Fraunhofer IKTS

Ultrasonic testing is a quality assurance process for volume testing of samples of various geometries and materials. It is often used to inspect joints such as welds. The characterization of layer thicknesses of up to 100 µm is also possible with this method.

The Accredited NDT Center of Fraunhofer IKTS offers ultrasonic testing with the following services:

• Manual and mechanized ultrasonic testing using immersion technology, contact technology, pulse-echo technology, transmission technology and, upon agreement, other techniques
• Use of different ultrasonic probes (vertical, S/E and angle, submersible and phased array probes) with different frequencies and other parameters
• Ultrasonic scan of flat, curved and also rotationally symmetrical samples of various materials with image reconstruction (realizable object parameters: weight < 100 kg, geometry < 500 x 300 x 300 mm)
• Detection of volume defects with material and process-dependent resolution of Ø ≥ 100 µm
• Development, optimization and testing of new ultrasound processes including access to simulation calculations
• Development of test concepts and creation of test instructions
• Validation of new ultrasonic testing methods, including those developed by the customer

The eddy current method is a fast electromagnetic testing method requiring little equipment. It can be used to check for defects and geometry variations as well as to determine layer thickness. One focus is on applications using high-frequency eddy current technology, which are mainly used to characterise low-conductivity materials such as superalloys and composites.

The Accredited NDT Center of Fraunhofer IKTS offers eddy current testing with the following services:

• Manual eddy current testing on components of different geometries
• Eddy current scans, also with very high spatial resolution (around 200 µm), on flat components
• Material characterization and testing for surface defects
• Determination of conductivity after calibration as well as layer thickness and wall thickness
• Determination of the drying behaviour of thin functional layers
• Investigations with various sensors (absolute and differential sensors), measuring frequencies up to 100 MHz
• Development of test concepts and creation of test instructions
• Validation of new eddy current testing methods, including those developed by the customer

Sound emission analysis is one of the acoustic non-destructive testing methods in which individual sound emissions are recorded and examined. This takes advantage of the fact that sudden changes in the structure of a material, which can be triggered by various loads (mechanical, thermal, chemical), emit such sound events. The position of the component damage can be determined from the transit time of the signals from the sound-emitting source to a large number of sensors. In composite materials, for example, the underlying damage mechanism such as fibre cracking, fibre-matrix cracking or delamination can be determined by analyzing the signals recorded.

The Accredited NDT Center of Fraunhofer IKTS offers noise emission testing with the following services:

• Sound emission testing on components of different geometries and materials
• Investigation of the material behaviour under mechanical stress on coupon samples
• Development of test concepts and adapted testing and monitoring systems for customer-specific applications of sound emissions
• Validation of the use of the procedure to support the release of the procedure to the operator

Visual inspection can be used to examine surfaces for visible defects. The process also includes checking for changed colours, which indicate defects in the structure, for example in weld seams.

The Accredited NDT Center of Fraunhofer IKTS offers visual inspection with the following services:

• Check for visible defects such as surface cracks, pores, discolouration
• Testing using endoscopy in places that are difficult to access (e.g. in pipes) and on components with complex geometry
• Optical enlargements

Methods and measurement systems for condition monitoring (CM) ensure the availability of machines and systems. The regular recording of machine states can provide crucial information, especially for maintenance strategies (Maintenance, Repair and Overhaul, MRO). This enables better control and optimization of the processes. In structural health monitoring (SHM), however, static components are monitored automatically. The resulting recording and evaluation of damage mechanisms ensures the reliability and optimal functioning of critical components.

Increase Reliability, Reduce Costs

Fraunhofer IKTS is a competent contact when it comes to monitoring materials, components, assemblies, machines, systems and the associated manufacturing processes. This includes both the analysis and evaluation as well as the optimization of components and processes.

For this purpose, IKTS develops sensors (optical, acoustic, electromagnetic), systems and procedures for condition monitoring, which are implemented during the construction and commissioning of systems. Retrofitting into existing infrastructure is also possible. Designed for harsh operating conditions, they implement non-destructive defect detection, real-time monitoring and the associated condition-based maintenance or system optimization.

The tailor-made solutions contribute to improving system safety while observing all standard requirements and are created in close collaboration and development work with manufacturers and users.

Complex testing tasks in the context of Industry 4.0 provide a huge amount of process data of different origins that must be analyzed. With suitable approaches, a wealth of additional information can be obtained from existing or specifically recorded parameters. Fraunhofer IKTS optimizes established and new NDT processes to show customers specific ways to use this valuable information.

Pattern Recognition

With pattern recognition, test objects can be classified based on their measurement signals, e.g. B. actively or passively obtained acoustic signals, images or other parameters such as temperature values. The focus of pattern recognition is to assign meaning to this complex data, e.g. E.g. “The gear is free of errors” or “The valve has reached 80% of its service life”. Fraunhofer IKTS has had extensive experience in the field of pattern recognition for many years. The algorithms developed have already been successfully tested and used in a variety of applications, including in mechanical engineering as well as in the automobile, glass, paper, textile and watch industries. In addition to the PC-based solution, a self-sufficient and modular device for mobile measurements was developed to which various sensors or microphones can be connected.

Technical details:

• Independent of the test method, measuring principle and type of sensors used
  
• Combination of different sensor data possible

Fields of application:

• Good/bad analyses
  
• Lifespan predictions
• Detection of cracks, inclusions and impacts
• Wear monitoring
• Condition monitoring of components, machines and systems
• Monitoring of production processes

Machine Learning

Machine learning, a branch of artificial intelligence, is used to learn from an existing, usually extensive amount of data. This process does not occur through “memorization,” but rather through recognizing patterns and regularities in known examples, the training data. In the training process, generalized models are formed that can be used to classify new, unseen data. Fraunhofer IKTS uses special machine learning methods for this, such as deep learning for training deep neural networks (DNN), expectation maximization algorithm (EM) for hidden Markov models (HMM) or convex optimization for support vector machines (SVM). Special training software makes it easy to learn new models, for example for further series of the same component or comparable components.

Range Of Services:

• Recognition and training software
  
• Hardware modules
• Data analysis and evaluations
• Customer-specific development of complete systems

NDT Assistance Systems

Another focus in the context of Industry 4.0 is assistance systems that are intended to support people in using technology. For this purpose, Fraunhofer IKTS is developing a cognitive user interface for controlling test systems. This enables a natural language dialogue with the testing system. This means that the examiner does not need any prior knowledge nor does he have to learn commands. It adapts independently to the examiner's way of working and to the inspection tasks. In addition, it learns individual user behaviour and can be controlled via various communication options (e.g. voice). This makes it easier for testers to operate the test device when access to the test specimen is difficult or under complicated environmental conditions (e.g. contaminated environments).

The cognitive user interface developed at Fraunhofer IKTS has the advantage of being self-sufficient. It requires neither an internet connection nor a radio network. In addition, the hardware module does not use any resources of the test device. The data is only kept on the device and is not transferred to external servers or a third-party cloud, which ensures absolute data security. This makes it also suitable for trustworthy and local applications with sensitive data.

Technical Details:

• Can be used without a wireless network or internet connection
  
• No transmission of user voice input to third-party servers
• Enables contactless (“hands and eyes free”) communication

Fields of application:

•  Maintenance, Repair and Operations (MRO) in:
• Large-scale technical infrastructures
• aviation
• Industrial and plant engineering
• Human-machine interaction
• Control of devices and systems 

Range of services:

• Recognition software
• Hardware
• Training software
• Customer-specific developments