Development of Orbital TIG Welding Robot System for the Pipe

Size: px

Start display at page:

Download "Development of Orbital TIG Welding Robot System for the Pipe"

Frank Dickerson
6 years ago
Views:

1 World Academy of cience, Engineering and Technology International Journal of Medical and Health ciences ol:7, No:12, 2013 Development of Orbital TIG Welding Robot ystem for the Pipe Dongho Kim, ung Choi, Kyowoong Pee, Youngsik Cho, eungwoo Jeong, oo-ho Kim International cience Index, Medical and Health ciences ol:7, No:12, 2013 waset.org/publication/ Abstract This study is about the orbital TIG ing robot system which travels on the guide rail installed on the pipe, and s and tracks the pipe seam using the L (Laser ision ensor) joint profile data. The orbital ing robot system consists of the robot, er, controller, and L. Moreover we can define the relationship between ing travel speed and feed speed, and we can make the linear equation using the maximum and minimum amount of metal. Using the linear equation we can determine the ing travel speed and the feed speed accurately corresponding to the area of captured by L. We applied this orbital TIG ing robot system to the stainless steel or duplex pipe on DME (Daewoo hipbuilding and Marine Engineering Co. Ltd.,) shipyard and the result of radiographic test is almost perfect. (Defect rate: 0.033%). Keywords Adaptive ing, automatic ing, Pipe ing, Orbital ing, Laser vision sensor, L, ing D/B. F I. INTRODUCTION OR maintaining global top brand of shipbuilding company, high-added value ship types are needed to be researched and developed with priority of high technique. High-added value ship types like LNGC, offshore plant and passenger ship have large quantities of nonferrous metal such as stainless steel and duplex pipe which demands high-level ing quality. However, the number of high skilled er is very scarce, and their labor cost is increasing, it provides subsequently process delay for the ship. Besides, manual ing by er depends on handcraft, hence, working speed is low, and ing quality is not uniform. In automatic pipe ing development of Korea, an example of applying ing trace line and ing condition to the pipe flange using laser vision [1], and the other example of orbital TIG ing applying ing line tracing using L [2] have been presented. Also, the example of tracking the robot around the pipe in GMA ing applying arc sensor to ing line tracing has been proposed in [3]. We, DME robot team has finished to setup rotatable six-axis robot combining L to stainless steel and duplex pipe,assembly parts in our shipyard, and L interfacing and ing database application is included in this research. This study presents robot system developments. The robot system do trace ing line using L running on the rail installed on circumference direction of fixed pipe, while the system s the v-groove seams of pipes to be in ships and offshore plant. The orbital TIG ing robot system may enable plain er, who has less enough skill to high quality ing, to easily replacing a few high cost TIG er A. ystem Outline II. YTEM CONFIGURATION Orbital TIG-ing robot is largely made up of robot, ing machine, system controller, and L. Once L obtains the inflection point and surface of gap, mismatch, and remaining area of the joint after ing on pipe joint, the system controller compares ing condition to ing database using shape information, and sends commands to the ing machine. Then the user can monitor ing status with GUI (Graphic User Interface) and UP (User Pendant) immediately. Orbital ing robot system configuration is shown in Fig. 1. Fig. 1 ystem configuration of orbital TIG ing robot system B. Robot The mechanical parts of orbital TIG ing robot is composed of robot body, feeder, water-cooling torch, robot drive head, torch weaving part, AC(Automatic oltage Controller), guide rail, and bunch of cables. The size of robot is 250mm (Horizontal) by 250mm (ertical). The height of robot body is designed less than 150mm. D, Kim et. al. are with the Daewoo hipbuilding and Marine Engineering Co. Ltd., eoul Korea (phone: fax: ; donghokim@dsme.co.kr). International cholarly and cientific Research & Innovation 7(12)

2 World Academy of cience, Engineering and Technology International Journal of Medical and Health ciences ol:7, No:12, ~146mm field of view and 0.5mm precision on width direction. Also, maximum 30frame/sec of measuring speed is supported. The communication is allowed via R-232 and R-422. International cience Index, Medical and Health ciences ol:7, No:12, 2013 waset.org/publication/ Fig. 2 Inner structure of robot and embedded motor driver C. Robot Controller The robot controller operates the ing robot system. In addition, it has function of ing process control, ing seam control, optimal control of ing condition, ing robot control, ing machine control, and L control together. Fig. 3 tructure of robot controller The robot controller is divided into three modules, which are ing machine controlling module, L controlling module, and robot controlling module. The function of three parts of modules is as follow. The ing machine controlling module interfaces the ing machine and the robot through LOCALNET communication protocol made by FRONIU. [4] L controlling module controls the L with R-422 protocol. The robot controlling module controls the robot body and integrates the ing machine controlling module, L controlling module and other add-one devices. Along with combining various modules, this module takes a charge of ing robot system management, ing locus creation, optimal ing condition creation and matching with ing database. D. L (Laser ision ensor) The L is made by DME robot team. This model has E. Welder Fig. 4 L (Laser ision ensor) The ing machine is TRANTIG 4000 which is made by FRONIU. This is only for TIG ing process, and emits maximum 400 ampere current. Along with communication module, ROB 5000, ing conditioning controls are done with LOCALNET protocol. III. DEELOPMENT OF WELDING PARAMETER A. Basic Welding Parameters As ing condition is independent with ing position at each pipe section and joint shape of ing object, we developed ing database through experiments selecting optimal ing condition for guaranteeing uniform ing quality. Orbital TIG ing requires high work experience. Without enough ing skill, it is very hard to reproduce the uniform quality of ing itself. Therefore, experiments coping with various situations are necessary for systemizing and defining the knowledge of the professional er. In this study, basic ing condition for pipe diameter and ing path is developed with professional er. The basic ing condition is initially set to ing condition controlling which are the amount of metal and heat input to area. The joint shape and the robot position are additionally referenced to ing condition equations for ing condition controlling. For controlling the amount of metal, heat input to area, the width of weaving, ing condition control is done selecting current, voltage, robot moving speed, feeding speed, and the width of weaving as main parameters. Table I below describes basic ing condition of 6.35mm (thickness) stainless steel pipe. International cholarly and cientific Research & Innovation 7(12)

3 World Academy of cience, Engineering and Technology International Journal of Medical and Health ciences ol:7, No:12, 2013 International cience Index, Medical and Health ciences ol:7, No:12, 2013 waset.org/publication/ TABLE I BAIC WELDING CONDITION OF TAINLE TEEL PIPE Welding parameters 1st Pass 2nd Pass 3rd Pass Current(A) Welding oltage() Travel peed(mm/s) Wire Feed peed(mm/s) Width(mm) Weaving Dwell(s) Weaving(s) B. Automatic Welding Procedure Automatic procedure of orbital TIG ing is done with five steps using joint shape of ing object and robot position of each pipe section. The first step is selecting the pipe. Determining the number of ing path at the second step, the area information is extracted turning the robot around the pipe circumference. Once ing condition based on ing speed and feeding speed corresponding with joint surface is chosen automatically in third step, then the user determines AC(Auto oltage Control: same as ing voltage) value manually with GUI or UP, and manipulates feeding position. This is resulted from proper AC value is hard to be regulated with surroundings (electric grounding, pipe materials). At the fifth step, ing seam line is traced obtaining joint shape in real time while ing, and weaving width is determined measuring the groove width. Eventually, the amount of metal is calculated from the equation of the amount of metal applying cross sectional surface area. C. Determine Amount of Weld Metal for Each Welding Pass (ing area) corresponding to the path could be obtained with joint (sensed joint area while ing). Equations concerning with this study is inferred by experiments and theory. Table II presents the amount of metal of 16"(thickness: 4.78mm) and 24"(thickness: 6.35mm). TABLE II METHOD FOR MULTI PA WELDING Pipe Pass Area of Weld : (mm 2 ) 16" 4.78t 24" 6.35t 1P Final *1) joint +10 1P Fill 0.9 joint 2P Final 11 1P Fill 0.4 joint 2P Fill 0.7 joint 3P Final joint - *2) max +7 *1) joint : ensed joint area while ing *2) joint - max : ensed joint area while ing of 2nd Pass-Max area of D. Control Amount of Weld Metal The semi-automatic orbital ing machine has a defect in control of the amount of metal. The user response of speed and the user posture give limitation of sensing the surface area of joint and manipulating the amount of metal. The semi-automatic machine also shows non-uniform bead shape by initial ing condition. In this study, the robot system supplies the amount of metal to ing joint adjusting ing speed and feeding speed using cross sectional surface area through L. Even cross sectional area is not uniform in inside joint, flat quality of ing can be produced expanding maximum and minimum amount of metal adjusting two parameters simultaneously which are ing speed or feeding speed over cross sectional surface area comparing one parameter, ing speed or feeding speed. Generally, adjusting only one parameter is often utilized. It is due to the face that by both of two parameters is more difficult than by only one parameter to carry out controlling the amount of metal. In this study, equations concerning with two parameters are derived taking advantage of maximum and minimum amount of metal. For automatic control of the amount of metal, ing speed over joint surface area of ing object and feeding speed of it are necessary. Equations derived as following procedure. Considering the amount of metal based on joint cross sectional surface area, basic ing parameters are required. Note that both maximum and minimum ing speed are determined based on ing speed and feeding speed of basic ing condition. Parameters for automatic control mode of the amount of metal are (ing area), (cross-section area of ), ( feeding speed). Equation (1) describes the maximum and minimum ing speed, and (2) is about maximum and minimum feeding speed. [ max] = [ ref.] + α [ min] = [ ref.] α where α is manipulating range of ing speed. The unit of speed is (mm/s). [ max ] = [ ref.] + β [ min] = [ ref.] β where α is manipulating range of feeding speed. The unit of speed is (mm/s). Note that the product of ing surface and ing speed is equal to the product of feeding speed and cross sectional area. Physically, the product means [volume of metal/sec], whose unit is mm 3 /sec. = (3) Arranging to (3), it turns out is the function of ing speed and feeding speed. = (4) (1) (2) International cholarly and cientific Research & Innovation 7(12)

World Academy of cience, Engineering and Technology International Journal of Medical and Health ciences ol:7, No:12, 2013 International cience Index, Medical and Health ciences ol:7, No:12, 2013

Applying the definition of maximum and minimum amount of metal to (4), linear equations, which are function of the area of metal, ing speed, feeding speed, are produced.

4 World Academy of cience, Engineering and Technology International Journal of Medical and Health ciences ol:7, No:12, 2013 International cience Index, Medical and Health ciences ol:7, No:12, 2013 waset.org/publication/ where is constant. Applying the definition of maximum and minimum amount of metal to (4), linear equations, which are function of the area of metal, ing speed, feeding speed, are produced. The function is eventually for speed and feeding speed over desired amount of metal. = = where is maximum ing area is minimum ing area Using (5) and (6), the linear equation (7) is derived as shown in Fig. 5. Fig. 5 Derivation of ing speed and feed speed using Min/Max area of = a + b where a = b = + ubstituting (7) with (3), ing speed and feeding speed are derived. The final equation for the area of metal to ing speed is (8), and the area of metal to feeding speed equation is written on (9); (5) (6) (7) b = a E. Welding Test + For conducting automatic ing of orbital ing robot, outer diameter 16"(400A) and outer diameter 24"(600A) pipe, whose material are identically stainless steel 316, are tried to test for verifying ing robot, L, control algorithm. In this case, the thickness of 16"(400A) pipe is 4.78mm, and the thickness of 24"(600A) pipe is 6.35mm. Both pipe have bevel angle, 70 degree, and are ed to manual mode in root path experiment. hielding gas is purely 100% AR gas. Wire is AW ER316L product, which is diameter, 0.9mm. Fig. 6 Welding test TABLE III AMPLE WELDING CONDITION OF 600A PIPE Welding Parameter Case of 1 Pass Current (A) 100~110(Auto) AC() Auto (after Manually Fixed) Travel(mm/s) 0.9~1.2(Auto) Wire Feed(mm/s) 11~16(Auto) Weld Min(mm 2 ) 5.8 area Max(mm 2 ) 11.3 Auto Width(mm) (depend on the Groove width) Weave Dwell(s) 0.7 Exc.(s) 0.4 (9) b = a (8) Fig. 7 Welding quality of manual ing vs. automatic ing International cholarly and cientific Research & Innovation 7(12)

World Academy of cience, Engineering and Technology International Journal of Medical and Health ciences ol:7, No:12, 2013 I. APPLICATION TO HIP YARD A.

All thirty pipes are tested, and defective rate is 0.033% in those. It is lower defective compared to average defective proportion, 0.5%. Moreover, productivity is improved to 30%. Fig.

5 World Academy of cience, Engineering and Technology International Journal of Medical and Health ciences ol:7, No:12, 2013 I. APPLICATION TO HIP YARD A. Application for tainless teel The field test of robot in DME shipyard has been tried during one month. Materials to have been tested is stainless steel 304 and 316(4.5T~12.7T). All thirty pipes are tested, and defective rate is 0.033% in those. It is lower defective compared to average defective proportion, 0.5%. Moreover, productivity is improved to 30%. Fig. 9 Before and after ing quality of Duplex steel International cience Index, Medical and Health ciences ol:7, No:12, 2013 waset.org/publication/ Fig. 8 tainless steel pipe ing in DME shipyard B. Application for Duplex tainless teel The duplex steel also has been tested in DME shipyard. However, joint ing, whose thickness is over 30mm, was carried out. 30T-ing has many passes (minimum 40 PA); hence, control of heat input to area is very important. Thus, ing has been accomplished maintaining the temperature, 150 Celsius degree between passes.. CONCLUION In stainless steel ing requiring high skilled er, We, DME robot team developed automatic ing robot installed on the rail which is tangential direction to the circular pipe. Using L, obtaining cross sectional surface area of pipe joint, uniform ing quality was guaranteed controlling ing speed and feeding speed with the acquired surface area. Furthermore, the database for stainless steel and duplex pipe was created and successfully applied to our shipyard. Apart from stainless steel and duplex steel materials, carbon steel, titanium, hastelloy, and etc. might be expected to the automatic ing system developed in this study, and ing condition researched in this study will be hopefully provided in those materials. REFERENCE [1] Y. H. Kang,. G. Lee, H. J. ong,. J. Na, T. I. Cha, J. Y Park : A study on the development of automatic ing system for pipe ing, 2001 Annual Fall Meeting of KW, (in Korean). [2] M. G. Kang, J. H. Kim, Y. J. Park, G. J. Woo : Laser ision ystem for Automatic eam Tracking of tainless teel Pipe Welding Machine, International Conference on Control, Automation and ystems [3] K. Y. Bae, J. H. Lee,. W. Chung : A tudy On control of pool and torch position in GMA ing of steel pipe by using sensing systems, Journal of KW, ol. 16, No 5, October, 1998, (in Korean). [4] Fronius user manual for TransTig4000 and Rob5000. International cholarly and cientific Research & Innovation 7(12)

New generation of welding and inspection systems

New generation of welding and inspection systems Throughout the pipeline industry, and particularly in offshore and spool base production, welding requirements are shifting toward higher quality, greater