and a VistaPlus image intensifier. The series of radioscopic images from the different areas of the reference sample are shown in the following figures next to the suitable thermographs.

Thermographic inspection

Prior to test the sample was uniformly heated in a controlled furnace up to 40 ºC above room temperature (measured 22 ºC) for 40 minutes. After that, colour and grey scale thermograms were taken within the following 20 minutes at controlled time intervals (Fig. 2) using a FLIR ThermaCAM P/N 545 camera (temperature range from -20 to 350 ºC with nominal sensitivity of 0.1 ºC at 30 ºC).

Fig 2: Colour and grey scale thermograms of the reference sample after different time intervals (general views of the sample).

The details of the most significant indications out of the different thermograms have been classified together with the suitable RT images, shown as a reference, for further discussion (Figs. 3 to 6).

Fig 3: Thermogram after 15 min and radioscopic images of the areas P, AB, R, X, H and J (1/2 and 3/4 water content).

Fig 4: Thermogram after 15 min and radioscopic images of the areas AD, AG, AQ and AN (1/2 and 3/4 incomplete foamed reinforcement).

Fig 5: Thermogram after 15 min and radioscopic images of the areas L, N, AL and AJ (1/2 and 3/4 incomplete foamed reinforcement).

Discussion

Applying this IR inspection technique allows the detection and resolution of differing water contents in honeycomb cells (Fig. 3) as well as incomplete foamed reinforcement areas (Figs. 4, 5). Only in the 39 mm thick half of the reference sample, some details of the foam bonding line between cores can also be appreciated with just reasonable confidence, like areas with excess foam (BF) and, though with some difficulty, the lack of foam in area G-G, too (Fig. 6). None of the other discontinuities in the reference sample can be clearly detected and different cell configuration (hexagonal or rectangular) does not appear to significantly affect inspection results.

Fig 6: Thermogram after 5 min (detail) and radioscopic image of the area G-G (lack of foam in core bonding line).

Related to this capability, though not affecting the aim of the present study, both IR and RT results have clearly shown the unwanted loss of water at positions H and J for some reason during the manufacturing of the sample as well as some other minor (basically geometrical) differences with the original drawings of the reference sample. Regarding the above mentioned IR detectable discontinuities, correlation between these and the suitable RT indications are fairly good in spite of the clearly better sizing capability when evaluating the suitable radioscopic images (Figs 3 to 6). Though not being the only one, the key factor of uncertainty is the absence of an objective correlation between the color or grey scales used along with temperature or temperature differences, in attempting to size or to make a quantitative evaluation of a given IR indication.

Another important point related with the detectability of those discontinuities is the clearly better reliability achieved when recording colour scale thermograms rather than grey scale ones (Fig. 2). Water ingress is in any case detectable in grey scale thermograms though just appearing as slightly darker spots in comparison with the indications at areas of foam reinforcement but, in this regard, the readiness to resolve between those two discontinuities is far better when interpreting colour thermograms and, specially, when the water content is small (Fig. 7).

As already well known, the optimum time interval for IR inspection heavily affects inspection reliability and must be specifically established each time attending to, basically, the nature of the part, the part geometry (in the case of our reference sample, core thickness appears to be the key variable) and the nature of the discontinuity to detect. Detection of water and incomplete foam reinforcement in the 39 mm thick half of the reference sample can be reliably performed from about 5 minutes after the part has left the furnace, but not before 15 minutes when testing the 60 mm thick half, and these suitable inspection conditions stand for a following period of about 15 to 20 minutes. However, if looking at other type of discontinuity, the indication of the foam bonding line between the cores, at least in the thinner half of the panel, becomes reasonably inspectable just between 5 and 10 minutes after heating (Figs. 2 and 6).

Fig 7: Thermograms (grey scale and colour) of a horizontal stabilizer showing two areas of water ingress

Conclusions and Remarks

Within the scope of the applied IR technique and, specifically for the considered type of composite materials and components under test, the results obtained have brought some.light on the practical aspects of the suitable inspections to be performed at aircraft maintenance and repair facilities.

Related proposals for improvement must be formulated after taking into account the actual practical constraints affecting inspection reliability, affordability and associated inspection costs. When attending to the practice of the IR inspections and the factors affecting inspection reliability and inspection/repair related costs in an aircraft maintenance or repair workshop rather than in a research environment, these can actually be simple findings, but of the highest importance and usefulness. For the specific object of detecting water ingress in honeycomb panels:

1. The systematic use of color thermograms instead of grey scale ones will enhance inspectors confidence for the identification of water ingress. In practice, reliability related to avoiding false calls becomes much more important than a not really needed accuracy in determining the actual size of the suitable affected areas.

2. The specification of a suitable standard reference sample and the availability of the key technical information about the part under test will also warranty the inspection reliability requirements. Specified reference samples must fit the significant part features (materials, thickness, configuration, etc.) and also include the suitable discontinuities (different contents of water, in this case). On the other hand, the inspector must know about the actual configuration of the part tested, which should include all the repairs performed, etc. (related to the traceability of the part).

At the same time, ongoing further developments, evaluation and implementation of alternative IR inspection techniques for in service maintenance will both broad the method application scope and improve suitable inspection reliability.

References

1. ASNT Nondestructive Testing Handbook, 3 rd Ed. Vol. 3. “Infrared and Thermal Testing”, 2001.
2. V. Dattoma et al., NDT&E International 34, (2001) 515-520 “Thermographic investigation of sandwich structure made of composite material”.

© NDT.net - info@ndt.net

|Top|