Lightweight Design worldwide

, Volume 11, Issue 1, pp 48–53 | Cite as

Eddy current testing in CFRP production

  • Georg Bardl
  • Richard Kupke
  • Henning Heuer
  • Chokri Cherif
Production Quality Assurance

Eddy current testing has established itself as a nondestructive test method in CFRP production. By exploiting the electrical conductivity of carbon fibers, it is possible to detect fiber orientation, surface weight, area density and fiber volume content as well as ripples and folds in stacks, preforms and composite components. This article presents recent applications and discusses future perspectives for deployment in CFRP production.

Forming Multi-layer Semifinished Products

Nondestructive testing of multi-layer stacks, preforms and composite components for correct fiber orientation and freedom from defects is a recurrent task in CFRP production. The shearing of textile materials that results when flat semifinished products are draped into multi-curved 3-D geometries, changes local fiber orientation and hence the rigidity of the composite components. What is more, areas with extreme shearing also increase fiber volume content and are prone to the formation of folds and dry spots that need to be eliminated through suitable process control. Fiber orientation must be known in particular for calculating rigidity and deformation.

Optical methods allow fiber orientation to be inspected quickly and precisely for single-layer fabrics or the uppermost layer of multi-layer stacks. However, one of the major challenges in process development is that the forming behavior of multi-layer stacks such as textile sheet materials, prepregs or organic sheets is essentially different from the forming behavior of single-layer semifinished products. Owing to the complex friction and glide conditions between the individual layers, defects occur when forming stacks that are not seen when forming single-layer semifinished products. Similarly, shearing and fiber orientation in the individual layers can differ considerably when compared with single-layer forming [1]. A lack of information about the paths of the fibers in the inner layers is therefore generally reflected in additional safety factors and hence in increased component weight and material costs.

Exploiting Electrical Conductivity

Eddy current testing has proved to be an alternative method of testing for carbon- fiber materials and is currently deployed in a large number of processes in carbon-fiber processing. The basis for eddy current testing is the electrical conductivity of the individual carbon fibers, which results in the formation of conductive networks in the carbon fiber materials. Figure 1 shows the principle of eddy current testing. A sensor uses a transmitter coil to induce a high-frequency, alternating magnetic field — the so-called primary magnetic field — which induces eddy currents in electrically conductive bodies. These in turn generate a secondary magnetic field, which counteracts the inducting field and which is registered by a receiver coil. The higher the local conductivity within the range of the sensor, the stronger the induced eddy currents and the stronger the magnetic field that is registered by the receiver coil. In the simplest setup, the transmitter and receiver coils are identical; to provide higher sensitivity, separate sending and receiving coils are usually used.
Figure 1

Principle of eddy current testing (© TU Dresden)

Scanning the component surface with a sensor — similar to an ultrasonic examination — provides an image of local conductivity. Figure 2 shows the eddy current result from inspecting a (0°/90°) biaxial non-crimp fabric with a missing carbon fiber yarn. The locally reduced conductivity caused by the missing thread can also be clearly seen in the seventh layer from the top.
Figure 2

Eddy current image of a non-crimp fabric stack with missing carbon fiber bundle: The missing fiber bundle can be clearly seen in the 7th layer (© TU Dresden)

Inline and 3-D Testing

The fundamental difference between eddy current testing for carbon fiber materials and eddy current testing for metals is the higher testing frequency. Frequencies in the range of between 1 and 10 MHz must be used in order to generate a sufficiently high density of eddy currents in weakly conductive carbon fibers — and these require specialist measurement technology. Initial investigations into eddy current testing for CFRPs were conducted in the 1970s [2]; high-resolution testing and evaluation systems for examining carbon fiber materials have been commercially available since 2010 [3]. Since eddy current testing allows very high testing speeds (it is used at speeds of up to 150 m/s in metal processing), it can be integrated as inline testing into many production line processes. Applications include the production of non-crimp fabrics, weaves, carbon fiber nonwovens as well as continuous processes such as pultrusion, prepreg production or the monitoring of spread fiber tows for constant fiber quality. Figure 3 shows a sample system for the nondestructive monitoring of carbon-fiber nonwoven production. Four sensors contactlessly and nondestructively determine the surface weight at four different positions across the width of the fabric.
Figure 3

Inline testing of the surface weight of carbon fiber nonwovens with four eddy current sensors (Suragus GmbH)

The difference to eddy current testing on metals is the higher testing frequency.

For the testing of double-curved, three-dimensional components, an industrial robot is used for guiding the sensor. Figure 4 shows, on the left, a testing cell developed at Fraunhofer IKTS Dresden, where digitization, path planning and eddy current testing is performed fully automatically for 3-D CFRP components. On the right, a testing system at the ITM is depicted, where an eddy current testing device from IKTS was integrated into a Kuka small-scale industrial robot. For scanning complex 3-D components, a dedicated path planning software was developed.
Figure 4

Testing cell for automatic surface capture, path planning and eddy current testing (left) (© Fraunhofer IKTS) and industrial robot with integrated eddy current testing system (right) (© TU Dresden)

Automatic Detection of Thread Paths

The nondestructive eddy current testing of 3-D components opens up completely new perspectives, as it enables the analysis of fiber orientation in multi-layer preforms and CFRP components. Figure 5 shows the eddy current image of a four-layer CFRP component with layer orientation (+45°/-45°/0°/90°). The superposition of the eddy current signal from the individual layers results in a striped pattern in the eddy current image. Since the conductivity at the edges and boundaries of carbon fiber yarns is lower, the individual rovings appear as a periodic change in conductivity, allowing fiber orientation to be identified.
Figure 5

Eddy current image of a four-layer CFRP component (© TU Dresden)

In order to read the local orientation of the carbon fiber yarns from the 3-D eddy current image, a software was developed at ITM that automatically determines the local orientation for the individual layers and reconstructs the paths of the individual yarns. Figure 6 shows the reconstructed thread paths in the four layers. With optical methods, it would only be possible to identify the uppermost layer.
Figure 6

Detected yarn orientations for the four layers (© TU Dresden)

Software automatically reconstructs the local orientation of the individual layers.

From the detected yarn orientations, the shear angle of the textile fabric can be determined. Two different types of non-crimp fabric were used to produce the component displayed: a bidiagonal ±45° non-crimp fabric and a biaxial (0°/90°) non-crimp fabric, for both of which the measured shear angles are shown in Figure 7. The different shear behavior of the two fabrics can be clearly seen, and the greater shear angles on the component’s vertical side areas indicates regions with higher fiber volume content.
Figure 7

Measured shear angle (© TU Dresden)

Eddy current testing thus also allows cost-effective and rapid validation of the fiber path calculated using draping simulation. Figure 8 shows, on the left, a comparison of the two bottommost layers with a draping simulation. The yarn path in the real component deviates from the simulated yarn path, in particular in the areas around the edges where there is relative movement between the layers during draping. While the red yarns (measurement) exactly follow the gray yarn patterns that can be seen in the eddy current image, the blue yarns (simulation) are slightly askew in some areas. The measured actual yarn paths can therefore be used for a more exact design.
Figure 8

Comparison between draping simulation and measurement (left), automatic parameterization of the FE mesh (center) and FE simulation with actual thread paths at a compressive force of 140 N on the inner ring (right) (© TU Dresden)

Figure 8 (center) shows an example of the automatic import of measured yarn directions in a finite element (FE) model for the second layer. On the right, the result of an FE simulation is depicted.


Eddy current testing is currently starting to be deployed in automotive series production, and online and offline eddy current testing systems are in use in a large number of CFRP production processes. This has led to an improved understanding of eddy current distribution in carbon-fiber materials, and will predictably result in the development of specialized sensors with greater resolution and measurement depth, adapted to the fiber structure and anisotropic conductivity of carbon fiber materials. Typical detection depths are currently between 8 and 12 carbon-fiber layers. A recent paper presented sensors that can detect small fiber waves with an amplitude of 3 mm up to the 18th layer [5]. Another foreseeable trend will be the development of sensor arrays to speed up the testing process similar to those familiar from ultrasonic testing. A challenge that still needs to be tackled in this respect is the implementation of flexible, three-dimensional testing heads that adapt to the surface geometry in order to prevent sensor liftoff for curved component.

The method supplies information that was previously difficult to access.

Furthermore, the range of detectable characteristics in CFRPs is far from being exhausted. All effects that have a direct or indirect influence on the conductivity of the fibers, or on the electrical contact between the fibers, fiber bundles and layers, could potentially be detected with eddy current testing. It has already been shown that local burns — so-called hot spots — can be detected [6], and also that aging and load history can be determined in the eddy current signal [7]. At the same time, it is expected that the use of eddy current testing for process monitoring and quality assurance will further grow. The possibility of being able to detect fiber orientation and defects in the multi-layer structures of three-dimensional preforms and CFRP components in a nondestructive way provides a method that can supply information for design, process development and quality assurance that was previously difficult to access and will conceivably raise the quality of CFRP structures while continuing to reduce raw material cost.



The work on 3-D fiber orientation detection described here is part of the “3-D Fast” project that is funded by the European Regional Development Fund (ERDF) and the Free State of Saxony (funding code 100224749). The authors would like to thank the aforementioned institutions for the provision of financial resources. Parts of the findings represented here are based on work performed by Martin Schulze and Matthias Pooch (Fraunhofer IKTS) and Matthias Hübner and Andreas Nocke (TU Dresden), whom the authors wish to thank for kindly making them available.


  1. [1]
    Nezami, F. N.: Automatisiertes Preforming von Kohlefaserhalbzeugen mit aktiven Materialführungssystemen zur Herstellung komplexer Faserverbundstrukturen. (Automated preforming of carbon-fiber semifinished products with active material guidance systems for the production of complex fiber composite structures.) Dissertation, TU Dresden, Faculty of Mechanical Science and engineering, 2015Google Scholar
  2. [2]
    Owston, C. N.: Eddy current methods for the examination of carbon-fibre-reinforced epoxy resins. In: Materials Evaluation 34 (1976), pp. 37–244Google Scholar
  3. [3]
    Schulze, M. H.; Heuer, H.; Küttner, M.; Meyendorf, N.: High-resolution eddy current sensor system for quality assessment of carbon-fiber materials. In: Microsystem Technologies 16 (2010), No. 5, pp. 791–797CrossRefGoogle Scholar
  4. [4]
    Bardl, G.; Nocke, A.; Cherif, C.; Pooch, M.; Schulze, M.; Heuer, H.; Schiller, M.; Kupke, R.; Klein, M.: Automated detection of yarn orientation in 3-D-draped carbon-fiber fabrics and preforms from eddy current data. In: Composites Part B: Engineering 96 (2016), pp. 312–324CrossRefGoogle Scholar
  5. [5]
    Mizukami, K.; Mizutani, Y.; Todoroki, A.; Suzuki, Y.: Detection of in-plane and out-of-plane fiber waviness in unidirectional carbon-fiber-reinforced composites using eddy current testing. In: Composites Part B: Engineering 86 (2016), pp. 84–94CrossRefGoogle Scholar
  6. [6]
    Heuer, H.; Schulze, M.; Pooch, M.; Gäbler, S.; et al.: Review on quality assurance along the CFRP value chain — Non-destructive testing of fabrics, preforms and CFRP by HF radio wave techniques. In: Composites Part B: Engineering 77 (2015), pp. 494–501CrossRefGoogle Scholar
  7. [7]
    Gäbler, S.: Untersuchung des Anwendungspotenzials der Hochfrequenzwirbelstrommesstechnik zur Charakterisierung dielektrischer Eigenschaften von Epoxidharzen und Faserverbundmaterialien. Dissertation, TU Dresden, Faculty of Electrical EngineeringGoogle Scholar

Copyright information

© Springer Fachmedien Wiesbaden 2018

Authors and Affiliations

  • Georg Bardl
    • 1
  • Richard Kupke
    • 2
  • Henning Heuer
    • 3
  • Chokri Cherif
    • 1
  1. 1.TU DresdenGermany
  2. 2.Suragus GmbHDresdenGermany
  3. 3.Fraunhofer Institute for Ceramic Technologies and Systems (IKTS)DresdenGermany

Personalised recommendations