CLET - Closed Loop Control of the Laser Welding Process

RESEARCH GROUP OverlappedLaserWeldin2

  • A.J. Huis in 't Veld
  • R.G.K.M. Aarts
  • Ir. A. R. Konuk (contact person)


  • April 2008 – March 2011


laser welding, robotic control, real-time process control, laser process, process control, optical emission, laser induced breakdown spectroscopy, image processing


In recent decades the laser welding process has become widely used in industry. Automotive, aeronautical, consumer goods, packaging are typical application areas. Quality of laser welded seams and speed of operation contributed to its diffusion over a wide range of industries.

Besides its advantages over other welding techniques, laser welding has also some drawbacks due to its nature. Welding defects may occur such as holes, pores, lack of penetration, and spatter. To avoid these defects a correct parameterization of laser welding process is essential. By controlling the parameters like laser power, welding speed, focal distance, shielding gas flow, lateral position a better weld quality could be obtained.

The goal of the CLET project is to find a reliable method for detecting and avoiding faults in laser welding. The best method should detect the welding defects in real-time and inspect the whole production without time increment.KeyholeWelding3D

This project is a European 7th Framework Program and it is being developed by a consortium of RTD’s and SME’s, see the project’s web page at The project partners are:

  • Fundación CARTIF (Spain)
  • CNR-INFM Laser Innovation Technology Transfer Training (Italy)
  • University of Twente (Netherlands)
  • Palacky University in Olomouc (Czech Republic)
  • Josdan Soldadura Láser y Ajuste S. L. (Spain)
  • Precitec KG (Germany)
  • Vatrans Zlin (Czech Republic)
  • Flexweld (Netherlands)


Within the boundaries of this project goals are summarized below:

  • Real-time defect detection system for welding with Nd:YAG and CO2 lasers. Considered materials are steel, stainless steel, zinc-coated steel and aluminium
  • Development of a feed-forward controller able to set the process parameters according to the signals generated by the defect detection system.
  • Rejection of perturbations that can cause defects. In this way the final product quality will be assured.
  • Simplification of adaptation to new specimens.


Methods based on analyzing electromagnetic emissions in ultraviolet, visible, and near infrared spectrum ranges can be classified in two main categories:

  • Broadband intensity analysis
  • Real-time spectrum analysis

Laser weld monitoring is a very common method employing optical sensor devices like photodiodes to measure the intensity of the emissions is a specific spectral range. The signals generated by them are analysed to obtain information about the possible weld defect. Real-time spectrum analysis is a relatively new method examining the spectra of the weld emission. The emission is received by an optical collimator and analysed with a high speed spectrometer device. An example plot of an emission from AISI304 stainless steel welding with a 1000W disk laser is given. During the welding operation the system acquires continuously spectra that have to be processed in real-time and specific features in the spectra must be related to the status of the welding process.

Retrieved information is then used in a supervisory controller to adjust the parameters of weld process. Such adjustments may be done in real-time as well as by feedforward control in such a way that the quality related signals always keep their values close to those values associated to good quality welds. In this way, defects are avoided in a automatic fashion and, moreover, the reparameterisation of the facility for new kinds of specimens becomes easier.


  • Laser welding emission analysis with spectrometer in relation to weld status
  • Real-time control system implementation


The research leading to these results has received funding from the European Community's Seventh Framework Programme (FP7-SME-2007-1) under grant agreement n° 222279.