The European project that seeks to optimize the joining of complex steels with greater precision, flexibility and real-time quality control.
The STWIN project entitled “Advanced, non-destructive technologies to improve steel stir welding process through artificial intelligence and smart digital twin” is an international initiative that develops advanced technologies to improve steel stir welding. It is funded by the European Research Fund for Coal and Steel (RFCS) program, no. 101112504.
What problem does STWIN address?
The steel industry, especially in the metal construction and transportation (more specifically automotive) sectors, demands more efficient and reliable joining processes. Friction stir welding (FSW) is a solid-state technique that can join metals without melting them and offers advantages such as being a more robust and sustainable method compared to traditional methods. However, automated application of this technique to complex structures and a wide range of steel grades and thicknesses requires more flexible systems capable of real-time quality assurance.
In addition, the project seeks to address challenges present in the European industry such as:
- Improve the productivity of welding processes.
- Improve working conditions for welders and operators.
- Addressing the shortage of skilled personnel in the steel industry
Project objective
STWIN aims to develop a flexible friction stir welding (FSW) system capable of:
- Automatically fabricate complex three-dimensional structures.
- Process multiple joint configurations and a wide variety of steel grades and thicknesses.
- Integrate real-time quality control technologies based on non-destructive techniques.
- Incorporate artificial intelligence and digital twin solutions to optimize the process.
This project aims to move towards a manufacturing approach with fewer defects and greater robustness, generating process stability and repeatability.
Technological approach
The project combines several technological strategies:
- Friction Stir Welding (FSW): A solid-state mechanical joining process in which a rotating tool generates frictional heat to plasticize materials and create a fusion-free weld.
- Real-time quality monitoring: Use of in-process measured parameters (such as rotational speed and forces) along with sensor measurements (acoustic emissions, eddy current and temperature distribution) to nondestructively assess the integrity of the weld as it is being made.
- Artificial Intelligence and digital twin: The data collected by the sensors are processed by advanced algorithms based on neural networks. This allows providing direct feedback to the welding system, optimizing process control.
This approach allows correlating actual process data with possible physical defects in the weld, which facilitates quality prediction and automatic process adjustments.
Demonstrations and use cases
STWIN plans to demonstrate its technology through three specific cases:
- Joints followed by a bending process between steels of very high yield strength and various thicknesses, intended to reinforce the structures of loading cranes.
- High-strength steel joints with cast iron, with curved paths, for application in automotive subassemblies.
- Aluminum-steel dissimilar joints for use in the manufacture of trailers.
These scenarios seek to validate the practical application of the system in real conditions.
International participation
The project is coordinated by Fundación Cidaut, a technology center located in Boecillo (Valladolid).
The complete list of participating European entities is:
- CIDAUT FOUNDATION
- BWI NPO BELGISCH INSTITUUT VOOR LASTECHNIEK VZW
- BWI IND BELGISCH INSTITUUT VOOR LASTECHNIEK INDUSTRIE BV
- ISEND SA INGENIERIA Y SISTEMAS DE ENSAYOS NO DESTRUCTIVOS SA
- GRANALU TRANSFORMATIONS SL
- HIAB AB
- ELEMENT SIX LIMITED
- SSAB EMEA AB
- HORSE POWERTRAIN SPAIN SL
- STIRTEC GMBH
These organizations bring expertise in advanced welding, automation, robotics, non-destructive testing and industrial manufacturing.
What does STWIN bring to the table?
- Powers an automated and versatile welding process for complex metal structures.
- Introduces real-time quality control using non-destructive techniques.
- Integrates artificial intelligence and digital twin to optimize welding system decision making.
- Seeks to validate improvements applicable in real industrial environments.
