System for in-process quality assurance of resistance spot welding using neural networks

Project Information

SQUAW

Grant agreement ID: BRPR980677

Project closed

Start date 1 September 1998

End date 31 August 2001

Funded under

• Specific research and technological development programme in the field of industrial and materials technologies, 1994-1998

Total cost

No data

EU contribution

No data

Coordinated by

HARMS & WENDE GMBH & CO KG
Germany

CORDIS provides links to public deliverables and publications of HORIZON projects.

Links to deliverables and publications from FP7 projects, as well as links to some specific result types such as dataset and software, are dynamically retrieved from OpenAIRE .

Deliverables

SQUAW: System for in-process quality assurance of resistance spot welding using neural networks

The project aimed to develop a guaranteed and reliable on-line quality assurance system using artificial intelligence, which would detect defective joints during the resistance spot welding process. Resistance spot welding is relatively easy to integrate in automatic plants. Therefore it is one of the most important welding methods used in industrial series and mass production with welding machines and welding robots. It is also used in manual single-piece production in the workshop sector, when sheet-sheet, wire-wire or wire-sheet joints are to be made. The spectrum of applications includes car body construction in the automotive industry, apparatus engineering, production of electrical appliances and household devices, as well as the production of small electrical components. Prototype: Originally two prototypes were supposed to be done: - One for the application in electrical and - One for the application in automotive industry, and furthermore the results concerning their performance under production conditions worked out during the field tests. During the project’s execution it was found out, that separate prototypes for electrical and automotive industry as previewed are not necessary to be constructed. Therefore, only one prototype was constructed. It was tested in both applications. The prototype includes two main modules: the hardware which includes modules for data measurement, signal conditioning and data recording and the software modules for measurement fault detection, data pre-processing and for the online quality prediction. Results from the field tests With the prototype it is possible to carry out tests under production conditions. The prototype can create multiple projects to observe various welding processes. Each project can observe up to 256 welding programs and up to 256 welding spots per welding object. Additionally the specific parameters such as welding time, sample rate and some others can set in several dialogs.