A Clever Way to Monitor Welding Processes

Steinbeis experts and project partners develop a multi-module sensor system

An efficient process for joining metal parts, arc welding continues to play an important role in the manufacture of structural components and assembled parts. There are a number of reasons for this, including the continuous productivity and quality improvements that have been made in process and plant technology, but also ongoing developments in process automation thanks to the introduction of industrial robots. Working alongside its project partners, the Steinbeis Innovation Center for Intelligent Functional Materials, Welding and Joining Techniques, Implementation has developed a new multi-module sensor system that makes it possible to regulate and monitor arc welding processes inline during the production of high-quality welded seams.

This slideshow requires JavaScript.

The increasing use of automation is amplifying the importance of welding defect-free joints in reproducible quality. The quality and reproducibility of seams, layers, and component design are strongly influenced by a number of factors, including the types of materials involved, joint geometry, component temperature, and welding heat energy. Any changes in these factors require continuous adjustments in welding process parameters (current, voltage, speed, wire feeds, process gas quantities, and the distance between the torch and components), and complex interactions and interdependencies between these parameters affect the quality of welds. Often, it is only possible to make adjustments by going through a highly complex process of correcting parameters – or relying on experienced specialists.

Four partners – one goal

To simplify this process and avoid weld defects, it was decided to develop a multi-module sensor system that would make it possible to regulate and monitor arc welding processes inline. The task was taken on under an R&D project conducted by Intelligent Functional Materials, Welding and Joining Techniques, Implementation, the Steinbeis Innovation Center from Chemnitz, in collaboration with three project partners: welding machine specialist Weber, mold maker Frank, and the professorial chair of measuring and sensor technology at TU Chemnitz. As Steinbeis entrepreneur PD Dr.-Ing. habil. Khaled Alaluss also highlights, improving the weldability of materials that are typically difficult to weld was a crucial goal from, among other things, a scientific and commercial perspective: “Finding a solution to this fuels market demand, not just for the welding and hybrid application of such materials – due to the advantages they offer in terms of joining technique and welding technology – but especially when it comes to the lightweight design of electric vehicles, instrument engineering, and systems technology.”

To develop and set up the sensor measurement system, the experts developed a process engineering concept as a basis for taking contactless temperature measurements in order to monitor welding processes using eddy current sensors. They also developed an enhanced communication protocol to make it possible to combine multiple subsystems and form more extensive sensor matrices. The measurement system consists of a transmitter plus two receiver coils for determining component temperatures. It is suited to frequencies ranging from 10 to 250 kilohertz. Given the measurement potential of the sensor technology, temperatures are measured around the seam edge, the heat-affected zone (HAZ), and the base material of the component. The algorithm, which was developed by the project team to use a preconfigured database to calculate temperatures and distances, produces accurate temperature readings at temperatures of 750°C or below, with a defined accuracy of distance measurements ranging from 4 to 8 millimeters.

Step by step from concept to prototype

To integrate the sensor measurement system into the welding torch technology and its corresponding system, in both technical and structural terms, the experts developed a module unit comprising a torch component for shielding arc radiation and heat radiating from components. This consists of holding and adjustment devices on the torch neck, plus a cooling and heating unit. In addition to allowing for sufficient flexibility regarding the position of equipment and ensuring sensors are effective – especially at higher welding energy outputs, thanks to efficient cooling – this also keeps operating temperatures stable at 80°C and below.

To create a reliably functioning and intelligent monitoring system, the experts also determined the exact positioning of the sensor system as the welding torch moves forward. A simulation experiment was conducted to gauge and assess the positions, functionality, and effectiveness of sensors. The prototype of the multi-module sensor system was then used to capture reproducible temperature measurements on welded component joints: It was possible to measure temperatures of up to 750°C along the seam edge, the HAZ, and the base metal of component geometries welded together from different materials (steel, titanium, aluminum).

To develop AI for the system, the experts created and subsequently implemented a concept for a control and regulation unit based on an inline monitoring system that draws on a database to set parameters for the welding process. This control unit was connected to the welding power source, the wire feeding unit, and the welding robot system to allow the multi-modular sensor system to detect component heating and cooling processes when using different materials and welding seam joints on different joint geometries. This made it possible to monitor and regulate the automated arc welding process for defined tasks involving acceptable process parameters.

The final step was to validate the sensor system prototype. To do this, welding and materials testing was carried out on different joint geometries using different materials in order to assess the operational behavior of the system and functional modules. This involved capturing raw sensor data and temperatures by looking at variable frequency intervals (10 to 250 kilohertz) during adjustment of the process parameters and sensor positions. This confirmed the smooth functioning of the system, the good handling of robotic elements, and good accessibility during the welding of different seam joints. Finally, it also demonstrated the practical nature of the multi-module sensor system for regulating and monitoring arc welding processes inline during the production of high-quality seam joints.

This project – Development of an Intelligent Multi-Module Sensor System for Inline Regulation and Monitoring of the Arc Welding Process in the Production of Quality Seam Joints – is funded by the Federal Ministry for Economic Affairs and Climate Action following a resolution passed by the German Bundestag.