دانلود رایگان ترجمه مقاله روش های کامپوزیت برای آنالیز کیفیت در جوشکاری لیزر خودرو (نشریه IEEE 2003)

تکنیک های کامپوزیت برای تحلیل کیفیت در جوشکاری لیزر خودرو

Composite Techniques for Quality Analysis in Automotive Laser Welding

Abstract

Real-time monitoring of the laser-based applications is becoming a main issue for quality analysis in the steel manufacturing industry. The paper suggests a solution achieving an automated real-time quality inspection in laser welding applications. A composite system composed of soft computing and traditional techniques has been considered for its positive impact on the reduced computation compared with more traditional approaches.

1 Introduction

Early detection of defects in metal manufacturing industries is becoming a main economical issue for its impact on quality analysis of the artifact and the industrial process.

Automated laser welding is having an increasing growth and diffusion due to the powerful and versatile process that allows for joining metals or non-metals at high speed with a relatively low heat input [l]. Lasers can produce welds in air, vacuum, controlled atmospheres, and pressurized chambers with a high reliable and automated process. Moreover, laser light may be focused to very small areas [2].

In general, the quality analysis of the process is assessed by offline inspection of each welded component. Such inspection phase is a time-consuming activity that rarely provides useful feedback to the expert user on how to improve the process itself (e.g., by tuning some parameters). In addition, external inspection of the welded artifact requires a costly procedure, most of times carried out with ultrasonic devices [3]. A different approach envisage an on-line quality analysis implemented directly during the process. A set of information is extracted from the sensors observing some critical parametedphenomena associated with the process [4]. Following this philosophy we address the laser welding quality analysis with a selftuning classification system that detects defects directly during the welding process. Moreover, the suggested procedure enables the process engineer to assess the quality of the process and study it aging effects for a subsequent process tuning. The quality analysis system is a composite system, namely a synergic composition of traditional algorithms for signal processing and sot?- computing techniques [5]. During development of the solution we addressed not solely the performance issue (in tenns of detected defects) but also the required computational burden in order to satisfy real-time requirements.

In particular, the industrial process under monitoring refers to the laser welding of automotive components carried out at the CW-FIAT laboratories [17]. The specific test bed is a steel gear, a critical part in the gearbox for a passenger vehicle (see Figure I).

Seven types of gear, built joining two rings (i.e., a light syncro gear and the principal gear), are actually butt welded using a CO2 laser. The test-bed is part of the Intelligent Manufactory Systems project SLAPS (Selftuning and User-independent Laser Material Processing Units) [ 161 a European Union effort to advance the applied research on the development of intelligent laser processing unit [15].

The structure of the paper is as follows. Section 2 introduces a general background on laser welding and a description of the typical defects. Section 3 presents the features extracted from our industrial set-up and in section 4 is described the overall composite algorithm describing also the classifier selection, the tuning methodology and the experimental results.