This preview shows page 1. Sign up to view the full content.
Unformatted text preview: Research article CAD interface for automatic robot welding programming
J. Norberto Pires T. Godinho and P Fer reira . Introduction
Robotic welding is by far the most popular application of industrial robots (Books, 1991; United Nations and International Federation of Robots, 2000). Very good research works, achieving very interesting results, were done since the early 1980s, focusing issues like the welding process itself (AGA AB, 2001; Jones et al., 1998; Loureiro et al., 1998; Pires and Loureiro, 2001; Touret, 1997; Choi et al., 1998), sensors (Bolmsjo, 1997a, b; Drews and È Starke, 1986; Yada et al., 1986), welding power sources and robotics welding (Adolfsson et al., 1999; Agren, 1995; Bolmsjo et al., 1995a, b; È Books, 1991; Dahle n and Bolmsjo, 1996; Pires Â È Â et al., 2002; Sa da Costa and Pires, 2001), programming, etc. Nevertheless, when robots are used in the shop ¯ oor to perform welding tasks, there is still too much additional work to guarantee that they perform with the required quality. The process also takes too long, which means that robots take longer to be programmed and set-up than really doing some interesting work. And that is a problem, since actual concepts like availability and agility (Adolfsson et al., 1999; Agren, 1995; Bolmsjo, È 1997b; Bolmsjo et al., 1995a, b; Books, 1991; È Â Dahlen and Bolmsjo, 1996; Kusiak, 1986, È Â 2000; Pires et al., 2002; Sa da Costa and Pires, 2001) are key issues of modern manufacturing. Machines that store high level of ¯ exibility, like robot manipulators, but fail in terms of agility because automation integrators and machine builders were unable, when designing a speci® c machine, to expose that ¯ exibility to users, may become less interesting. This is one of the major issues that robot manufacturers must address clearly in the near future. When robots are used to weld, the programming problem arises immediately. In this paper, we address this issue by presenting a CAD interface enabling fast and simple programming. The paper is organized as follows. First, the CAD interface is introduced and explained, along with the experimental system used to test the proposed solutions. An industrial test case example is then brie¯ y presented and discussed. Finally, conclusions are drawn together with guidelines for future work. 71 The authors J. Norberto Pires is at Mechanical Engineering Department, University of Coimbra, Portugal. ´ T. Godinho is at San Jose SA, Lisboa, Portugal. P. Ferreira is at Mechanical Engineering Institute, Coimbra, Portugal. Keywords Robotics, Computer aided design Abstract Industrial robots play an important role in industry, due to their exibility. Many applications (almost all that require human intervention) may be performed with advantages by robots. Nevertheless, set-up operations, necessary when changing production models, are still tricky and time-consuming. It is common to have detailed data of working pieces in computer aided design (CAD) les, resulting from product design and project. This information is not used satisfactorily, or even not used at all, for robot programming. In this paper, we propose a solution capable of extracting robot motion information from the CAD data. Electronic access The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister The current issue and full text archive of this journal is available at www.emeraldinsight.com/0143-991X.htm Industrial Robot: An International Journal Volume 31 · Number 1 · 2004 · pp. 71–76 q Emerald Group Publishing Limited · ISSN 0143-991X DOI 10.1108/01439910410512028 CAD interface for automatic robot welding programming Industrial Robot: An International Journal Volume 31 · Number 1 · 2004 · 71–76 J. Norberto Pires et al. CAD interface
Currently, since the vast majority of companies use CAD programs to design their products, information from CAD ® les could be used to generate robot welding programs. The application presented here enables the user to work on the CAD ® le, de® ning the welding and approach/escape paths between two consecutive welds, and organize them in the desired welding sequence. When de® nition is complete, a small program converts it to robot commands that can be immediately tested for detailed tuning. A set of tools is then available to speed up necessary corrections, which can be made online with the robot moving. After a few simulations (robot performing all motions without welding) the program is ready for production. The whole process can be completed in just a few hours, representing a huge reduction of programming and set-up time. Following the current developments available in the Industrial Robotics Laboratory of the Mechanical Engineering Department of the University of Coimbra, and meeting a request from the industry, a solution was designed using AUTOCAD and DXF ® les (currently standard ® les for all CAD software tools). If the user follows some basic rules, and produces a DXF ® le with all the information needed, then the application developed is capable of extracting that information from the CAD ® le, and is able to produce a robot program almost ready for production. The user starts by having a 3D drawing of the piece to weld and of the table used to hold the piece. The 3D models should be very precise in dimensions and in part positioning. Then the user should draw (Plate 1) all the trajectories required to fully weld the piece as desired, using the available layers, i.e. using one layer for each trajectory, which is composed of a start-point and an end-point, both with orientation, and the type of motion (welding trajectory or approach/escape trajectory). The welding parameters (velocity, voltage, torch distance to the surface) are introduced in the selected layer, just by adding labels with the corresponding values. The weld layers should then be renamed for easy identi® cation. Following this simple small set of rules, the designer can add to his 3D models 72 information about how the welding process should be done. The DXF ® le generated by the CAD application (AUTOCAD in our case) includes all the information added to specify the welding process. Since the DXF ® le is an ASCII ® le, it is very easy to extract the above-mentioned information, using a simple application (Plate 2) that identi® es each added welding layer, and the related welding parameters, and stores all that information in a known way. Here the de® nition shown in Figure 1 was used. The generated dat ® le is used as input for the application shown in Plate 3. This application shows the available de® nition with the help of several push-down buttons, and enables the user to change the welding parameters, correct points and orientations, simulate the whole process using the real robot and the real piece to weld. The simulation is very realistic, making the ® nal program ready for production. Test case example
The laboratory set-up (Plate 4) is composed of an ABB IRB1400 M94/S4 industrial robot, equipped with a ESAB A350 welding power source controlled from the robot using our own software (we do not use any of ABB WeldWare software modules), and a RPC-based software interface to a PC (RAP, 1996), explored using Â the software package PCROB (Pires and Sa da Costa, 2000; Pires et al., 2002). PCROB is a collection of software tools that enable control of the robot from a PC. In our case we use a Pentium II PC running Windows 2000/ SP4. The piece used in the proposed example (Plate 4) is a scale model of a piece very common in companies building structural components for the construction industry. A 3D drawing of the selected piece and holding table (Plate 5) is then necessary to proceed. Subsequently, the welding trajectories should be added as explained earlier to completely de® ne the welding process. The test weld is very simple and is shown in Plate 5 connecting the topside from left to right. The welding parameters were selected using a welding database (The Welding Institute, 2000) by taking into consideration the material in the test piece. CAD interface for automatic robot welding programming Industrial Robot: An International Journal Volume 31 · Number 1 · 2004 · 71–76 J. Norberto Pires et al. Plate 1 Example of welding trajectories using available layers Plate 2 Application to extract information from a DXF CAD le Conclusion
In this paper, we presented a CAD interface capable of adding in the programming process of welding applications. The solution has proved to be very interesting, permitting a big reduction of the programming process just by extracting welding de® nition from the CAD ® le, and enabling a fast pre-program that is very easy to simulate and adjust. The system was tested using linear welds on a piece, built to scale from real structural pieces for the construction industry, and showed to be accurate and very easy to use. Further development is needed in a way to de® ne all the situations that may arise in industrial environments. Nevertheless, the results obtained show that a general solution is possible just by adding simple rules to match each case. Further developments should be carried out just by making the solution compatible with other robots/controllers. This requires standard software and platforms, refusing to use proprietary licenses and protocols. 73 The results obtained in this test case were very promising, since it was possible to make the weld less than 5 min after ® nishing the CAD model of the system, just by generating the DXF ® le, extracting the information to the simulation application (Plate 3) and make a few adjustments to solve any small positioning errors, and testing everything using the simulation facility of the tool represented in Plate 3. A video of the system performing can be found in http:/ /robotics. dem.uc.pt/norberto/cv/videos/pedro_welding. wmv, where only the welding process is shown. CAD interface for automatic robot welding programming Industrial Robot: An International Journal Volume 31 · Number 1 · 2004 · 71–76 J. Norberto Pires et al. Figure 1 De nition of the welding le obtained from the DXF CAD le 74 CAD interface for automatic robot welding programming Industrial Robot: An International Journal Volume 31 · Number 1 · 2004 · 71–76 J. Norberto Pires et al. Plate 3 Program used to adjust program obtained from CAD le Plate 4 Aspect of the experimental system and test piece 75 CAD interface for automatic robot welding programming Industrial Robot: An International Journal Volume 31 · Number 1 · 2004 · 71–76 J. Norberto Pires et al. Plate 5 3D drawing of welding table and selected piece References
Adolfsson, S., Bahrami, A., Bolmsjo, G. and Claesson, I. ¨ (1999), “On-line quality monitoring in short-circuit gas metal arc welding”, Welding Journal, Vol. 75, pp. 59s-73s. AGA AB (2001), “AGA AB, Facts about rapid ArcTM. Rapid MeltTM. High-Productivity MIG/MAG welding”, AGA AB, Sweden. Agren, B. (1995), “Sensor integration for robotic arc welding”, PhD thesis, Lund University. Bolmsjo, G. (1997a), “Knowledge based systems in robotized ¨ arc welding. knowledge based systems – advanced concepts”, in Tzafestas, S. (Ed.), Techniques and Applications, Chapter 17, ISBN 9810228309, World Scienti c, Singapore, pp. 465-95. Bolmsjo, G. (1997b), “Sensor system in arc welding”, ¨ Technical Report, Lund Institute of Technology, Production and Materials Engineering Department. Bolmsjo, G., Olsson, M., Nikoleris, G. and Brink, K. (1995a), ¨ “Task programming of welding robots” , Proc. of the Int. Conf. on the Joining of Materials, JOM-7 , May-June, pp. 573-85. Bolmsjo, G., Olsson, M., Nikoleris, G. and Brinkm, K. (1995b), ¨ “Task programming of welding robots” , Proc. of the Int. Conf. on the Joining of Materials, JOM-7 , May-June, pp. 573-85. Books, B. (1991), “Welding robots – the state-of-the-art”, Welding and Metal Fabrication,. Choi, S.K., Yoo, C.D. and Kim, Y.S. (1998), “Dynamic simulation of metal transfer in GMAW. Part 1: globular and spray transfer mode welding”, Vol. 77 No. 1, pp. 38-44, “part 2: short circuit transfer mode”, pp. 45-51, Welding Journal. ´ Dahlen, P. and Bolmsjo, G. (1996), “Human factors in the ¨ justi cation of advanced manufacturing systems”, International Journal of Human Factors in Manufacturing, Vol. 6 No. 2, pp. 147-62. Drews, P. and Starke, G. (1986), “Development approaches for advanced adaptive control in automated arc welding”, Internal Report XII-970, Mechatronics Department, Aachen, Germany. Jones, L.A., Edgar, T.W. and Lang, J.H. (1998), “Images of steel electrode in Ar-20% shielding during constant current gas metal arc welding”, Welding Journal, Vol. 77 No. 1, pp. 135-41. Kusiak, A. (1986), Modelling and Design of Flexible Manufacturing Systems, Elsevier, Amsterdam. Kusiak, A. (2000), Computational Intelligence, Wiley, New York, NY, USA. Loureiro, A., Velindro, M. and Neves, F. (1998), “The in uence of heat input and the torch weaving movement on robotized MIG weld shape”, International Journal for the Joining of Material, Vol. 10 No. 3/4, pp. 86-91. Pires, J.N. and Loureiro, A. (Eds) (2001), Proceedings of the International Workshop on Robotic Welding Systems and Process Monitoring, (http://robotics.dem.uc.pt/ welding/), January 2001, Coimbra, Portugal, Journal Robotica, January 2001. ´ ´ Pires, J.N. and Sa da Costa, J.M.G. (2000), “Object oriented and distributed approach for programming robotic manufacturing cells”, IFAC Journal on Robotics and Computer Integrated Manufacturing,. Pires, J.N., Loureiro, A., Godinho, T., Ferreira, P., Fernando, B. and Morgado, J. (2002), “Object oriented and distributed software applied to industrial robotic welding”, Industrial Robot: An International Journal, Vol. 29 No. 2, pp. 149-62. RAP, Service Protocol De nition, ABB Flexible Automation (1996). ´ Sa da Costa, J. and Pires, J.N. (2001), “Future welding robot ´ developments”, Journal Robotica, Vol. 41. The Welding Institute (2000), Standard Data for Arc Welding, Abington Hall, Abington, UK. ´´ Touret, S. (1997), “Le procede T.I.M.E”, Soudage et Techniques Connexes, pp. 29-33. United Nations and International Federation of Robots (2000), “World industrial robots 1996: statistics and forecasts”, ONU/IFR 2000, New York. Yada, Y. et al. (1986), “Development of real time welding control system by using image processing”, Internal Report XII-998, Mechatronics Department, Aachen, Germany. 76 ...
View Full Document
- Spring '10