Lightweight Design worldwide

, Volume 10, Issue 6, pp 28–33 | Cite as

Active Thermography for Automated Testing of Composite Structural Components

  • Markus Thurmeier
  • Alexander Stock
  • Michael Fischlschweiger
Production Component

While the use of active thermography for non-destructive testing supplies valuable information on component quality, the measurement results are still very often evaluated manually. Ottronic now presents a method that automatically classifies and interprets measured data. Active thermography is thus paving the way for intelligent, self-optimised production lines.

The new generation of Audi’s A8 model, Figure 1, is a recent example of the use of a wide range of materials in a vehicle’s structure. Multi-material design is the logical solution for taking account of the different functions and loads of individual components in the best possible way. Lightweight design plays an important role in increasing the energy efficiency of the entire vehicle. Besides conventional measures for optimising the weight of components when viewed in isolation, particularly systematic approaches help to reduce material and hence cost and weight.
Figure 1

The new Audi A8 (© Audi)

Lightweight Design Requires a Systematic Approach

In combination with the rest of the body, Figure 2, the rear wall of the A8 forms a torsion ring with a significant effect on the vehicle stiffness, which is of particular benefit to the long version of the model.
Figure 2

CFRP rear wall in the new Audi A8 (© Audi)

The rear wall consists of carbon fibre reinforced polymer (CFRP). The arrangement of the highly rigid fibres and the number and arrangement of the layers from which the rear wall is built up has been optimised in terms of load on the component. Figure 3 shows how the structure and the topology of the component takes into account and models the various load paths. Areas exposed to high loads have thicker walls. The optimised use of materials saves costs and weight at points where lower forces need to be transferred.
Figure 3

Load-appropriate structure of CFRP layers in the rear wall of the new Audi A8 (© Audi)

Mechanics Determine Quality

The special mechanical properties of CFRPs and other FRCs originate in the interaction between fibres and the polymer in which they are embedded. The layered structure of FRC components allows the flexible design of stiffness properties that can be highly anisotropic and arranged differently depending on location [1, 2, 3]. However, these advantages over more homogeneous materials depend on the quality of the composite structure. Deviations from the planned orientation of the fibres, dry areas resulting from missing polymer, delamination between the individual layers and variations in the fibre content are only a few examples of potential quality variations that can have drastic effects on the mechanical properties of the component. Many quality attributes are determined by the production process [4]. For this reason, a large number of quality-determining measures accompany the formation of FRC components — from development through to production. Non-destructive testing in particular is gaining in importance in this area.

Active Thermography for Extensive Testing

Modern testing methods, such as active thermography, supply comprehensive quality information on components like the CFRP rear wall. Active thermography is a test method in which the elements to be tested are subjected to brief thermal variations using sources of heat. Areas in the test object with different thermophysical properties will also have an effect on the way heat is transferred locally. These differences are picked up by corresponding infrared (IR) detectors and visualised with the help of appropriate algorithms. The resulting images show variations in the thermal properties of the test objects and can be interpreted and evaluated by the operator on the basis of experience and correlation.

Comprehensive studies on a range of quality attributes and material combinations of FRC show that the method is not only suitable for testing the quality of FRCs but is actually sometimes preferable to other non-destructive testing methods [5, 6, 7, 8]. While economic considerations support this view, the method has other advantages that can be highlighted. First, active thermography is based on measuring the thermophysical properties of test objects. It is precisely these properties that are influenced by the quality attributes and variations that are sought. Second, the method does not emit any ionising radiation.

Testing Large Aircraft Parts

The fact that aircraft manufacturer Boeing has authorised its supplier FACC to use active thermography as a test method — showing that the aerospace industry is adopting this method on an industrial scale in parallel with the automotive industry — can be attributed to advanced developments in sensors, data interfaces and electronics [9]. An advantage of active thermography is the speed at which large-format test objects can be measured. This is the reason why the rapid development of infrared cameras (greater temporal, geometric and thermal resolution) and the inexpensive availability of large computing capacities are contributing to the deployment of commercial applications in a range of industries.

Divergent fibre orientation, dry areas or delamination can have drastic effects on the mechanical properties of the component.

Optimised Test Results for Visual Assessment

The high level of sensitivity of active thermography to differences in properties supplies test data with high informational content. Analysis methods and algorithms have been developed for evaluating the test results that visualise the data in the form of results images. This step is necessary in order to provide human operators with information that they can comprehend and interpret. The large body of the thermographic data supplied by the IR camera are compressed into individual results images. The compression methods that have been developed are optimised to contain as much of the total information in the data as possible. Algorithms that abstract the original temperature information are ideal for this purpose [10]. The following analogue example from the field of medicine illustrates the interaction between test results and humans. While a doctor uses an X-ray image to interpret the state of health of patients, compressed results images from active thermography provide an experienced test engineer with comprehensive information for interpreting the quality of test objects. Figure 4 shows such a results image. The different shades of grey are an abstract representation of the data recorded by the IR camera. Combining this with knowledge of the component and of the production process ensures the success of the interpretation of the results images. The large-format dark area suggests delamination in the example shown. The smaller dark areas suggest thinner areas that can occur in the production process as a result of localised rubber folds. Such test results provide important information in different phases of product life. In addition to component testing, active thermography is also used successfully for material and semi-finished goods testing as well as for process validation [11].
Figure 4

Typical results image for FRC structures with classic analytical methods of active thermography (© Ottronic)

Differences in thermophysical properties are picked up by infrared detectors and visualised.

Self-optimisation of Production Chains

The character of the testing method enables it to be used for other industrial purposes. This non-invasive and non-destructive approach that can measure various test areas quickly and flexibly makes it ideal for interim and final testing in manufacturing processes. Making the test results available to the processing systems for intervention in the manufacturing process creates a closed control loop consisting of the manufacturing process and quality testing, and this forms the basis for self-optimising production chains.

For this purpose, it must be possible to automatically classify the data from active thermography. Every class contains values of the measured parameter that can be linked to a test object attribute. Two classes can be used, for example, to differentiate between expected and divergent value ranges. Further classes allow other variations to be differentiated. While humans are able to classify and evaluate complex visual impressions using their knowledge and experience, automated classification is based on clear criteria and rules. These need to be determined and defined by humans before being “taught” to the machine in a further step.

There are two goals associated with the approach to this task. First, the measurement setup for changing environmental conditions needs to be thermodynamically independent. Second, measurement information is required to which clear and robust criteria and rules can be applied. Ottronic Regeltechnik has introduced successful concepts based on its system for active thermography.

Automated Results Evaluation

Figure 5 shows how systems from Ottronic meet the first goal for controlled, thermodynamic boundary conditions. While humans can recognise changing environmental influences that affect test information as such and can take them into account, changing thermodynamic conditions make automated classification difficult. The closed-loop testing system prevents the influence of rapid and unintentional changes in temperature through external sources of thermal radiation. An additional advantage of the closed-loop testing room is the safe operation of thermal power sources and integrated robotic systems.
Figure 5

System ATIIS from Ottronic Regeltechnik GmbH for active thermography in a closed test room (© Ottronic)

Thermodynamics as the Basis

The second goal is concerned with finding a thermodynamic measurement parameter that allows simple assignment of thermophysical properties for every point on the test object. If the required assignment can be based on direct physical correlations, it is possible to trace clear and solid relationships between the value of the measured parameter and the cause of the variation in the thermophysical property.

The first attempts at flashlight-based determination of thermophysical properties were made back in the 1960s [12]. Taking the basic observations in [12] as a starting point, Ottronic has now succeeded in developing a new product for active thermography that uses a new type of automated evaluation technology and software to measure the dimensions of physically based parameters of the test object. One example of such a parameter is thermal diffusivity α. Appropriate sets of rules and operators can be used to identify defects, defect sizes and other crucial quality information about the test object directly. Figure 6 shows a results image with the measurement parameter α where the test object section corresponds to that in Figure 4. The colours represent different values for thermal diffusivity. Values increase across the colour gradient from dark blue to dark red. Unlike Figure 4, the dark-red areas can essentially be differentiated from the dark-blue areas without any knowledge of the test object or experience of production. The low values of the blue areas indicate an obstruction to thermal conduction that can only result from distinct material transitions or cavities. The high values of the dark-red areas can be explained, for example, by extraneous materials with high temperature conductivity or by thin walls. In this case, the identified quality variations are also due to the production process [13, 14]. However, knowledge about production is no longer necessarily required to evaluate the results.
Figure 6

Typical results image for FRC structures with analysis based on physical properties (© Ottronic)

This example shows that it is relatively easy to classify different areas based on physical relationships. This makes it easier to specify rules for automated classification, often without the need for time-consuming empirical studies.

Automatic Evaluation of Quality

Additional rules need to be set up for an interpretation in terms of a good/bad evaluation of the variations classified automatically by the test system. Direct assignment of test object properties to a physical measurement parameter also helps in this case. Following initial classification of the good/bad decision and after the system has “learned” the necessary rules for this, interpretation can be performed automatically following classification.

A closed control loop consisting of the manufacturing process and quality testing permits self-optimising production chains.

Figure 7 visualises automatic classification based on the information presented in Figure 6. The test area has been automatically subdivided into three classes. The green class represents properties lying within the expected boundaries. The blue class points to obstructed — and the red class to faster — thermal transfer. Based on these classes, it is also possible to perform interpretation automatically in terms of a good/bad decision using simple criteria. For example, entire classes can be categorised as bad. Similarly, the size of a contiguous area of the same class — or the number of such areas — form the basis for the good/bad decision.
Figure 7

Visualisation of automatically classified test results (© Ottronic)

The example of the automated interpretation of results can also be used to illustrate how correlation is easier to achieve via the direct relationship of the measured parameter to the properties of the test object than when the measurement results are available in an abstract form, such as differences in contrast and relative variations in information.

This automated classification and interpretation of test information, which has been made possible for the first time with the new test system from Ottronic, represents a milestone for the application of active thermography integrated into the process chain.

Electric Vehicles with Greater Range

Further possible uses for the new system are demonstrated by an application in the field of e-mobility. Energy efficiency plays a very special role for electric-powered vehicles, in particular through its correlation with range. As discussed at the beginning of this article, lightweight design is one of the possibilities to exert a positive influence on this. If powerful batteries are used, heat management in their proximity becomes an important parameter in energy optimisation. Since active cooling consumes large amounts of energy, the ability to dissipate heat via the structure surrounding the batteries plays a vital role. The possibility of using large-format active thermography to determine these structures allows an individual — and hence optimised control loop — to be developed to dissipate heat from the battery. This enables thermal management to be optimised and makes a positive contribution towards the range of the overall vehicle. In addition, taking account of individual and locally different values for α allows highly efficient quality assurance to be performed.

This and other applications in the automotive and aerospace industries show how the combination of established processes and methods and new ways of thinking can turn a robust and simple principle of measurement, such as active thermography, into a versatile testing method with many possible uses. Improved test equipment and new analytical algorithms can contribute to new and better results in many areas, ranging from classic manual test applications to high-end applications in automated intelligent production lines. Applications for optimising weight and in the thermal management of electric-powered vehicles show the wide range of possible applications for active thermography. Directly measured thermophysical parameters as well as information about the mechanical characteristics of components obtained through interpretation contribute to knowledge about the status and quality of individual structures and entire systems and thereby to their optimisation.


  1. [1]
    Rettenwander, T.; Fischlschweiger, M.; Machado, M.; Steinbichler, G.; Major, Z.: Tailored patch placement on a base load carrying laminate: A computational structural optimisation with experimental validation. In: Composite Structures 116 (2014), pp. 48–54CrossRefGoogle Scholar
  2. [2]
    Rettenwander, T.; Fischlschweiger, M.; Steinbichler, G.: Computational structural tailoring of continuous fibre reinforced polymer matrix composites by hybridisation of principal stress and thickness optimisation. In: Composite Structures 108 (2014), pp. 711–719CrossRefGoogle Scholar
  3. [3]
    Machado, M.; Fischlschweiger M.; Major Z.: Strength of single-lap-joint assemblies of continuous unidirectional carbon fibre-reinforced thermoplastic matrix tapes under tensile loading. In: Journal of Composite Materials 49 (2015), pp. 1977–1987CrossRefGoogle Scholar
  4. [4]
    Nagy, J.; Reith, L.; Fischlschweiger, M.; Steinbichler, G.: Influence of fiber orientation and geometry variation on flow phenomena and reactive polymerization of €-caprolactam. In: Chemical Engineering Science 128 (2015), pp. 1–10.CrossRefGoogle Scholar
  5. [5]
    Arndt, R.; Hillemeier, B.; Maierhofer, C.; Rieck, C.; Röllig, M.; Scheel, H.; Walther, A.: Zerstörungsfreie Ortung von Fehlstellen und Inhomogenitäten in Bauteilen mit der Impuls-Thermografie. In: Bautechnik 81 (2004), pp. 786–793CrossRefGoogle Scholar
  6. [6]
    Fischlschweiger, M.: Integrated defect classification in manufacturing of carbon fibre reinforced thermoplastic polymer matrix composites. International Conference on Processing and Manufacturing of Advanced Materials, Graz, 2016Google Scholar
  7. [7]
    Fischlschweiger, M.; Stock, A. Thurmeier, M.: Integrated Defect Classification in Manufacturing of Carbon Fibre Reinforced Thermoplastic Polymer Matrix Composites. In: Materials Science Forum, 879 (2017)Google Scholar
  8. [8]
    Maierhofer, C.; Röllig, M.; Ehrig, K.; Meinel, D.: Validierung der aktiven Thermografie mittels CT zur Charakterisierung von Inhomogenitäten und Fehlstellen in CFK. In: DGZfP-DACH-Jahrestagung Mi.3.C.1. (2015)Google Scholar
  9. [9]
    Schachinger, A.: Prüfkompetenz bei Leichtbauteilen: FACC erhält Boeing Qualifikation für aktive Thermografie. Online: Thermografie, access: 31.08.2017.
  10. [10]
    Zauner, G.; Mayr, G. Hendorfer, G.: Application of wavelet analysis in active thermography for nondestructive testing of CFRP composites. In: Proceedings of SPIE, 6383 (2006)Google Scholar
  11. [11]
    Thurmeier, M.; Fischlschweiger, M.; Stock, A.: Defying Hostile Environments. In: Kunststoffe international 6-7 (2016), pp. 32–34Google Scholar
  12. [12]
    Parker, W. J.; Jenkins, R. J.; Butler, C. P.; Abbott, G. L.: Flash method of determining thermal diffusivity, heat capacity, and thermal conductivity. Journal of applied physics, 32 (1961), pp. 1679–1684CrossRefGoogle Scholar
  13. [13]
    Machado, M.; Fischlschweiger, M.; Major, Z.: A rate-dependent non-orthogonal constitutive model for describing shear behaviour of woven reinforced thermoplastic composites. In: Composites Part A: Applied Science and Manufacturing 80 (2016), pp. 194–203CrossRefGoogle Scholar
  14. [14]
    Machado, M.; Murenu, L.; Fischlschweiger, M.; Major, Z.: Analysis of the thermomechanical shear behaviour of woven-reinforced thermoplastic- matrix composites during forming. Composites Part A: Applied Science and Manufacturing 86 (2016), pp. 39–48CrossRefGoogle Scholar

Copyright information

© Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2017

Authors and Affiliations

  • Markus Thurmeier
    • 1
  • Alexander Stock
    • 2
  • Michael Fischlschweiger
    • 3
  1. 1.Fibre-reinforced Polymers and Advanced Development of BEVsAudi AGIngolstadtGermany
  2. 2.Ottronic Regeltechnik GmbHFohnsdorfAustria
  3. 3.Ottronic Regeltechnik GmbH and Ottronic Technology LaboratoryFohnsdorfAustria

Personalised recommendations