VIRTUAL AND AUGMENTED REALITY FOR QUALITY IMPROVEMENT OF MANUAL WELDS P. Tschirner, S. Nordbruch", A. Gräser" ABSTRACT The large problem to create manual welds of constant high quality results from missing optical information during the actual welding process. Due to the extreme brightness conditions e.g. in arc welding and the use of protective glasses even experienced welders can hardly recognize details of the welding pool, the welding seam and the environment. This paper describes a new research project for the development of a system for the support of the welder. KEYWORDS Virtual and augmented reality, manual arc welding, quality control INTRODUCTION Welding is one of the most important industrial manufacturing methods. The creation of manual welds is often used in production of unique pieces where an automation is impossible or uneconomic, e.g. in shipbuilding with its complicated hull geometry. A common welding method is the inert-gas arc welding e.g. manual gas tungsten-arc welding (GTAW) or manual gas metal-arc welding (GMAW). In all these processes the electrical arc transfers energy to the welding seam. This arc has an extraordinary high brightness and ultraviolet radiation and for observation a welding helmet with suitable protective glasses is necessary. These glasses absorb the radiation and darken the entire scene, so even experienced welders can hardly recognize details of the welding pool, the welding seam and the environment. Figure 1 shows a view of the welder recorded with a CCD camera through standard protective glasses. Figure 1: Simulated view of the welder recorded with a CCD camera and standard protective glasses. University Bremen, Institute of Automation (IAT), Kufsteiner Str. NW1, 28359 Bremen, Germany A further disadvantage from protective glasses is the restricted field of view of the welder. Furthermore the welder has no additional information, e.g. the actual parameters of the welding power supply. This is particularly of importance if the welding had to be done in far distance from the welding power supply. During the welding process the requirements at the concentration and reliability of the welder are very high because of the insufficient optical information. Due to these missing optical information the creation of manual welds with constant high quality becomes very difficult. To increase the manufacturing quality and economic efficiency a support system for the welder is required. This can be achieved by improving the visual information for the welder as well as by supplying additional information with methods of virtual and augmented reality. The described system is based on the approach for the optimization of PGMAW using visual online observation of the droplet transfer in combination with recording of electrical welding parameters described by Nordbruch, Tschirner and Gräser e.g. in [1], [2]. VIRTUAL AND AUGMENTED REALITY Virtual reality is a new technology, which uses a computer model to generate and present an artificial environment to a person, evoking the impression of actually moving in this envi ronment. Augmented reality is a new form of human-machine interface, which inserts information via head-mounted displays in the users field of view. The insertion is context dependent, i.e. compatible to and derived from the observed object, e.g. the real field of view of a mechanic is extended by the insertion of instruction sheets. Apart from this the use of wearable computers opens new augmented reality application fields, in which high mobility as well as actual process values, measured or simulated data are required. DESCRIPTION OF THE SYSTEM The system combines a conventional welding helmet with modern technology. The system consists of a welding helmet combined with two High-Dynamic-Range-CMOS-cameras (HDRC) for observation of the welding scene a head-mounted display for visualization of the welding scene ⚫ a wearable computer for image recording and image processing a computer for measuring electrical welding parameters and automatic calculation of characteristic welding process parameters The basic set-up of the system is shown in figure 2. |