Abstract

The inspection of aerospace CFRP components with ultrasonic techniques is subject to very challenging requirements in terms of ensuring a reliable and time efficient non-destructive examination. The different approaches such as squirter, immersion and contact techniques are discussed under these aspects, which seem to be contradictory at first view.

Assessing the advantages and the weak points of each of these methodologies in view of the time and cost constraints of the fabrication process, the conclusion resulted in an examination strategy, which defines the optimum range of application for each of the a.m. methodologies. This strategy also defines the requirements in order to optimize the application of each of these methodologies.

As it is the least known of the methodologies mentioned, the paper places some emphasis on the presentation of the contact technique with a multisensor approach. This system combines high frequency conventional and low frequency phased-array probes and facilitates pulse-echo technique and backwall monitoring - even in the nonparallel sections - as two complementary sources of information about the structural integrity of the component.

The development of this technique and its optimization performed have helped to determine the optimized parameters of such techniques and the optimum number of channels between "very high" associated with high cost fast examination and "low" as a low-cost version, which in turn requires a considerable time for the examination.

Most of this work leading to the implementation of such system was performed in the framework of the German research project MaTech, partly sponsored by the Federal Ministry for Education and Research. Project partners are Airbus in Bremen, Ingenieurbüro Dr. Hillger in Braunschweig, TUHH-Technologie GmbH and intelligeNDT Systems & Services in Erlangen.

The experience achieved with this system has proven its compliance with the performance as planned under the aspect of reliability and time-efficiency.

Introduction

The use of CFRP components is strongly progressing with the introduction of more and more advanced aircraft types in the military and the commercial aircraft market. The share of CFRP components in an aircraft has risen up to 20 %, as the example of the Airbus A 380 clearly demonstrates (Fig. 1).

Fig 1: Composite applications in commercial aircraft.

Fig 2: Major monolithic CFRP and thermoplastics application of an A380

The inspection of aerospace CFRP components poses some considerable challenges to the application of NDE to be integrated into the fabrication process in an optimum way. The boundary conditions for an optimum solution are quite different accounting for the variance of the characteristics of the different parts to be examined:

As an example, in terms of consequences for the methodology to be employed, honeycomb structures can only be examined by the squirter technique mainly for preventing water from intrusion, as would be with immersion techniques.

Comparison of different approaches

In context with the boundary conditions valid for the different components, the advantages and disadvantages of the three categories of available methodology have to be thoroughly studied. They can be summarized as follows:

• The squirter technique works predominantly in the through-transmission mode. However, recent development facilitates the use of pulse-echo technique additionally. Still, questions such as the signal-to-noise-ratio and the near surface resolution need to be answered to complete satisfaction. In a necessary multiple sensor arrangement, the water jets tend to interfere with each other, if they are too close to each other, disturbing the laminar water flow in the jet for an impracticably large width of the system if more squirters should be used. This makes the system bulky with a complex rastering scheme, if the geometry of the component is not ideally rectangular.
• The immersion technique requires a precise adjustment of the angular positioning of each probe due to the relatively large influence on the refraction angle in the case of even slight deviations of the plain component geometry such as some curvature systematically occurring at the peripheral parts of the component. The problem of precise adjustment also applies for maintaining a constant distance between probe and component surface. This would mean an unreasonable effort of 3-D-controlled positioning individually for each of the multiple probes.
• The contact technique allows for the arrangement of a multiple probe system with relatively simple probe holder system allowing the probes to individually adapt to the geometry they are riding on. If probes with different parameters such as frequency are to be used, they can easily be accommodated whereas the other two techniques may require different jet distance or immersion distance conditions, which would even more complicate the complex system.

Consideration of the components and the corresponding inspection methods

For the purpose of a coarse assessment of the advantages and the weak points of each of the three methodologies in view of the time and cost constraints of the fabrication process, the components are classified into different categories as follows:

In view of the characteristics of the different methods and their special capabilities, the different components have to be analyzed in terms of their geometry. Some parts with complex geometries, but generally smaller dimensions, call for a small number of probes. This can be facilitated with water jet coupling (Squirter) in view of the curved surface geometry and a complex motion control with up to 7 axis for each sensor. The alternative available is based on advanced technology utilizing robotics approach.

Fig 3: Solution for the inspection of components with complex geometries.

Besides using one pair of squirter probes this approach allows for the application of one, max. two probes in contact technique. Furtheron, the geometric conditions for precise positioning are quite stringent in order to facilitate the use of through- transmission techniques looking for the shadowing of the through-transmission signal by a potential defect or - with the same principle - backwall reflection technique with incidence perpendicular to the backwall.

The parts of a smaller degree of complexity and/or with - to a larger extent – plain surfaces may be examined using the squirter technique, the immersion technique or the contact technique with a smaller number of probes, allowing for pulse-echo and backwall reflection techniques, eventually also through-transmission techniques.

Large parts with a low degree of complexity, i.e. with practically parallel surfaces pose other requirements, which are quite contradictory for themselves.

• The integration into the fabrication sequence leaves a short time interval for the examination performance relative to the large dimensions of the components to be scanned
• For the purpose of safety, the pulse-echo technique and through-transmission or pitch-and-catch technique via backwall signal monitoring should be combined. For the latter case, the method has to cope with nonparallel contours
• The dimension of the minimum detectable defect is relatively small, which means a small rastering index
• The relatively high sound attenuation would call for lower frequencies
• The resolution of defects even in immediate subsurface location would call for higher frequencies
• The demands for the positioning accuracy are very stringent

Examination strategy

When attempting to draw a conclusion in terms of an examination strategy, which defines the optimum range of application for each of the a.m. methodologies, we can highlight the following items:

For the squirter technique, the maximum number of squirter jets is four with sufficient distance from each other in order to prevent interference. The scanning velocity is relatively high with up to 1000 mm per second, which helps only for larger dimensions. For complex geometries, only one squirter pair can be employed, otherwise the movement patterns and the control system would be too complicated. Besides, the complex geometry does not allow for the a.m. high scanning velocity.

The immersion technique also allows for a high scanning velocity and the coupling does not pose any problems as soon as air bubbles have been removed. The removal of bubbles also from the opposite surface is necessary in order to maintain a constant backwall signal. The system has no problem with the edges of components, as there is no mechanical constraint as would be for the contact technique. However, for components with even a medium degree of complexity, a higher number of probes would require an individual mechanic positioning and orientation in the case of even slightly curved surfaces, which makes the system more complicated and difficult to balance in terms of obtaining a unique sensitivity. The alternative is to employ phased-array probes, which can adapt to the contour by the capability of beam control.

The contact technique offers the largest range of application depending on the component geometry. This technique, in a one- or two-probe arrangement, can be applied as alternative to the squirter technique in the case of complex geometries. The best performance of such system is achieved in combination with a robotic approach for scanning.

However, the technique also allows for the integration into a multisensor array adapting to contour variations in a smaller range by individual gimbal holder arrangement, which on the other hand requires attention to the precise z-(radial) position.

Before a more detailed consideration of the potential range of application of this system, this technique shall be presented in more detail.

Capabilities of the contact technique

Fig 4: Vertical Stabilizer on the Test Bench.

The first system utilizing the contact technique has been designed to examine large CFRP components like the shell of the vertical stabilizer (fig. 2). The dimensions of such component are typically in the range of up to 20 m in height, and from 2 to 4 m in width.

It is obvious that, in the interest of an at least reasonable time frame for the examination of such component, inspection systems need to have a large index pattern realized by a wide footprint of the system with sufficient beam overlap in the index direction. The size of defect to be detected is a 6 x 6 mm² square reflector in the orientation of a delamination, i.e. parallel to the layers in the structure. This criterion means that a defect should be hit at least twice, resulting in a 3 mm index rastering pattern.

Conceptual considerations

For this application, the comparison of the capabilities and the potential of the three techniques clearly spoke in favor of the contact technique approach. The probe system combines high frequency conventional probes for probe-near/ surface-near examination in pulse-echo technique as well as the continuos monitoring of the through-transmission by the back-wall reflection in the case of smaller wall thickness. Pulse-echo-technique and the monitoring of the backwall reflection for larger wall thickness as well as backwall reflection in the case of nonparallel contour is the task for the low frequency phased-array probes.

Fig 5: Basic Methodology Concept.

The intent of the use of phased-array probes for the monitoring of the backwall echo is the following. As soon as the backwall contour is nonparallel, the beam of a conventional straight beam probe is deflected at the backwall and no or a minimal backwall signal is obtained. In this case the lack of backwall reflection does no more allow the interpretation of backwall loss as a delamination. The phased-array probe allows to vary the beam angle and therefore to obtain a backwall signal even in nonparallel sections. The backwall signal would only be missing if a delamination is interrupting the beam propagation.

In a study jointly performed by Airbus and intelligeNDT Systems and Services, the basic parameters of such system have been determined and the UT-system in its basic features has been designed.

This design of the probe system was performed in terms of optimizing the number of conventional and the number of phased-array probes. The following criteria were taken into account:

• the required indexing and
• data sampling,
• rastering pattern optimization,
• probe holder design,
• manipulator design,
• inspection time and
• cost-time benefit considerations

Subsequently, a research program, partly sponsored by the Federal Ministry for Education and Research has been launched.

The first purpose was to lead the multisensoric approach through another phase of optimization in terms of parameters like frequency, damping (spectrum), transducer sizes, beam shape in the interest of beam overlap. The target was to achieve the maximum resolution in order to be able to detect delamination of the first layers counting from both surfaces.

The second purpose was to establish the design principles for a multiple multichannel and multipurpose UT equipment (M³-UT-System) Saphir^plus including the optimized modular software package supporting the full-scale examination of components like the Airbus 380 vertical stabilizer.

Optimization of the multisensoric approach

These activities were mainly relying on an iterative process including modelling, experimental investigations on testblock fully representative for original components with real or realistic defects and the design of new probes as prototypes.

The original concept of combining the advantages of two different probe types for different wall thickness layers was kept.

Subsystem with conventional probes
After the optimization of the probe parameters including wedge design for optimum resolution and commensurate sound attenuation proven by experimental investigations, these probes had to be combined into an array in order to provide the required beam overlap.

For this reason, the probes were arranged within a lateral distance of ~ 3 mm which means a slightly distorted triangular pitch of the probe positions. Each probe array with such arrangement consists of 8 probes. In the interest of an even larger system footprint, a total of 12 arrays were combined which means a total lateral coverage of the system of 280 mm.

From this probe arrangement, the main requirement for the UT system is the number of channels for the conventional probes, which has to be 96.

Subsystem with phased-array probes
The use of the phased-array system is restricted to the lower section of the vertical stabilizer, which - with the larger wall thickness - constitutes a smaller part of the entire component.

For this reason, the reduction of scanning time due to a very large lateral coverage is of less importance.

Therefore and in view of the larger number of channels required for each phased- array probe, the number of 10 phased-array probes was selected.

The required lateral beam overlap could be maintained by an indexing of the activation of a subset of elements within each phased-array probe.

The total lateral coverage amounts to 80 mm.

Fig 6: Multisensor Examination Module.

Combination into multisensor arrangement
These subsystems had to be combined into one system, which operates the conventional probes in the section with smaller wall thickness. In these areas, the conventional probes have two tasks: pulse-echo testing and backwall monitoring. In the section with the larger wall thickness, the conventional system is operating in pulse-echo mode only, the phased-array system works in pulse-echo mode in the larger wall thickness range as well and is monitoring the backwall echo continuously. In this section, both subsystems are operating simultaneously.

M³-UT system Saphir^plus
The main system features of the multiple-multichannel and multipurpose UT equipment (M³-UT-System) Saphir^plus are:

Fig 7: Saphirplus M3 -UT System.

• Modular architecture up to 448 channels
• Combination and interaction of phased-array and conventional probes in one system
• Various data reduction modes: RF, Pixeling, ALOK, Gates
• Precise beam pattern control (vertical, horizontal or combined skew)
• Equipment parameters for most challenging requirements
• Extensive self-check
• Automated fast self calibration of all equipment functions in accordance with national standards
• Extensive data base of individual probe and probe system data for easy set-up, system planning and operation
• Real-time scanning for only one scanning movement without rastering (no meander)
• Different access levels to the system according to the individual qualification both for data acquisition and analysis

Specific software features

Within the large range of the modular Saphir^plus software package some modules which are of specific assistance for the fast and reliable assessment of the CFRP- components integrity shall be named:

The so-called D-scan is principally a C-scan with the measured wall thickness represented by color code.

The routine combined top, side and end view display with zoom function and A- scan as identified by cursor positioning was slightly adapted for the specific conditions of CFRP-examination, like separated backwall display (otherwise the C- scan would be completely blotched). The capability of the software to combine a large number of channels into one display helps to generate a clear uninterrupted overview over a larger section of the component as if it had been raster-scanned.

Experience highlights

Among the large experience achieved with the system since the start of its operation in 2001, the most significant results were obtained in an internal Round- Robin Test using a section of a CFRP component. Aside from intentionally inserted defects representing delamination in all kinds of occurrence, porosities were inserted additionally. The reliable detection of these porosities was the decisive criterion for the level of performance of the different systems, specially with regard to minor composite porosities and even more or less singular porosities.

Fig 8: Results of Squirter system: Porosities remain undetected.

Fig 9a: Results of Multisensor Multichannel System in same area: With switched off TGC the near-surface delaminations and the delaminations in the volume can be distinguished because of the different amplitude/color.

Fig 9b: Results of Multisensor Multichannel System: The pulse-echo technique enables the detection of porosities in the upper area of the CFRP testpiece between the 1st and 2nd layer.

The test piece was examined by squirter system and by the multisensor multichannel system.

The results, as displayed in fig. 8 and fig.9, clearly demonstrate that the detection of the delaminations did not pose any difficulty for both systems. However, as is self-evident, the through-transmission technique as used by the squirter technique, cannot give any information about the through-wall position of the defects. This information can easily be obtained by the pulse-echo technique integrated into the contact multichannel system.

The contours of the defects are definitely more clearly depicted by the contact technique, which can easily be derived from the finer index used. A defect contouring could be useful, if defect characterization is desirable in the interest of optimizing the fabrication process.

Finally and very important, the porosities do not show up clearly in the display of the results of the squirter technique, whereas they can easily be recognized in the display of the multichannel contact technique. This is a relevant difference in the systems’ capabilities, as porosities, even minor composite ones have gained considerable interest in the structural integrity assessment of airplane CFRP components.

The high signal to noise ratio as obtained with the multichannel contact technique has been confirmed by the large and growing experience gained in the in-process examination of airplane CFRP components.

Conclusion

Based on the stringent requirements for an ultrasonic system in terms of timely performance and high sensitivity and defect detection reliability a system has been generated, which is seemingly complex and bulky. However, this system helped to reduce the inspection time by a factor of up to 10 related to systems used up till now. Therefore, the higher cost of such system is compensated by its time-saving performance and reliability within a very short period.

References

1. G. Engl, F. Mohr, Siemens NDT, Erlangen, Germany and A. Erhard, BAM, Berlin, Germany; “The Impact of Implementation of Phased-array Technology into the Industrial NDE Market”, 2^nd International Conference on NDE in Relation to Structural Integrity for Nuclear and Pressurized Components, New Orleans, May 2000

© NDT.net - info@ndt.net

|Top|