HEAT TRANSFER IN U-FORM FOLDED STRUCTURAL STEEL PLATE USING GMAW WELDING

Author: Josemairon Prado Pereira
Faculty of Engineering, Department of Mechanical Engineering, Universidade Estadual Paulista “Júlio de Mesquita Filho” – Bauru Campus, SP, Brazil.

Abstract
Several factors affect performance in the fabrication and assembly of metal structures, such as those inherent to the manufacturing process, equipment, welding parameters, and consumables. In this business, folded plate in carbon steel is widely used in order to obtain high quality welded joints. For this, three elements need to remain with low variability and within reliable limits: the welding torch parameters related to recoil, the welding speed, and the cross-section geometry. This work focuses on folded steel U-type plate 100 mm x 50 mm x 2.65 mm, applied to the GMAW welding process, employing the UP6 Motoman robotic arm, on butt joints using AWS ER70S-6 wire electrode, with 75%Ar-25%CO2 gas shielding and welding parameters maintaining current density. In this text, the finite element method was adopted to investigate the simulation of the temperature distribution by ANSYS V.2022 and compare it with that acquired by the TC-08 equipment, by means of the K-type thermocouple. Thus, it was verified that the temperature distribution propagated in a correlated manner to the specimen model, after ascertaining that the robot programming must be adequate to the linear speed corresponding to the torch angular speed, specifically evidencing that the folds are more sensitive to heat in a different manner than the flat plate, demonstrating heat sinks. This simulation was effective and can be applied to other types of folded profiles.

Keywords: Distortion. Welding Simulation. Scanning. Welded joint. Metallic structures.

1. Introduction

The industries that manufacture metallic structures widely employ welding with union of structural elements, whose process as GMAW-MAG inserts high thermal energy in a tiny area in the joint of steel profiles, so that the dimensioned design remains valid in these continuities and remains unchanged (Pramod et al. 2022)
. In this context, three factors need to be with low variability and high reliability, such as the welding speed, the DBCP, and the geometric parameters of the cross section (Tomar, Shiva, e Nath 2022)

In this, the heat generated by the energy source in a small region given by the formation of the fusion puddle caused by the wire-electrode, which spontaneously dissipates, in a radial way, through the sheet metal, with an abrupt change of direction of the normal plane of the sheet in the bending of the profile, suffering alteration in the temperature distribution and happening a delay and commutation of heat between them, by convection and radiation. Most of the researches consulted are based on flat plates, however, the reality in the welding of folded plates or rolled profiles presents a different behavior, in face of thermal development (Hu e Tsai 2007)

Such occurrence extends along the structural element, and heat propagation causes the complex interaction between the design geometric constrictions and deformations, culminating in the residual stresses imposed by the welding thermal cycle (Arias 2022; Pramod et al. [s.d.])

The distortions cause unforeseen problems of deformations, which hinder the fittings designed for the assembly of the steel structure, compromising the continuation of the work (Zhang et al. 2022)
.
Despite these setbacks, despite the heat of fusion introduced by microstructural hardening, if the structural profiles are obtained from the cold-forming of steel sheets, especially in the folded region, they offer sufficient strength to prevent displacements in these links (Sandeep et al. [s.d.])
.
Defining the welding procedures, such as their sequence, is one of the main requirements in controlling the temperature distribution induced by the process, so that several methods have been developed to study it with good precision of results, but with little practical relation (Ghafouri et al. [s.d.]; Pramod et al. 2020)
. The apparent limitation may be associated with the intricate analytical formulation involved in the models used, which include a large number of intervening variables and with a certain precise understanding of the implications of some of them (Mehran et al. 2020)
.
The methods commonly adopted to analyze the temperature distribution are by finite elements, which provide the values, and the measurement is made directly on the plate at the time of welding, by means of K-type thermocouples (Geng et al. 2017)
.
Therefore, with the current density maintained, it is essential to control some welding parameters, such as torch speed and the distance of the DBCP, when the robot travels along the joint of the structural steel folded profile, that, with its constant dimensions, make it possible to show the propagation of the temperature that starts from this welded joint (Bai, Zhang, e Wang 2013)
.
This work presents the welding laboratory results that acquired the temperatures, comparing them with those predicted by the finite element method, in joints composed of cold bent structural profiles.


2. Materials and Methods
​
2.1 Materials
CSN CIVIL 300 steels, in the form of a cold-bent U-shaped profile, was used as the base metal. The wire electrode of specification AWS ER70S-3, diameter 1.0 mm, participated as filler metal. Gas shielding of the welds was provided by a 75%Ar-25%CO2 mixture at a flow rate of 2,1x10-4m3.s-1 (l/min).
​2.2 Methods
​The composition of the welding joints was made by two "U" type profiles, in the dimensions of 100 mm x 50 mm x 2.65 mm, with a height of 85 mm each part, positioned on top along the joint cross section. Figure 1 schematically represents the assembly of these joints for welding:

Figure 1 - Assembling the joint for welding

The welds were performed using the conventional MAG process, with torch displacement done by a Motoman UP6 robotic arm and Yasnac XRC controller, adapted and programmed to perform the welding operations in a single pass, in the horizontal position (2G) and from left to right, as indicated in Figures 2 and 3, with a contact tip distance on the work piece of 1 mm, in all 96 experiments.

Figure 2 - Robotic arm attached to a welding machine

The instantaneous values of the current, welding voltage and the wire-electrode feed speed were compiled through the acquisition system, with data processing SAP-V4.23-IMC. Table 1 provides the average values of the variables:​

Figure 3 – Positioning the welding torch

The technique developed for temperature measurement consists of attaching thermocouples in the holes, with a diameter of 1 mm, as shown in Figure 4, before the welds are made for each specimen:

Figure 4 - Distribution of thermocouple attachment points

The points thus distributed recorded the temperatures during welding, using the TC-08 equipment connected to a PC, in whose files the information during the process was stored.
​
3. Results
All the specimens were at the initial temperature of 25°C, stable and homogeneous, but at the instant the arc is opened, this heat input causes the specimen to change from this point, as the welding torch moves at the speed set by the robot. And something interesting drew attention in the laboratory: when the torch passed through the first and second bend of the profile, at the same speed as in the planes, there was clearly some heat build-up, changing the light intensity in the fusion puddle, pointing out the difference between one and the other and evidencing the localized increase in thermal energy.
Another thing was that, in the first one, despite this occurrence, the resulting weld was more abated than the plane one, but in the second one, the defect of overbite and welding failure always occurred, with holes in the plate, at this point until the final line of this curve, and excess spatter evidencing, again, temperature change to high values. This led to the thought that, as the heat flow moves with the melting temperature, having just ahead the original temperature of 25°C, in the first curve, from the internal side, there is the occurrence of heat exchange with the environment and the adjacent sheet plane, due to the proximity of the sheets converging to the bend. This occurrence, for a certain time, causes the temperature in front of the puddle to be above 25°C; in the second, the same occurrence is aggravated with the edge effect, because the heat begins to return and adds up with the heat exchange between the plates joined by the bend, as can be seen in Figure 5.

Figure 5 – Heat exchange in the folded regions

Thus, it is noticeable that the basic positions of the robot, 1, 2 and 3, have the actual 540 mm walk, causing a slowdown in speed and increasing thermal energy in this very region, as shown in Figure 6.

Figure 6 – Translating the robot when welding

To solve the question, we took the normal path of the curve that has the correspondence in time of 1.35 s, in which we have the speed of 400 mm / s for final calibration of the robot and, thus, automatically avoid the welding problems cited; in function of the speed, we posted Table 2, obtaining a defect-free and homogeneous welding.

This method, with the use of thermocouples, allowed to know a sweep of temperatures superficially throughout the specimen, with precision, when compared to the values established by the finite element method. Figure 7 shows the images of the joints being welded under these welding conditions and those produced by the finite element simulation model.

Figure 7 - Thermal distribution in the web of the experimental folded profile next to that modeled by ANSYS

4. Discussion
It is common for research to consider the welding simulation with plates, which simplifies the mathematical problem, without having to relate the angular velocity of the torch, which is what Islam, Assidi and Warmefjord, among others, did (Assidi et al. 2009; Islam et al. 2014; Wärmefjord et al. 2012)
. Even though there is no record of using this type of cold-formed U-profile, the welding simulation revealed a high degree of assertiveness, below 2%, in line with the work done so far.
Chiocca noticed that the temperature in front of the welding torch can interfere with the speed at which the isotherms travel, altering the temperature distribution (Chiocca et al. [s.d.]).
For the temperature distribution to be equivalent to an unchanged velocity as of a flat plate, in the U-type formed profile, there is a need to adjust the linear velocity, in order to balance the system, so that it is equivalent to the angular one to minimize these effects, solve the problem in the welding process performed by the robot and still conserve the current density. In fact, up to 7°C in front of the welding torch the programmed speed can be admitted, but above this temperature it is necessary to perform a thermodynamic balance of the system in order to be equivalent.
Another aspect regarding the welding speed setting on the U-profile concerns the rotation radius that the UP6 Motoman robot joint performs, since it is not exactly the rotation in the DBCP, which is the result of the operation. Therefore, the degree of freedom of the robot, on the workpiece, must be considered and this welding speed must be related. This is also an extremely simple process when using the plate, as Beygi did, however, its behavior will not be like the one performed on the folded U-type structural profile (Beygi et al. [s.d.])

The welding thermal yield and heat input was determined, according to Kumar's work, with experimental equations(Kumar et al. [s.d.])
Furthermore, the use of the welding parameters coupled to the welding robot resulted in standardized and high-quality welds.
The temperatures measured by K-type thermocouples is a conventional method employed by most researchers, including Long and Yi, in order to validate the model performed by the finite element method (Long et al. 2022; Yi et al. [s.d.]).
The temperature distribution obtained an equality between the methods adopted, which led to similar distortions in the samples due to the single welding direction.
The CSN CIVIL 300 carbon steels showed temperature results with discrete imperceptible differences; now, the joint is an adiabatic region, where the thermal energy generated, heating up to the melting point, presented the perfectly symmetrical distribution. With the cooling that causes shrinkage, part of the deformations is prevented by the presence of the metal profile folds, giving rise to tensile residual stresses.
Numerous thermodynamic systems are formed from the fusion puddle to total thermal equilibrium. The tabs, besides increasing the resistance to distortion, hinder the dissipation of temperature by the edge effect, especially where the weld ends, originating an almost cyclic thermal effect at low temperatures.
As the heat energy input and cooling time of the sample after welding is directly linked with the temperature distribution, the cooling time in this case took around 25 minutes to reach room temperature.
Since the heat input is instantaneous and dissipative, the dimensional limitation of the folded plate prevents the radiation that would help cooling, even more in the presence of folds, with some recapture of the heat given off to the environment by convection, by the internal side of the section, complying with the zero and first law of thermodynamics. The temperatures that cause the expansions and shrinks are maintained with welding speed, resulting in the same distribution behavior. Therefore, it can be said that the influence on the temperature distribution is caused by the thickness of the plate used, the bending shape of the structural plate and the type of structural steel used.
​
Conclusion
CSN CIVIL 300 steels, presented equal temperature distribution for the used methods, with difference between the measured and the calculated, deviation less than 2.6 %.
The presence of tabs on the U-type folded structural steel plate serves as a thermal resistance, being a temperature conservative.
The performance of the welding process with the use of the robot is efficient, with respect to regular temperature distribution in favor of weld quality.
The melting temperature, as a function of the welding thermal energy, is always higher on the side where the welding path of the weld bead starts, caused by the heating, and ends with a discrete variation with 5% lower, when passing through the middle of the welding path to the second fold, when the edge effect occurs.
Through the proposed method, it was possible to evaluate the temperatures, consisting in a proposal for the industrial area, still to be improved and expanded for other cases.
In this context, this method shows itself viable, in its practicality and evaluation of the temperature in structural steel profiles cold formed type "U", consisting of a robot system perfectly adaptable and reliable for the industry of metallic structures, allowing the monitoring of the heat input, by the process of MAG welding, with the robot Motoman UP6.
The simulation of the GMAW-MAG welding process, by the finite element method, in ANSYS, mapped the temperature distribution with a high degree of assertiveness, similar to the results of researchers who worked with other types of materials.
The GMAW-MAG arc welding process with gas protection, in cold formed structural steel plates type "U", in the dimensions 100 x 50 x 2.65 mm, with the proposed joints and integration to the Motoman UP6 robot, with programming in the YASNAC controller language, was successful for the parts that went through the process, using the constant parameters provided in Table 1, resulting in a high-quality weld bead.
The ANSYS simulation of the welding process that modeled the temperature fields is ready for further studies of the displacements and residual stresses by the finite element method of the specimen.
Thus, the theoretical model validated the temperature acquisition by the thermocouples and ratified corresponding to the experimental one.
All experimental studies arising from the MAG welding process, in short-circuit transfer, presented results of heat input in joules per second. The yield (ηm) of the electric arc considered for calculation purposes was from the GMAW-MAG welding with 87%, based on the same equation of Wang's research, which used a very close value, through the formulations of thermodynamics, in experimental and numerical formats of the data from the experiments (Karimi, Wang, e Jelovica 2022).
As for the geometry of the cold-formed profile, using thin plates, it was proven that the consumable has a significant influence on the thermal contribution, and the AWS ER70S-3 wire electrode reveals an action on the fusion puddle with excellent efficiency.
It was evident that, in the structural profile of folded sheet type "U", the presence of the folds reasonably modifies the heat distribution.

5. Acknowledgments
To the Graduate Program in Mechanical Engineering at UNESP-Campus de Bauru-SP.
​
6. Bibliographic references
Arias, Francisco J. 2022. “Micro Heat and Mass Transfer by Bubble-Trains from the Gas Discharging at Nucleation Sites”. International Journal of Heat and Mass Transfer 194. doi: 10.1016/j.ijheatmasstransfer.2022.123042.

Assidi, M., L. Fourment, S. Guerdoux, T. Nelson-International Journal of, e undefined 2010. 2009. “Friction Model for Friction Stir Welding Process Simulation: Calibrations from Welding Experiments”. Elsevier 50:143–55. doi: 10.1016/j.ijmachtools.2009.11.008.
Bai, Xingwang, Haiou Zhang, e Guilan Wang. 2013. “Improving Prediction Accuracy of Thermal Analysis for Weld-Based Additive Manufacturing by Calibrating Input Parameters Using IR Imaging”. International Journal of Advanced Manufacturing Technology 69(5–8):1087–95. doi: 10.1007/S00170-013-5102-Y.
Beygi, R., E. Marques, … LFM da Silva-Concepts in Simulation of Welding, e undefined 2022. [s.d.]. “Heat Sources and Thermal Analysis”. Springer.
Chiocca, A., F. Frendo, L. Bertini-International Journal of Mechanical, e undefined 2021. [s.d.]. “Evaluation of Residual Stresses in a Pipe-to-Plate Welded Joint by Means of Uncoupled Thermal-Structural Simulation and Experimental Tests”. Elsevier.
Geng, Haibin, Jinglong Li, Jiangtao Xiong, e Xin Lin. 2017. “Optimisation of Interpass Temperature and Heat Input for Wire and Arc Additive Manufacturing 5A06 Aluminium Alloy”. Science and Technology of Welding and Joining 22(6):472–83. doi: 10.1080/13621718.2016.1259031.
Ghafouri, M., A. Ahola, J. Ahn, T. Björk-Journal of Constructional Steel, e undefined 2022. [s.d.]. “Welding-Induced Stresses and Distortion in High-Strength Steel T-Joints: Numerical and Experimental Study”. Elsevier.
Hu, J., e H. L. Tsai. 2007. “Heat and Mass Transfer in Gas Metal Arc Welding. Part I: The Arc”. International Journal of Heat and Mass Transfer 50(5–6):833–46. doi: 10.1016/j.ijheatmasstransfer.2006.08.025.
Islam, M., A. Buijk, M. Rais-Rohani, e K. Motoyama. 2014. “Simulation-Based Numerical Optimization of Arc Welding Process for Reduced Distortion in Welded Structures”. Finite Elements in Analysis and Design 84:54–64. doi: 10.1016/j.finel.2014.02.003.
Karimi, Mohammad Reza, Sheng-Hui Wang, e Jasmin Jelovica. 2022. “Experimental Investigation into Weld Cladding of Ni-WC MMC Overlays by CMT Process”. doi: 10.21203/rs.3.rs-1556833/v1.
Kumar, V., BK Roy, A. Mandal-Advances in Manufacturing Engineering, e undefined 2023. [s.d.]. “Thermal Modeling of Wire Arc Additive Manufacturing Process Using COMSOL Multiphysics”. Springer.
Long, Qing, Haiyu Qiao, Xiaodong Yu, Yayun Liu, e Chuanyang Wang. 2022. “Modeling and Experiments of the Thermal Degradation Behavior of PMMA during Laser Transmission Welding Process”. International Journal of Heat and Mass Transfer 194:123086. doi: 10.1016/J.IJHEATMASSTRANSFER.2022.123086.
Mehran, Ghafouri, Ahn Joseph, Mourujärvi Juho, Björk Timo, Larkiola Jari Ghafouri, J. Björk, Mehran Ghafouri, Joseph Ahn, Juho Mourujärvi, Timo Björk, e Jari Larkiola. 2020. “Finite Element Simulation of Welding Distortions in Ultra-High Strength Steel S960 MC Including Comprehensive Thermal and Solid-State Phase Transformation Models”. Elsevier. doi: 10.1016/j.
Pramod, R., S. Mohan Kumar, B. Girinath, A. Rajesh Kannan, N. Pravin Kumar, e N. Siva Shanmugam. 2020. “Fabrication, Characterisation, and Finite Element Analysis of Cold Metal Transfer–Based Wire and Arc Additive–Manufactured Aluminium Alloy 4043 Cylinder”. Welding in the World 64(11):1905–19. doi: 10.1007/S40194-020-00970-8.
Pramod, R., SM Kumar, B. Girinath, … AR Kannan-Welding in the, e undefined 2020. [s.d.]. “Fabrication, Characterisation, and Finite Element Analysis of Cold Metal Transfer–Based Wire and Arc Additive–Manufactured Aluminium Alloy 4043 Cylinder”. Springer.
Pramod, R., S. Mohan Kumar, A. Rajesh Kannan, N. Siva Shanmugam, e Reza Tangestani. 2022. “Fabrication of Gas Metal Arc Welding Based Wire Plus Arc Additive Manufactured 347 Stainless Steel Structure: Behavioral Analysis Through Experimentation and Finite Element Method”. Metals and Materials International 28(1):307–21. doi: 10.1007/s12540-021-01026-2.
Sandeep, KJ, PJ Teja, AK Choudhary, R. Jain-CIRP Journal of, e undefined 2022. [s.d.]. “Development of Correlation between Temperature, Liquid Life Span, Molten Pool, and Porosity during Wire Arc Additive Manufacturing: A Finite Element Approach”. Elsevier.
Tomar, Bunty, S. Shiva, e Tameshwer Nath. 2022. “A Review on Wire Arc Additive Manufacturing: Processing Parameters, Defects, Quality Improvement and Recent Advances”. Materials Today Communications 31. doi: 10.1016/j.mtcomm.2022.103739.
Wärmefjord, Kristina, Rikard Söderberg, Andreas Pahkamaa, Lennart Karlsson, e John Goldak. 2012. “Combining Variation Simulation with Welding Simulation for Prediction of Deformation and Variation of a Final Assembly”. Asmedigitalcollection.Asme.Org. doi: 10.1115/1.4005720.
Yi, B., X. Zhou, C. Shen, J. Wang-Journal of Manufacturing Processes, e undefined 2022. [s.d.]. “Computational Investigation on Transient Thermal Tensioning for Mitigation of Welding Induced Buckling by Elastic FE Method”. Elsevier.
Zhang, Yi, Wenxue Su, Hua Dong, Tao Li, e Hongyou Cao. 2022. “Effect of Welding Sequence and Constraint on the Residual Stress and Deformation of Thick Welded Butt Joint Made of Q345qD Steel” organizado por W. Zhu. Advances in Civil Engineering 2022:5966274. doi: 10.1155/2022/5966274.

Copyright © 2023 Josemairon Prado Pereira. Todos os direitos reservados.
É proibida a reprodução, total ou parcial, do conteúdo sem prévia autorização do autor da obra.