CrossMark

Available online at www.sciencedirect.com

ScienceDirect

Procedía Engineering 97 (2014) 790 - 799

Procedía Engineering

www.elsevier.com/locate/procedia

12th GLOBAL CONGRESS ON MANUFACTURING AND MANAGEMENT, GCMM 2014

The Effect of Welding Process Parameters on Pitting Corrosion and Microstructure of Chromium-Manganese Stainless Steel Gas Tungsten Arc Welded Plates

SUDHAKARAN. Ra*, SIVASAKTHIVEL. P.Sb, NAGARAJA.Sc and EAZHIL. K.Md

aSNS College of Engineering, Kurumbapalayam, Coimbatore -641107, India bSASTRA University, Thanjavur- 613401 ,lndia c,dSNS College of Engineering, Coimbatore -641107, India

Abstract

In this paper, the effect of gas tungsten arc welding (GTAW) parameters on the pitting corrosion on AISI 202 chromium manganese stainless steel was investigated. An empirical mathematical equation correlating pitting resistance equivalent number (PREN) with the welding parameters such as welding current, welding speed, welding gun angle and gas flow rate was developed. Central composite response surface methodology with four parameter and five levels was employed for conducting the experiments. The adequacy of the developed model was checked using ANOVA. The main effects of the process parameters on PREN of the welded joints were studied using surface and contour plots. The quality of the weld joint is also highly influenced by the microstructure of the weldment. The quality of the weldment gets deteriorated due to metallurgical changes such as micro-segregation, precipitation of secondary phases, presence of porosities, grain growth in the heat affected zone and loss of material by vaporization. This paper also investigates the influence of welding gun angle on the microstructure examination of the weldment. The microstructure study concentrated on the grain structure, presence of carbides and formation of ferrite, austenite and martensite in the weldment. The results obtained from the present investigation helps in quickly selecting the required process parameters to achieve the desired PREN and weld quality.

© 2014TheAuthors.PublishedbyElsevierLtd.Thisisan open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

Selection and peer-review under responsibility of the Organizing Committee of GCMM 2014

Keywords: Gas tungsten arc welding, stainless steel, grain structure, pitting resistance equivalent number

* Corresponding author. Tel.: 919894030121; fax: +914226465201. E-mail address: absudha@yahoo.com

1877-7058 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

Selection and peer-review under responsibility of the Organizing Committee of GCMM 2014 doi: 10.1016/j.proeng.2014.12.281

1. Introduction

In Gas Tungsten Arc Welding (GTAW) process, the interfuse of metals was produced by heating them with an arc using a non consumable electrode. It is widely used welding process finds applications in welding hard to weld metals such as aluminium, stainless steel, magnesium and titanium [1]. The increased use of automated welding urges the welding procedures and selection of welding parameters must be more specific for good weld quality and precision with minimum cost [2]. The bead geometry plays an important role in determining the microstructure of the welded specimen and the mechanical properties of the weld [3].The proper selection of the input welding parameters which influence the properties of welded specimen ensure a high quality joint. Stainless steels are corrosive resistance in nature finds diversified application. Even stainless posses good resistance, they are yet susceptible to pitting corrosion. The pitting corrosion is a localized dissolution of an oxide-covered metal in specific aggressive environments. It is most common and cataclysmic causes of failure of metallic structures. The detection and monitoring of pitting corrosion is an important task in determining the weld quality. The pitting corrosion is a random, sporadic and stochastic process and their prediction of the time and location of occurrence remains extremely difficult and undefined [4]. Stainless steels may also suffer from different forms of metallurgical changes when exposed to critical temperatures. In welding, the heat affected zone often experiences temperatures which cause sufficient microstructural changes in the welded plates. The precipitation of chromium nitrides, carbides and carbonitrides in the parent metal occur under various welding and environmental conditions and also depends on the grades of stainless steel. During GTAW process, the formations of coarse grains and inter granular chromium rich carbides along the grain boundaries in the heat affected zone deteriorates the mechanical properties. Kondapalli Sivaprasad et al. [5] studied the influence of welding parameters on pitting corrosion rate of AISI 304L welded sheets using pulsed current micro plasma arc welding process. They developed an empirical relation correlating various process parameters for pitting corrosion rate. They conducted experiments using five factor and five level central composite rotatable design matrix. They have studied the effects of process parameters using surface and contour plots. Siva and Murugan [6] studied the effects of plasma transferred arc welding parameters on the pitting corrosion resistance of Colmonoy 5 overlays deposited on austenitic stainless steel plates. They used Potentiodynamic polarization technique for conducting the corrosion tests. A mathematical model was developed using multiple regression technique and the effects were studied.

Subodh and Shahi [7] studied the influence of heat input on the microstructure of GTA welded AISI304 grade stainless steel joints. They used three heat input combinations for their study. They designated the heat combinations as low heat (2.56 kJ/mm), medium heat (2.784 kJ/mm) and high heat (3.017 kJ/mm). From their study, they have found that significant grain coarsening was inferred in the heat affected zone for all the joints. They have also found that the coverage of grain coarsening in the HAZ increased with the increase in heat input and average dendrite length and interdendritic spacing in the weld zone also increases with the increase in heat input. Huaibei et al [8] investigated the microstructure analysis of low carbon 12% chromium stainless steel at high temperature in the heat affected zone. The main problems attained during welding with low carbon 12% Cr stainless steel was coarsening and embrittlement. In their study, they did microstructures microstructural investigations in the heat affected zone under high temperature with different chemical constituents and heat inputs using thermal simulation tests. From their study they have inferred that the heat input influences the microstructure of steel in high temperature heat affected zone and the grain size grows up with the increase in heat input.

There were many works carried out to study the pitting corrosion resistance and the microstructure of various grades of stainless steel using different welding processes. The study reveals the work pertain to 200 series stainless steel was limited, especially 202 grade using GTAW. AISI 202 chromium manganese stainless steel contains nitrogen; exhibits effective strengthening addition was harder and stronger than 300 series. It posses one third more yield strength than AISI 304 stainless steel, hence it is used as an alternate for AISI 304 in manufacturing of propeller shafts for motor boats, medical instruments, kitchen equipment and food processing equipment [9]. The rise in popularity of these grades is linked to world Nickel price volatility and advances in steel production technology. In countries like India which has no indigenous supply of Nickel, there is a potential for growth of these grades. But before considering these grades for any application detailed investigation has to be carried out regarding its suitability, since there are certain instances where 200 series grades have failed due to pitting corrosion and other problems. Therefore, it becomes imperative to study the effect of welding process parameters on the pitting corrosion and microstructure of these grades before considering it for any application. Hence in the present work AISI 202 grade chromium manganese stainless steel is chosen with the following chemical composition.

Table 1 Chemical Composition of AISI 202

%C %Mn %Si %Cr %Ni %P %S %N %Fe 0.15 9.25 0.49 17.1 4.1 0.06 0.03 0.25 70.01

In this work, a pitting resistance equivalent number (PREN) in the weld bead was estimated form the alloying elements such as chromium, molybdenum and nitrogen. From the estimated values of PREN a statistical model was developed to predict PERN in terms of welding parameters. The study on the PREN revealed that the heat input has major effect on pitting corrosion. A detailed microscopic study has been made to study the effect of welding gun angle on the fusion zone, heat affected zone and base metal. The study on the structure of the specimen's revealed the influence of process parameters on grain structure, grain coarsening in the heat affected zone and weld metal. It also gave an idea about formation of martensite and ferrite veins in the weld metal.

2. Experimental procedure

The experimental design was based on central composite rotatable design matrix consisting of four factors and five levels each with full replication [10]. The experiments were conducted on randomized manner as per the design matrix using electric digital welding machine (Lincoln V 350 PRO). The uniform welding was maintained using a servo motor driven manipulator. The experimental set up consists of a traveling carriage with a table for fixing the work specimens. The power source was set at right. The welding gun was fabricated with an attachment for setting the welding gun angle and the required nozzle to plate distance. It was held stationary in a frame above the table. The nozzle to plate distance was set constant at 2.5 mm for all the experiments. To generate arc at this distance an high frequency attachment was used in the frame. The specimens AISI 202 plates were cut to size (100 mm X 30 mm) and 5 mm of thickness and the surface to be welded was cleaned to remove oxide scale and dirt. The flow rates of argon gas ranges between 5 - 25 liters/min was used for shielding purpose to protect the weld area from atmospheric gases. The experimental setup used for conducting the experiments is shown in Fig. 1 and the welded specimens are shown in Fig 2.

Fig. 1. experimental setup Fig. 2. welded specimens

3. Research methodology

The welding parameters that can be independently controlled were identified based on their influence on PREN and the microstructure. The parameters considered for experimentation are welding gun angle (9), welding speed (V), welding current (I) and shielding gas flow rate (Q). The working ranges of all these selected parameters were set by conducting trial runs. The range of parameters for 9, V, I, and Q identified through trail experiments are given in Table 2. The upper limit of the parameter is coded as 2; lower limit, as -2; and the coded values for

intermediate values were calculated from the following relationship

v _ 2(2X~ (Xmax + Xmin ) Xi - (1)

Xmax Xmin

Where Xi is the required coded value of a variable X. X is any value of the variable from Xmin to Xmax. The process parameters and their levels with its limits and notations are given in Table 2. The process parameter at the intermediate (0) level constitute the centre points, while lower value (-2), higher value (+2) and intermediates (-1 and +1) with the other two parameters at the intermediate level constitute the star points [12].

Table 2 Process Parameters and its Levels

Parameter Unit -2 -1 Levels 0 1 2

Gun Angle ( 6) Deg 50 60 70 80 90

Welding Speed (V) mm/min 80 90 100 110 120

Welding Current (I) Amperage 70 80 90 100 110

Shielding Gas flow rate (Q) liters/min 5 10 15 20 25

4. Recording the response

PREN is a theoretical way of estimating the pitting corrosion resistance of stainless steel based on the chemical compositions of an alloy. The most commonly used equation to determine the PREN is give by Eq. (2)

PREN = %Cr + 3.3(%Mo) + 16(%N) (2)

By conducting wet analysis on the welded specimen the accurate chemical composition of all the welded specimens were determined. The IS 228 test method was employed to determine the chemical composition. PREN was estimated from the chemical constituents using the equation 2. The PERN evaluated for each welded specimens along with its constituents is given in the table 3

4. Experimental design

The experimental design chosen was central composite rotatable design for fitting a second-order quadratic equation based on the criterion of rotatability. The design plan chosen consists of 31 experiments. It is four parameters and five levels central composite rotatable design comprising of 31 sets of coded conditions. The design matrix for the experiment comprises of full replication of 16 factorial design, 8 star points and 7 centre points. These corresponds to first 16 rows, rows from 17 to 24 and the last seven rows respectively, in the design matrix shown in Table 3 Thus through these 31 experimental runs, the linear, quadratic and two way interactive effects of the process parameters on the PERN can be estimated. Experiments were conducted at conducted at randomized manner to minimize the errors creeping into the experimental procedure. The design matrix and measured value of PREN is shown in Table 3.

Table 3 Experimental Design and PREN evaluation of Welded Specimens

10 11 12

20 21 22

-1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -2 2 0 0 0 0 0 0 0 0 0 0 0 0 0

-1 -1 1 1 -1 -1 1 1 -1 -1 1 1 -1 -1 1 1 0 0 -2 2 0 0 0 0 0 0 0 0 0 0 0

-1 1 17.5

-1 -1 18

-1 -1 17.8

-1 1 18.1

1 -1 18.2

1 1 18.5

1 1 18.6

1 -1 18.3

-1 -1 17.9

-1 1 18.3

-1 1 17.7

-1 -1 17.9

1 1 18.3

1 -1 18.1

1 -1 17.8

1 1 17.6

0 0 17.9

17.8 18.1 17.8 18.1

17.8 18.3 18.2 18.1 18.3 18.2 18.1 17.8 17.6

22.775

20.855

21.731

21.706

22.325

23.693

22.061

22.425

22.381

22.594

22.708

22.186

22.399

21.681

22.433

22.045

22.724

22.107

22.021

22.074

22.508

21.961

22.721

22.161

22.566

22.186

22.485

22.129

5. Development of mathematical model

The general equation of a quadratic polynomial, which relates the response variable 'y' and the process variable 'x' under investigation, is given by equation 3. [12]

- S i = 1

■ s i = 1

-Z Wj i< J

where b0=constant, bi=linear term coefficient, biquadratic term coefficient, and bij=interaction term coefficient

The coefficients of the quadratic equation were calculated using statistical software MINITAB R14 [13]. The statistical mathematical model was developed after eliminating the coefficients with least significant. This was done by conducting student's t test. According to this test when the calculated value of 't' corresponding to the coefficients exceeds the standard tabulated value for the probability criterion kept at 0.75, the coefficient becomes significant [14]. The mathematical model was given by Eq.4

PREN = 22.280 - 0.074(0. - 0.104(V) + 0.075(I) + 0.166(Q)- 0.146(0.1 -0.219(0.2-0.317(0.3 + 0.192(VI) + 0.094(VQ) + 0.103(IQ) (4)

The adequacy of the model was tested using the analysis of variance technique (ANOVA). The calculated F - ratio [15] of the model developed does not exceed the standard value for a desired level of confidence, i.e., 95% and the calculated value of R - ratio of the model developed is above the standard tabulated value for the same confidence level. Hence the model is considered to be adequate. The results of ANOVA are presented in Table 4. It is evident from the table that the model is adequate.

Table 4 adequacy of the model

First order term Second order term Lack of fit Error term

Parameter - F ratio R ratio

_SS_DOF_SS_DOF SS_DOF SS_DOF_

PREN 1.185 4 0.405 10 1.572 14 0.318 7 2.474 10.584

SS - Sum of Squares, DOF- Degree of Freedom Mean Sum of Squares = Sum of Square Terms/DOF F ratio = Ms of Lack of Fit/ Ms of Error Terms

R ratio = Ms of First Order Term & Second Order Term/ MS of Error Term Percentage Points of the F Distribution for PREN F ratio (14, 7, 0.05) = 2.76 R ratio (10, 7, 0.05) = 3.14

6. Results and discussion

A mathematical model was developed to predict PREN by relating it with welding parameters such as welding current, welding speed, welding gun angle and gas flow rate. The direct effect of these welding parameters on the PREN was studied using the developed empirical model. The direct effects of these process parameters on PREN were calculated and plotted are shown in Figs. 3, 4, 5, 6, 7 and the cause and effect were analyzed. The trends of the potted graph of the direct effect of these process parameters helps to determine which parameter are statistically significant. The effect of all the parameters on PREN is discussed below.

6.1 Effect of welding process parameters on PREN

Fig. 3. effect of welding gun angle on PREN

Fig. 4. effect of welding current on PREN

Fig. 5. effect of welding speed on PREN

Fig. 6. effect of shielding gas flow rate on PREN

Figure 3 depicts the direct effect of welding gun angle on PREN. From the figure it is observed that the value of pitting resistance equivalent number decreases as the welding gun angle increases. This is due to the reason, at lower gun angles the exposure of the weld metal to the arc is less which decreases the heat input and increases the cooling rate. This results in decreasing the critical pitting temperature. The decrease in critical pitting temperature results in increase in pitting resistance equivalent number. At higher gun angles, cause the heat input to increase which decreases the cooling rate and increases the critical pitting temperature. The increase in pitting temperature decreases the pitting resistance equivalent number.

Figure 4 depicts the direct effect of welding current on PREN. From the figure it is observed that the value of pitting resistance equivalent number decreases as the current increases. This is due the fact that welding current is directly proportional to the heat input. At lower current, heat input is low which increases the cooling rate. This results in decreasing the critical pitting temperature. The decrease in critical pitting temperature results in increase in pitting resistance equivalent number. At higher currents, there is an increase in heat input which decreases the cooling rate and increases the critical pitting temperature. The increase in pitting temperature decreases the pitting resistance equivalent number.

Figure 5 depicts the direct effect of welding speed on PREN. From the figure it is observed that the value of pitting resistance equivalent number increases as the welding speed increases. This is due to the reason; welding speed is inversely proportional to the heat input. At lower speeds, heat input is high which decreases the cooling rate. This results in increasing the critical pitting temperature. The increase in critical pitting temperature results in lower pitting resistance equivalent number. At higher speeds, there is a decrease in heat input which increases the cooling rate and decreases the critical pitting temperature. The decrease in pitting temperature increases the pitting resistance equivalent number.

Figure 6 depicts the direct effect of shielding gas flow rate on PREN. From the figure it is observed that the value of pitting resistance equivalent number increases as the gas flow rate increases. This is due to the reason; shielding gas always carries away some heat during welding. The amount of heat carried is lesser at lower levels and increases as the gas flow rate increases. Hence the heat input is high at lower levels of gas flow rate than at higher levels. Due to this effect there is a decrease in critical pitting temperature at higher levels of gas flow rate which increases PREN and increase in critical pitting temperature at lower levels of gas flow rate which decreases PREN.

6.2 Study on the microstructure of the welded specimens

The metallographic studies were carried out on the welded specimens obtained when the welding gun angle is maintained at their maximum and minimum levels whereas the other parameters are maintained at their middle levels.

Fig.7. microstructure of the base metal

Figure 7 shows the micro graph of the base metal of AISI202 grade stainless steel plate. The AISI202 stainless steel base metal consists of equiaxed austenite grains. There are some small amounts of dispersed carbide particles at boundaries of austenite grains. Hence the base metal has fully austenitic structure with small amount of dispersed carbide particles at the grain boundaries.

Figure 8 shows the microstructure of the specimens welded when the welding gun angle is at 50° and the other parameters such as welding current, welding speed and shielding gas flow rate are maintained at 90 amps, 190 mm/min and 15 liters/min. It can be observed from the microstructure of specimen that the base metal has equiaxed grains of austenite. The area of heat affected zone is relatively small. This is because exposure of weld metal to the arc is less at this gun angle. The weld metal has coarse grains and major portion of austenite is retained. The tempered martensite is distributed coarsely throughout the zone. There is also presence of ferrite in the microstructure of the weld metal. Figure 9 shows the microstructure of the specimen welded keeping welding gun angle at 90° and other parameters namely: welding current, welding speed and shielding gas flow rate at 90 amps, 190 mm/min and 15 liters/min.

Fig.8. microstructure of the weld metal, HAZ and base metal of specimen with gun angle at lower level

Fig. 9. microstructure of the weld metal, HAZ and base metal of specimen with gun angle at higher level

From the microstructure, it can be observed that area of the heat affected zone is more compared to the specimen welded using 50° gun angle. This is because; exposure of the weld metal to the arc is more at 90° gun angle. It can also be observed from the microstructure of the weld metal that there is less coarsening in the weld metal compared to the previous specimen. The microstructure of the weld metal also reveals that major portion of austenite has been retained. There is a small presence of tempered martensite throughout the weld zone. There is also presence of ferrite in the austenite matrix.

7. Conclusions

The following conclusions were arrived at based on the present investigations.

1. The welding current and welding gun angle has strong negative effect on pitting corrosion as PREN values decreases as the above parameters increases.

2. The welding speed and shielding gas flow rate has strong positive effect on pitting corrosion as the PREN value increases as the above parameters increases.

3. The base metal of AISI202 grade stainless steel has equiaxed grains of austenite with presence of small carbide particles at the grain boundaries

4. When the welding gun angle is maintained at low level of 50°, the area of heat affected zone decreases due to less exposure of weld metal to the arc. In the weld metal zone coarse grains are observed and major portion of austenite is retained. The tempered martensite is distributed coarsely throughout this zone

5. When the welding gun angle is maintained at high level of 90°, there is an increase in area of the heat affected zone due to more exposure of the weld metal to the arc. Less coarsening is observed in the weld metal. Major portion of austenite is retained along with a small amount of ferrite in the austenite matrix. There is a presence of tempered martensite throughout the zone

6. It can be finally concluded that AISI202 grade stainless steel has sufficient pitting corrosion resistance and moderate grain growth in heat affected zone. In the weld metal zone it has major portion of austenite with small amount of ferrite and tempered martensite. The presence of martensite increases hardness of the weld metal. The presence of ferrite in austenite matrix shows that the weld metal has sufficient strength, toughness, phase stability and resistant to corrosion.

[1] H. B. Cary, Modern Welding Technology, Prentics Hall, New Jersey, 1989

[2] Z. Samati, Automatic Pulsed MIG Welding, Metal Construction. (1986) 38R- 44.

[3] P.J. Konkol and G. F. Koons, Optimization of Parameters for Two Wire AC- ACSAW, American Welding Journal, 27 (1978) 367s - 374s.

References

[4] Fong - Yuan Ma, Corrosive Effects of Chlorides on Metals, Pitting Corrosion, Nasr Bensalah (Ed.),Intech open (2012)

[5] Kondapalli Siva Prasada, Chalamalasetti Srinivasa Rao and Damera Nageswara Rao, Effect of welding parameters on pitting corrosion rate of pulsed current micro plasma arc welded AISI 304L sheets in 1N HCl. Journal of Computational and Applied Research in Mechanical Engineering, 1 (2013) 1 -11.

[6] K. Siva and N. Murugan, A Study on the Influence of PTAW Process Parameters on Pitting Corrosion Resistance of Nickel based Overlays, Procedia Engineering. 64 (2013) 1147 - 1156.

[7] A. S. Subodh Kumar, A.S. Shahi, Effect of Heat Input on the Microstructure and Mechanical Properties of Gas Tungsten Arc Welded

AISI304 Stainless Steel Joints, Materials and Design. 32 (2011) 3617 - 3623.

[8] Huaibei Zheng, Xiaoning Ye, Laizhu Jiang, Baosen Wang, Zhenyu Liu and Guodong Wang, Study onMicrostructure of Low Carbon 12% Chromium Stainless Steel in High Temperature Heat Affected Zone, Materials and Design. 31 (2010) 4836 - 4841.

[9] International Stainless Steel Forum (ISSF), New 200 Series Steels - An Opportunity or Threat to the Image of Stainless Steel, world stainless .Org

[10] D.C. Montgomery and E. F. Peck, Introduction to Linear Regression Analysis, John Wiley and Sons, New York, 2005.

[11] R. Sudhakaran, V. Velmurugan, P.S. Sivasakthivel and M. Balaji, Prediction and Optimization of Depth of Penetration for Stainless Steel Gas Tungsten Arc Welded Plates using Artificial Neural Networks and Simulated Annealing Algorithm, Neural Computing and Applications. 22 (2013) 637 - 649.

[12] W.G. Cochran and G. M. Cox, Experimental Designs, John Wiley and Sons, New York, 1987.

[13] Mini Tab R(14), Minitab Inc, USA, 2004.

[14] R. Sudhakaran, V. Velmurugan, P.S. Sivasakthivel and M. Balaji, Modeling and Analysis of Ferrite Number of Stainless Steel Gas Tungsten Arc Welded Plates using Response Surface Methodology, International Journal of Advanced Manufacturing Technology. 64 (2013) 14871504.

[15] R. Sudhakaran, V. Velmurugan, P.S. Sivasakthivel, The Effect of Process Parameters on Depth of Penetration in Gas Tungsten Arc Welded 202 Grade Stainless Steel Plates using Response Surface Methodology, The Journal of Engineering Research. 9 (2012) 64 - 79.