Objectives and content
The proposed MAIL project is aimed at basic improvement in the reliability of micro-spot laser welding of metal parts in production, by the deveopment of a self-learning real-time feedback control system.
The basic problem of this technology is how to attain the Par tPerMillion PPM) defect level needed for wider industrial application in micro-mechancs and electronics. The two most important defect sources are addressed in this project:
- Defects due to position tolerances of the products: an "Aim and Shoot" system that recognizes the joint position and adjusts the actual laser spot position will be developed.
- Defects resulting from the variation in surface conditions of the parts: a generic multi-sensor realtime monitoring and the following modules will be developed:
- advanced pattern recognition for joint classification (allowing process feedback through the line operator),
- adaptive control, based on process modelling ant Neural Network self learing,
- real-time closed loop control of the beam energy. The to be developed system has a modular architecture, the monitoring part can be combined with one of the three control modules, depending on application.
This modularity will allow customised applications, covering several market segments. The long term prospect of this development is to partly replace the present soldering technology by the laser spot welding technique: clean production and green products. The 1st phase -Process Identificationcontains investigation into the relevant physieal parameters for joint assessment & classification. In the 2nd phase - Closed Loop Control - adaptive feedback as well as closed-loop control during the laser pulse will be developed. The Neural Networks module will allow, through its self-learning capability, quick start-up of a new application In the 3rd phase - Testbed Applications -the developed system will be tested in three prototype test-beds.
Funding SchemeCSC - Cost-sharing contracts
43800 Valls - Tarragona