Study on laser welding of austenitic stainless steel by varying incident angle of pulsed laser beam

Study on laser welding of austenitic stainless steel by varying incident angle of pulsed laser beam

Optics and Laser Technology 94 (2017) 296–309 Contents lists available at ScienceDirect Optics and Laser Technology journal homepage: www.elsevier.c...

5MB Sizes 0 Downloads 37 Views

Optics and Laser Technology 94 (2017) 296–309

Contents lists available at ScienceDirect

Optics and Laser Technology journal homepage: www.elsevier.com/locate/jolt

Full length article

Study on laser welding of austenitic stainless steel by varying incident angle of pulsed laser beam Nikhil Kumar a,⇑, Manidipto Mukherjee b, Asish Bandyopadhyay a a b

Mechanical Engineering Department, Jadavpur University, Kolkata 700032, India Mechanical Engineering Department, C.V. Raman College of Engineering, Bhubaneswar 752 054, India

a r t i c l e

i n f o

Article history: Received 14 August 2016 Received in revised form 2 April 2017 Accepted 9 April 2017

Keywords: Laser welding Response surface methodology Modeling and optimization Microstructure Mechanical properties

a b s t r a c t In the present work, AISI 304 stainless steel sheets are laser welded in butt joint configuration using a robotic control 600 W pulsed Nd:YAG laser system. The objective of the work is of twofold. Firstly, the study aims to find out the effect of incident angle on the weld pool geometry, microstructure and tensile property of the welded joints. Secondly, a set of experiments are conducted, according to response surface design, to investigate the effects of process parameters, namely, incident angle of laser beam, laser power and welding speed, on ultimate tensile strength by developing a second order polynomial equation. Study with three different incident angle of laser beam 89.7 deg, 85.5 deg and 83 deg has been presented in this work. It is observed that the weld pool geometry has been significantly altered with the deviation in incident angle. The weld pool shape at the top surface has been altered from semispherical or nearly spherical shape to tear drop shape with decrease in incident angle. Simultaneously, planer, fine columnar dendritic and coarse columnar dendritic structures have been observed at 89.7 deg, 85.5 deg and 83 deg incident angle respectively. Weld metals with 85.5 deg incident angle has higher fraction of carbide and d-ferrite precipitation in the austenitic matrix compared to other weld conditions. Hence, weld metal of 85.5 deg incident angle achieved higher micro-hardness of 280 HV and tensile strength of 579.26 MPa followed by 89.7 deg and 83 deg incident angle welds. Furthermore, the predicted maximum value of ultimate tensile strength of 580.50 MPa has been achieved for 85.95 deg incident angle using the developed equation where other two optimum parameter settings have been obtained as laser power of 455.52 W and welding speed of 4.95 mm/s. This observation has been satisfactorily validated by three confirmatory tests. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction Austenitic stainless steel has a wide range of applications in nuclear structural fabrication, valve bodies and vessel internals because of their excellent mechanical properties. Joining process is required for this and laser welding is such a joining process. Laser welding has several advantages when compared to the conventional welding. It is non-contact type and its localized and narrow heat zone can create high quality result. Common re-working and after-work procedures are no more required. Laser welding has been widely applied in various industries including automotive, microelectronics, aerospace, medical, optoelectronics, microsystems etc. Kuryntsev and Gilmutdinov [1] have studied the laser welding of type 321 stainless steel and have found that the defo-

⇑ Corresponding author. E-mail addresses: [email protected] (N. Kumar), [email protected] com (M. Mukherjee), [email protected] (A. Bandyopadhyay). http://dx.doi.org/10.1016/j.optlastec.2017.04.008 0030-3992/Ó 2017 Elsevier Ltd. All rights reserved.

cused laser beam has increased the volume of weld pool that in turn to reduce the requirement for preparation of edge and gap between workpieces. Yan et al. [2] have investigated the microstructure and mechanical properties of tungsten inert gas, laser and laser-TIG hybrid welded 304 stainless steel. They have found that laser welded sample has highest tensile strength and smallest dendrite size than all other. Experimental investigation on dissimilar pulsed Nd:YAG laser welding of AISI 420 stainless steel to kovar alloy has been reported in [3] and they have found that the start of solidification in the kovar side of weld zone has occurred by means of epitaxial growth. In the work of Ai et al. [4] a defect-responsive optimization method for the fiber laser butt welding of dissimilar materials has been investigated. The genetic algorithm (GA) is applied to solve the model. The dissimilar laser welding of AISI 316L stainless steel to Ti6-Al4-6V alloy via pure vanadium interlayer has been studied by Tomashchuk et al. [5]. The effects of laser power, scanning speed, defocus distance, beam incident angle and line energy on weld bead geometry and

297

N. Kumar et al. / Optics and Laser Technology 94 (2017) 296–309

shearing force of laser welded dissimilar AISI 304L and AISI 430 stainless steel has been investigated in [6]. An investigation has been made by Keskitalo et al. [7] to study the laser welding of duplex stainless steel with nitrogen gas as shielding gas. The result suggests that nitrogen increases austenite levels in the weld metal and improved toughness levels. A study of simulation of laser butt welding of AISI 316L stainless steel sheet using various heat sources has been performed and experimental validation also has been done in [8]. The simulated thermal cycles, residual stress and distortion has been validated by experiments. In the research of Chen et al. [9], the influence of processing parameters on the characteristic of stainless steel/copper laser welding has been studied. Hao et al. [10] have investigated the effects of beam oscillating parameters on the weld morphologies. They have found that the difference in cross-section width from top to the lower gradually has reduced to disappear with the increase in oscillating frequency. An attempt has been made to improve the quality of the weldment between nickel titanium (NiTi) and AISI 316L stainless steel wires in [11]. A pulsed wave Nd:YAG laser system has been used for the welding of CP Ti and stainless steel sheets and the effect of pulse profiles used in laser welding on weld appearance, weld geometry, microstructure, hardness variation, joint strength and failure mode of weld have been investigated in [12]. Tan and Shin [13] have studied the multi-scale modeling of solidification and microstructure development in laser keyhole welding process for austenitic stainless steel. The model predictions are validated with the experimental results and the effects of the welding parameters are analyzed based on numerical and experimental results. Optimization of CO2 laser welding of DP/TRIP steel sheets using statistical approach has been conducted in [14]. In the article of Matsunawa et al. [15] the observation of keyhole as well as weld pool dynamics and their related phenomena to reveal the mechanism of porosity formation and its suppression methods have been studied. A numerical simulation model has been developed by Cho et al. [16] to study the temperature profile characteristics of weld bead and molten pool dynamics of high power disk laser welding process. Numerical and experimental study of molten pool formation during continuous laser welding of AZ91 magnesium alloy has been reported in [17]. A mathematical model has been developed by Zhou et al. [18] to analyze the heat transfer, fluid flow and keyhole dynamics during pulsed keyhole laser welding. A numerical and experimental investigation of laser welding of titanium alloy (Ti6Al4V) for modeling the temperature distribution to predict the heat affected zone, depth and width of the molten pool has been analyzed in [19]. Shanmugarajan et al. [20] have studied the effect of process parameters such as laser power, welding speed, shielding gas and laser beam mode on microstructure and mechanical properties of laser welded sample of type 304B4 borated stainless steel. Torkamany et al. [21] have analyzed the pulsed Nd:YAG laser welding of pure niobium plate to titanium alloy Ti-6AL-4V sheet in butt joint. The effect of pulsed Nd:YAG laser welding parameters and subsequent post-weld heat treatment on microstructure and hardness of AISI 420 stainless steel have been studied by Baghjari and Mousavi [22]. In the research of Chen et al. [23] the effect of laser-beam offsetting on microstructural characteristics and fracture behaviour of the laser butt joint of titanium alloy have been studied. An experimental procedure has been developed by Atabaki et al. [24] to join thick advance high strength steel plates by using the hybrid laser/arc welding (HLAW) process. An investigation has been made by Sun et al. [25] to analyze the laser butt joint of Al/steel dissimilar materials. Within scope of literature review, it has been observed that almost limited or no information is available on the effect of laser incident angle on the mechanical and microstructural properties of pulsed laser welding of AISI 304 stainless steel sheets in a butt joint configuration. It is one of the important parameter that may be co-

related with responses. This research aims to find the optimum incident angle for which the laser optic lens will be protected and also to understand the physical mechanisms responsible for the joint quality of the laser beam butt welding process of stainless steel plates. In the present work, 3 factors-5 levels experiments have been planned using response surface methodology (RSM) design matrix and analyzing the responses of interest by developed mathematical models based on experimental results. The second order mathematical equations have been developed for predicting the desired weld quality. In addition to statistical evaluation of the welded joints, metallurgical and mechanical analyses have been carried out on laser welded three specimens with incident angle 89.7 deg, 85.5 deg and 83 deg incident angle. A 3-D responses surface and contour plots have been developed to find the combined effect of input parameters on responses.

2. Response surface methodology Response surface methodology is a useful design of experiment method that is gaining popularity. This includes a review of basic experimental designs for fitting linear response surface models, in addition to a description of methods for the determination of optimum operating conditions. The steps of response surface methodology are: (i) Developing experimental strategy for selecting independent variables. (ii) Statistical modeling to build an approximate relationship between the response and process variables. (iii) Optimization for finding values of process variables producing desirable values of the response. When all the independent variables are measurable, controllable and continuous during experiments, response surface, y can be expressed with negligible error by:

y ¼ f ðxÞb þ  0

ð1Þ

where x = (x1, x2, . . ., xk). 0 f ðxÞ = a vector function of p elements. b = a vector of p unknown constant coefficients.  = a random experimental error assumed zero mean. In RSM, an approximate model is needed to develop for the true response surface. The approximated model is constructed utilizing observed data from the process or system. Multiple regression analysis is commonly used for this. Usually, a second-order polynomial equation is used in RSM, which is given by

y ¼ b0 þ

k k X XX X bi xi þ bij xi xj þ bii x2i þ e i¼1

ð2Þ

i¼1

where parameters b0 bi, bij, bii are called regression coefficient for i = 0, 1, . . ., k and j = 0, 1. . .k. 2.1. Desirability function analysis It is an approach in which, individual responses are transformed to corresponding desirability values. Desirability value depends on acceptable tolerance range as well as target of the response. Unity is assigned, as the response reaches its target value, which is most desired situation. Beyond acceptable limit, desirability value assumes zero. In this study, individual desirability function posses one of the following two characteristics:

298

N. Kumar et al. / Optics and Laser Technology 94 (2017) 296–309

For goal of maximum, the desirability ðdi Þ will be defined by

di ¼

8 0; > > < > > :

yi Li Hi Li

wi

if responseðyi Þ 6 low v alueðLi Þ ; as responseðyi Þ v aries from lowðLi Þ to highðHi Þ if responseðyi Þ P high v alueðHi Þ

1;

ð3Þ For goal of minimum, the desirability will be defined by

di ¼

8 1; > > < > > :

Hi yi Hi Li

wi

if responseðyi Þ 6 low v alueðLi Þ ; as responseðyi Þv aries from lowðLi Þ to highðHi Þ if responseðyi Þ P high v alueðHi Þ

0;

ð4Þ A weight (w) can be assigned to a goal to emphasize the particular desirability function. Weights can be varied between 0.1 and 10. A weight greater than 1 gives more emphasis to the goal, while weights less than 1 give less emphasis. The simultaneous objective function, D, is a geometric mean of all transformed responses: 1  r r r  D ¼ d11  d12  . . .  d1n Rri ¼

n Y

Fig. 1. Schematic diagram of laser welding process and joint configuration.

! 1r r

di i

R

i

ð5Þ

top to bottom in the weld using Vickers’s microhardness testing machine (Make: LECO Co., USA; Model: LM248AT) at 50 gf load with 15 s dwell time. The measurement of tensile strength of welded samples has been conducted on Instron (Model-8801) as per ASTM E8 standard with strain rate of 2.8  104 s1. The schematic view of the tensile test specimen as per ASTM E8 with dimension is shown in Fig. 4.

i¼1

where n is the number of responses in the measure. Each response can be assigned an importance relative to the other responses. Importance ðr i Þ values varies from 1, the least important, to 5, the most important [26]. 3. Experimental set up and procedure

4. Results and discussions

AISI 304 type stainless steel has been chosen for experimental work, the dimensions of workpieces before welding 100 mm  20 mm  1.5 mm. The chemical composition of the material is shown in Table 1. The schematic diagram of laser welding process is shown in Fig. 1 and h indicates the incident angle. All the experiments have been conducted on JK600HP Nd:YAG laser generator (GSI, UK) integrated with ABB IRB 1410 robotic control. The welding of work piece has been conducted in pulse width of 5 ms and repetition rate of 25 Hz. Experimental set-up is given in Fig. 2. Samples are butt jointed and during welding technically zero gap between two sheets is maintained in each case. An argon gas jet emerges from the side nozzle which makes a fixed angle with the laser beam to avoid any external atmospheric contamination during welding. The laser beam has a spot size 0.75 mm. The ranges of input parameters are selected on the basis of trial experiments conducted by using one factor at a time approach. The chosen process parameters and their limits are given in Table 2. Eighteen experiments have been conducted as per central composite rotatable design (CCD) including 4 center points. Statistical software Design-Expert v10 has been applied to establish the design matrix. Fig. 3 shows a weld sample in butt joint configuration. Samples for the metallographic examinations have been prepared by polishing successively in 80, 120, 220, 320, 400, 1200, 1600, 2000 grade emery papers to remove the scratches. The compositions of the etchant are 2.4 gm. of CuCl2, 10 ml of 99% C2H5OH and 10 ml of 40% HCl. Furthermore, micro-hardness survey has been made on flat metallographic specimen across the joints and

The measured response is listed in Table 3. Design-Expert v10 software has been applied for analyzing the measured response and determining the mathematical model with best fit. The fitted quadratic polynomial model for response is statistically significant for the prediction within working range of welding parameters. Therefore, they will be used for further analysis. The maximum and minimum ultimate tensile strength are observed for sample no. 5 (P = 450 W, S = 5 mm/s and A = 85.5 deg) and sample no. 10 (P = 425 W, S = 5.5 mm/s and A = 83 deg) respectively. The metallurgical characteristics along with mechanical properties like hardness and tensile strength of weld sample no. 5, 10 and 16 have been presented in the following section. In general the laser welding is being performed with an incident angle of 89.7 deg, hence a comparison has been made between the sample no. 5 and 16 and sample no. 10 and 16. Since sample no 16 has been welded with an incident angle 89.7 (90 deg). 4.1. Development of mathematical model The adequacy of the developed model is tested using the sequential f-test, lack-of-fit test and analysis-of-variance (ANOVA) technique using the Design-Expert v10 software to obtain the bestfit model. The ANOVA tables also show the other adequacy measure R2, adjusted R2, adequacy precision R2 and predicted R2 for response is given in Table 3. The adequate precision compares

Table 1 Chemical compositions (wt.%) of AISI 304 stainless steel. Type

304 SS

Chemical composition C%

Si%

Mn%

P%

S%

Cr%

Ni%

Mo%

Cu%

Nb%

Al%

N%

0.079

0.2858

1.8

0.032

0.0194

18.56

8.20

0.265

0.292

0.0281

0.0063



299

N. Kumar et al. / Optics and Laser Technology 94 (2017) 296–309

Fig. 2. Robotic control laser welding set-up.

Table 2 Process control parameters and their limits. Parameters with units

Notation

Power, W Scanning speed, mm/s Incident angle, deg

P S A

Levels 2

1

0

+1

+2

407 4.16 81.29

425 4.50 83.00

450 5.00 85.50

475 5.50 88.00

492 5.84 89.70

Fig. 3. Top view of welded sample in butt joint configuration.

model terms are significant. The ‘‘lack-of-fit F-value” of 6.88 implies there is a 7.10% chance that a ‘‘lack-of-fit F-value” this large could occur due to noise. The ‘‘Predicted R2” of 0.9235 is in reasonable agreement with the ‘‘adjusted R2” of 0.9724 (i.e., the difference is less than 0.2). ‘‘Adequate precision” measures the signal to noise ratio. A ratio greater than 4 is desirable. The value of adequate precision of 25.13 indicates an adequate model. The model can be used to navigate the design space. The mathematical models for ultimate tensile strength, which can be used for prediction within same design space, are shown below: (a) In term of coded factors

UTS ¼ 575:434 þ 28:67P  15:135S þ 15:28A  26:10PA þ 13:52SA  55:71P2  83:83S2  22:69A2 Fig. 4. Laser welded sample for tensile test as per ASTM E8.

ð6Þ

(b) In term of actual factors

UTS ¼ 64732:16 þ 117:08P þ 2398:25S þ 760:81A the range of predicted value at the design points to the average predicted error [27]. The associated p-value of less than 0.05 for the model (i.e., p-value < 0.05, at 95% confidence level) indicates that the model terms are statistically significant. The lack-of-fit value of the model indicates non-significant, as this desirable. The ANOVA indicates that for the ultimate tensile strength model (Table 4), the laser power (P), welding speed (S), incident angle (A), interaction effect of laser power and incident angle (P  A), welding speed and incident angle (S  A), the quadratic effect of the laser power (P2 ), welding speed (S2 ), incident angle (A2 ) are the significant model terms. The interaction effect of laser power and welding speed (P  S) is not significant and thus, eliminated by backward elimination process to improve model adequacy. The ANOVA result for reduced quadratic model is shown in Table 5. The model F-value of 75.98 implies the model is significant. There is only a 0.01% chance that an F-value this large could occur due to noise. Values of ‘‘prob > F” less than 0.05 indicate

 0:41PA þ 10:81SA  0:08P2  335:31S2  3:63A2

ð7Þ

4.2. Validation of the developed model The developed response surface equation, derived from multiple regression analysis has been validated by conducting confirmatory tests. Three confirmatory experiments have been conducted and welding conditions have been chosen randomly. The tested results of experiments are presented in Table 6. It is obtained from Table 6 that there is a small error percentage between experimental and the predicted values form developed regression equation, which shows that the developed model can yield nearly accurate results. Fig. 5 shows the relationship between the actual and predicted values of responses. This figure also indicates that the developed model is adequate and predicted results are in good agreement with measured data.

300

N. Kumar et al. / Optics and Laser Technology 94 (2017) 296–309

Table 3 Central composite design for actual factors and measured experimental results. Experiment no.

Power, W

Welding speed, mm/s

Incident angle, deg

Ultimate tensile strength, MPa

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

450 425 492 450 450 475 475 450 425 425 450 425 475 475 450 450 450 407

5.00 4.50 5.00 5.84 5.00 4.50 5.50 5.00 5.50 5.50 5.00 4.50 5.50 4.50 5.00 5.00 4.15 5.00

85.5 88.0 85.5 85.5 85.5 88.0 83.0 85.5 88.0 83.0 85.5 83.0 88.0 83.0 81.2 89.7 85.5 85.5

577.81 443.82 466.94 334.00 579.26 429.66 408.23 578.42 404.30 308.64 565.23 369.48 432.23 492.48 491.16 537.28 348.60 374.70

Table 4 ANOVA for the fitted quadratic polynomial model for UTS of welded samples (before elimination). Source

Sum of squares

Df

Mean square

F value

p-value Prob > F

Remark

Model P S A PS PA SA P2 S2 A2 Residual Lack of fit Pure error Cor total

1.327E+005 11222.45 3125.26 3190.65 43.62 5450.72 1461.78 39262.43 88886.75 6512.97 1920.05 1786.99 133.06 1.346E+005

9 1 1 1 1 1 1 1 1 1 8 5 3 17

14741.39 11222.45 3125.26 3190.65 43.62 5450.72 1461.78 39262.43 88886.75 6512.97 240.01 357.40 44.35

61.42 46.76 13.02 13.29 0.18 22.71 6.09 163.59 370.35 27.14

<0.0001 0.0001 0.0069 0.0065 0.6811 0.0014 0.0388 <0.0001 <0.0001 0.0008

Significant Significant Significant Significant Not significant Significant Significant Significant Significant Significant

8.06

0.0581

Not significant

R2 = 0.9857 Adjusted R2 = 0.9697 Predicted R2 = 0.8911 Adequate precision = 22.79

Standard deviation = 15.49 Mean = 452.35 Coefficient of variance = 3.42 Predicted residual error of sum of squares (PRESS) = 14655.52

Table 5 ANOVA for the fitted quadratic polynomial model for UTS of welded samples (after backward elimination). Source

Sum of squares

Df

Mean square

F value

p-value Prob > F

Remark

Model P S A PA SA P2 S2 A2 Residual Lack of fit Pure error Cor total

1.326E+005 11222.45 3125.26 3190.65 5450.72 1461.78 39262.43 88886.75 6512.97 1963.67 1830.61 133.06 1.346E+005

8 1 1 1 1 1 1 1 1 9 6 3 17

16578.62 11222.45 3125.26 3190.65 5450.72 1461.78 39262.43 88886.75 6512.97 218.19 305.10 44.35

75.98 51.44 14.32 14.62 24.98 6.70 179.95 407.39 29.85

<0.0001 <0.0001 0.0043 0.0041 0.0007 0.0293 <0.0001 <0.0001 0.0004

Significant Significant Significant Significant Significant Significant Significant Significant Significant

6.88

0.0710

Not significant

Standard deviation = 14.77 Mean = 452.35 Coefficient of variance = 3.27 Predicted residual error of sum of squares (PRESS) = 10289.67

4.3. Effect of incident angle on weld pool geometry The macro view (at nominal magnification) of typical fusion zones (FZ) of 304 SS after pulsed Nd:YAG laser welding at 89.7 deg, 85.5 deg and 83 deg incident angle has been presented

R2 = 0.9854 Adjusted R2 = 0.9724 Predicted R2 = 0.9235 Adequate precision = 25.13

in Fig. 6. The incident angles have been considered as per the DoE (Table 3) and sample no. 5, 10 and 16 have been selected for analysis on the basis of obtained tensile strength responses of 304 SS weld joints (Table 3). Weld metals having maximum and minimum UTS have been considered along with one sample having

301

N. Kumar et al. / Optics and Laser Technology 94 (2017) 296–309

Table 6 Optimization validation test results. Exp. no

Welding condition

UTS (MPa)

P (W)

S (mm/s)

A (deg)

1

450

4.5

85.5

Actual Predicted jError%j

519.20 506.00 2.50

2

450

4.5

89.7

Actual Predicted jError%j

472.30 445.70 5.63

3

475

5.0

83.0

Actual Predicted jError%j

439.30 426.00 2.95

Fig. 5. Plot of actual vs. predicted of UTS results.

maximum angle of incidence (with tensile strength between maximum and minimum) for the analysis and it is also assumed that the rest should have properties in between these extreme limits. Fig. 6 clearly pointed out typical variation in surface depression and three distinguishable regions marked as fusion zone 1 (FZ1), fusion zone 2 (FZ2) and fusion zone 3 (FZ3) extended from fusion boundary (FB) to the weld center. These distinguishable fusion zones have been experienced probably due to the periodic pulse oscillation (on time and off time) of the pulsed Nd:YAG laser power source where the base metal initially melts at the on time or the high energy pulse duration and get partially solidified during the off time or low energy pulse duration. Periodic pulse oscillation (on time and off time) results in higher turbulence in weld pool which may reduce the temperature in front of the solidifying interface resulted in local undercooling [28]. The undercooling is a function of the dendrite tip radius and thus, as solidification growth velocity increases (tip radius decreases) the local under cooling of the tip also increases. The local undercooling can influence both segregation and solidification mode and at high solidification growth rates (critical growth rate), sufficient dendrite tip undercooling occurs to promote a shift in solidification mode from primary ferrite to primary austenite [29]. Despite the degree of growth rate among the weld metals, it can be postulated that the pulsed Nd:YAG laser welding is generally a rapid cooling process and thus all the weld metals should experience a certain degree of local undercooling and shift in solidification which ultimately creates the separate solidification boundaries in the fusion zone (Fig. 6). The dendritic structure growing from the FB has been degenerated at the solidification boundaries and favours the growth of new grains at the weld center in FZ2 and FZ3 (Fig. 6a). The new grains have been found to be regularly oriented towards the direction of higher thermal gradient (from FZ2 to FZ3) and generate finer grain structure at the FZ3 of weld metal produced at 89.7 deg incident angle [30]. However, Fig. 6(b) and (c) shows that the solidification structure and the weld pool geometry have been significantly altered with the deviation in incident angle. As the incident angle with

(a)

(b)

FZ3 FZ2

FZ3

FZ1

FZ2 FZ1

(c) FZ3 FZ2 FZ1

Fig. 6. Macrographs of 304 SS pulsed Nd:YAG laser welded joint cross-sections at (a) 89.7 deg, (b) 85.5 deg and (c) 83 deg, incident angle show solidification boundaries and dendritic growth pattern.

302

N. Kumar et al. / Optics and Laser Technology 94 (2017) 296–309

(b) Welding direction

(a)

Welding direction

(c) Welding direction

Fig. 7. Macrographs of 304 SS pulsed Nd:YAG laser welded joint bead surface at (a) 89.7 deg, (b) 85.5 deg and (c) 83 deg, incident angle show variation in ripples.

Fig. 8. Schematic representations of weld pool geometry along the Y-Z and X-Y plane for different incident angle of laser beam.

the surface decreases to 85.5 deg (Fig. 6b), the overall weld pool becomes wider and FZ3 acquires large overlapped area among the fusion zones. Parallel epitaxial columnar dendritic growth from the adjacent solidification boundary of FZ2 towards weld center has been observed in the FZ3. Further decrease in incident angle (83 deg) produces lack of penetration with high surface depression and shift in weld center from the joint interface (Fig. 6c). These variations in solidification patterns may be explained by the change in weld pool shape and size incurred due to the deviation in incident angle of laser beam. In general, laser welding results in the formation of a deep and narrow vapour cavity called keyhole at the joint interface due to high energy density which creates strong evaporation of alloying elements [31,32]. The material evaporates and the metal vapours leaving the surface exert a recoil pressure on the weld pool surface. When no material is added to the weld pool, the volume change

remains zero provided the average densities of the initial and the final phases are similar, thus, the top surface deformation is expected (Fig. 6). However, a large top surface deformation has been observed for partial penetration of weld metal at low incident angle (Fig. 6c) and this may be due to the volume expansion during melting followed by rapid solidification. Furthermore, Fig. 6(a)–(c) also show that the weld pool has a spread near the top and bottom surface for full penetration welds and for partial penetration weld narrow weld pool having only a wide spread near the top surface is observed. The weld pool widens near the top surface as well as near the bottom due to convective heat transfer or Marangoni convection [33,34]. During melting liquid metal being pushed outward from the weld center due to large Marangoni force creates a wide spread weld pool. The rapid movement of concentrated heat source along the weld line freezes the weld pool in its deformed state due to shrinkage creating surface depression. The weld pool spread is

303

N. Kumar et al. / Optics and Laser Technology 94 (2017) 296–309

(a)

(b)

BM

(c)

(d)

Fig. 9. SEM micrographs of (a) AISI 304 SS base metal; and fusion zones after pulsed Nd:YAG laser welding at (b) 89.7 deg, (c) 85.5 deg and (d) 83 deg, incident angle.

generally maximum at the top surface where the liquid metal flows outward from the keyhole and carries heat away from the center which creates a wide and elongated weld pool behind the heat source (opposite to the welding direction) as shown in Fig. 7. Clear variation in solid-liquid interface boundaries or isotherms (shown as yellow dotted line in Fig. 7) signifies the variation in weld pool shape/geometry among the weld metals. These variations are schematically presented in Fig. 8. Weld pool shape along the Y-Z plane surround the keyhole varies significantly with the decrease in incident angle. In general, the weld pool length decreases with distance from the top surface due to viscous effects and the weld pool is increasingly influenced by conduction heat transfer. At the 89.7 deg incidence angle, the keyhole generated almost vertically and the isotherms are compressed in the front of the heat source and elongated behind it due to the effect of moving heat source. The temperature within the keyhole is highest at the bottom and lowest near the top surface. This temperature gradient along the keyhole surface drives liquid metal flow from the hot keyhole bottom to the top, resulting in a fluid flow pattern in the weld pool where hot liquid metal moves along the keyhole walls to the top, moving outward from there and finally coming back inward and down along the solid-liquid boundary [35]. However, when incident angle decreases, an additional fluid flow may be occurred in front of the heat source where a part of the hot liquid metal moves along the inclined keyhole wall towards the top surface and is pushed forward due to the forward movement of the heat source. Extension of weld pool, stretches the overall weld pool volume and forces to change the shape, thus, with decrease in incident angle weld pool shape at the top surface (X-Y plane) has been altered from semi-spherical shape to tear drop shape. At 85.5 deg incident angle, weld pool shape is elongated along the welding direction and compressed at the edges to become nearly elliptical probably due to the azimuthal oscillation of the heat source [36]. Elliptical weld pool facilitates straight columnar dendritic grain growth during solidification (Fig. 6b). The weld pool further changes to tear drop shape at 83 deg incident angle due to higher welding speed and less power (Table 3), and it is associated with partially curved columnar dendritic grain

growth. The lower incident angle predominantly pushed the molten liquid metal along the keyhole wall towards the top surface which widens the weld pool width. However, comparatively higher inclination of heat source against the welding direction reduces the effective heat distribution at the bottom of the keyhole, resulting in partial penetration (Fig. 6c). 4.4. Microstructural analysis The microstructural analysis of base metal and three welded specimens (pulsed Nd:YAG laser welded for sample no. 5, 10 and 16 with incident angle of 85.5 deg, 83 deg and 89.7 deg) have been carried out using scanning electron microscopy (SEM). Fig. 9(a) shows the typical microstructure of 304 SS base metal mainly consisting fully austenitic structure along with few annealing twins. The detail micrographs of fusion zones (FZ) welded at 89.7 deg, 85.5 deg and 83 deg have been given in Fig. 9(b)–(d) respectively. Epitaxial columnar dendritic growth has been observed in micrographs except FZ of 89.7 deg incident angle where equiaxed grains or planar grain structure has been generated due to the decomposition of dendritic grains (Fig. 9b). Epitaxial columnar dendritic growth in other two FZs of 85.5 deg and 83 deg are generally begun at the initial stages of solidification from the fusion boundary towards the weld center (Fig. 9c and d). Furthermore, FZs of 85.5 deg and 83 deg show higher fraction of lacy d ferrite at the dendrite core surrounded by interdendritic c-phase compare to FZ of 89.7 deg incident angle which shows major proportion of vermicular and equiaxed d ferrite on the austenite matrix. The primary solidification mode (PSM) of all the weld metals is based on the Creq/Nieq ratios of the 304 SS base metal (i.e. 1.68) because no filler metal has been used in the study and can be generally categorised as FA mode [37–40]: FA mode: Nieq < 1.95

L ? L + d ? L + d + c ? d + c ? c:

1.48 < Creq/

The weld metals of 304 SS are generally solidified in FA mode of solidification which signifies the precipitation of primary ferrite,

304

N. Kumar et al. / Optics and Laser Technology 94 (2017) 296–309

(b) (a)

BM

WM

BM

FZ3

FZ2

FZ1

(c)

Fig. 10. Micro-hardness plots of welded joints (a) across the fusion zone; (b) from top to bottom and (c) the average micro-hardness values of weld metals at different incident angle.

Table 7 Tensile test results of base metal and welded joints. Sample specification

rYS (MPa)

rUTS (MPa)

eu (%)

DrUTS-YS (MPa)

n

K (MPa)

Location of fracture

BM 304 SS Sample no. 16 (89.7 deg) Sample no. 5 (85.5 deg) Sample no. 10 (83 deg)

352.06 324.42 321.86 240.41

632.67 537.28 578.42 308.65

49.77 35.87 47.25 3.83

280.61 212.86 256.56 68.24

0.338 0.397 0.469 0.327

253.9 176.10 142.88 137.52

BM FZ/HAZ FZ/HAZ FZ

rYS = yield strength; rUTS = ultimate tensile strength; eu = uniform elongation; DrUTS-YS = stress increment; n = strain-hardening exponent; K = strength coefficient.

plus three phase reaction (L + d + c) at the terminal solidification stage, and d ? c continuing below the solidus line [41]. The complexity of FA mode of solidification arises from the fact that, after a certain amount of primary solidification as ferrite, austenite precipitation occurs through a peritectic/eutectic reaction (L + d ? L + d + c) and changes the direction of micro-segregations. Moreover, the subsequent solid-state transformation of d ? c causes additional partitioning, altering the existing concentration profiles [42]. Thus, the final microstructure of FZs after complete solidification should consists of vermicular and lacy d-ferrite at the dendrite core enveloped by an interdendritic layer of austenite. The variation in morphologies in FZs can be explained by the variation in weld pool shape and growth kinetics. The magnitude

and direction of the maximum thermal gradient change continuously from the fusion line to the weld centreline in an elliptically shaped weld pool (Fig. 8). Since the average growth direction during solidification of a weld pool is approximately normal to the solid-liquid interface along the maximum temperature gradient, a particular columnar grain will not be favourably oriented during the entire solidification process. Therefore, many unfavourably oriented grains at the interface may become more favourably orientated before they are completely eliminated and thus they may survive and continue to grow towards the centerline, resulting in a finer fusion zone grain structure (Fig. 9b and c). For a tear drop shaped weld pool, there is almost an invariant direction of maximum thermal gradient at all points on the pool edge from the

N. Kumar et al. / Optics and Laser Technology 94 (2017) 296–309

305

Fig. 11. True stress–true strain curves of (a) 304 SS base metal and weld metals and (b) tensile specimen after fracture shows the fracture path of sample of 85.5 deg incident angle.

Fig. 12. Perturbation plot showing the effect of all factors on the UTS.

fusion boundary to the weld centreline which promotes coarse columnar grain growth as a result of favourable and uninterrupted grain growth towards the center (Fig. 9d). Furthermore, degree of grain fineness, type of grain growth and the thermal conditions in the immediate vicinity of the solid–liquid interface controls the solute segregation process during solidification which ultimately controls the microstructures and finally the properties. When the weld pool is nearly circular or semispherical at the 89.7 deg incident angle, the grains grow epitaxially from the fusion boundary in the direction of the extremely steep thermal gradient at the weld center produced by the vertically

moving heat source. The extremely steep temperature gradient associated with 89.7 deg incident angle at constant travel speed may induce negligible constitutional supercooling at the solidifying interface and produces planar weld metal grain structure [43] as observed in Fig. 9(b). The shape of the weld pool tends to become more elongated with decreasing incident angle to 85.5 deg and therefore, the columnar grains do not turn as much as in the case of a nearly circular weld pool and finally create an equiaxed grain structure at the weld center because of the periodic pulse oscillation. The direction of maximum temperature gradient becomes perpendicular to the weld interface, but because the weld pool is trailing a greater distance behind the arc, the temperature gradient at the centerline is no longer strongly directed towards the heat source which slightly decreases the steepness of the temperature gradient for constant travel speed. When the gradient is decreased slightly, any protuberance of solid metal on the interface will grow faster than the remaining planer interface because the solid is growing into supercooled liquid; that is, the solid protuberance exists at a temperature below that of the liquidus for that alloy. As a result, a cellular substructure develops in each epitaxially grown grain (Fig. 9c). The liquid ahead and alongside each cell contains greater solute content than the cell core which may yield greater degree of microsegregation during solidification than the planar structure [43]. Higher microsegregation generally facilitates precipitation of carbide and d-phase at the interdendritic spaces and weld center. Similar microsegregation should occur in weld metal of 83 deg incident angle however in this case, the weld pool takes a nearly teardrop shape due to the faster welding speed. The weld pool is extremely elongated behind the heat source and the directions of the maximum temperature gradient within the interface and centerline should have changed slightly. As a result, the grains grow from the fusion boundary and converge abruptly at the centerline

306

N. Kumar et al. / Optics and Laser Technology 94 (2017) 296–309

Fig. 13. (a) Contours plot and (b) response surface plot showing the effect of P and S on the UTS at A = 85.5 deg.

of the weld with little change in direction which yields weaker grain structure with possibilities of cracking. 4.5. Micro-hardness analysis Micro-hardness profiles across the fusion zone (horizontal direction), top to bottom (vertical direction) and the average hardness values of 304 SS weld metals are shown in Fig. 10. The hardness of weld metals is generally higher than that of the 304 SS base metal (200 Hv) (Fig. 10a and c). Among the weld metals at different incident angle, highest micro-hardness value (280 Hv) has been observed with 85.5 deg and lowest micro-hardness value (210 Hv) has been observed with 83 deg. Weld metal of 89.7 deg incident angle has in between micro-hardness values (Fig. 10a–c). The variation in hardness values among the weld metals can be explained from the relative microstructural differences. Lower hardness of 83 deg weld metal resulted from stable and higher fraction of c-phase formation along with lower d-ferrite content and relatively coarse columnar grain structure. Whereas, weld metals of 85.5 deg and 89.7 deg having relatively higher d-

Fig. 14. (a) Contours plot and (b) response surface plot showing the effect of P and A on the UTS at S = 5 mm/s.

ferrite at the dendrite core and relatively finer grain structure, resulting in higher hardness values. However, relatively higher micro-hardness of weld metal 85.5 deg (Fig. 9a) is probably due to higher alloy segregation which yields greater precipitation of d-phase and carbide in the structure. Again, Fig. 9(b) shows the micro-hardness distribution within the different fusion zones (namely FZ1, FZ2 and FZ3) of the weld metals. FZ3 hardness is slightly higher followed by FZ2 and FZ1. Comparatively fine grain structure of FZ3 (due to shift in solidification and periodic pulse oscillation) probably responsible higher hardness. 4.6. Tensile behaviour Uniaxial tensile test results of base metals and different welded joints of sample No. 5, 10 and16 are presented in Table 7. Furthermore, true stress vs. true strain curves has been considered to understand the tensile test results of different weld metals and given in Fig. 11(a). Tensile test of each welded joint have been car-

N. Kumar et al. / Optics and Laser Technology 94 (2017) 296–309

307

tions except weld metal of 83 deg incident angle where failure occurred from fusion zone. From Table 7 two major observations have been deduced; in general, weld metals have lower yield strength (rys), ultimate tensile strength (rUTS) and uniform elongation (eu) values than the base metal. Secondly, among the weld metals, weld metal of 85.5 deg incident angle has higher rys, rUTS and eu values compare to other weld metals almost comparable properties with base metal. The true stress-true strain relationship (Fig. 11a) reveals the variation in strain hardening behaviour among the weld metals which take place during uniform plastic deformation (equivalent to true plastic strain). Strain hardening behaviour can be understood from strain-hardening exponent (n) of different weld metals which is expressed by a power law equation [44] defined as r = Ken, where r is the true stress, e is the true strain, K is strength coefficient and n is the strain-hardening exponent [45]. Strain-hardening exponent (n) for each weld metals have been measured from the log-log plots of r-e and is presented in Table 7. The results show that rUTS increases with increase in nvalue; whereas rys does not vary significantly with n-value and ultimately demonstrates large strain hardening effect during uniform plastic deformation. Again, the stress increment (DrUTS-YS), expressed as the difference between the values of rUTS and rys (Table 7), reveals significant stress increment after yield during plastic deformation of the weld metals due to negligible variation in rys for the acceptable weld metals. Weld metals having higher n-value show relatively higher DrUTS-YS and eu (Table 7) which indicate large amount of energy absorption during plastic deformation up to rUTS and enhancement of the strength. The energy absorption during plastic deformation and significant strength enhancement in weld metals of 89.7 deg and 85.5 deg affect the fracture path and shifted the crack propagation towards the FZHAZ interface (Fig. 11b). Weld metal of 83 deg incident angle show drastic reduction in rUTS and according to Fig. 11(a) they have instantaneous fracture after yield through the fusion zone. Weld metal of 83 deg incident angle has natural notch (reduced fusion zone area) at the fusion zone as shown in Fig. 6(c) where local stress concentration is significantly high enough and may be the reason of sudden failure unlike other weld metals. It is likely to be mentioned here that the notch has been created due to the partial penetration. Fig. 15. (a) Contours plot and (b) response surface plot showing the effect of S and A on the UTS at P = 450 W.

ried out at constant cross-head speed of 0.5 mm/min (strain rate of 2.8  104/s) up to the fracture and the fracture has been generally localized at the FZ-HAZ interface regardless of the welding condi-

4.7. Effects of process parameters on response using response surface and contour plots Fig. 12 shows the perturbation plot, which determines the effect of all the welding parameters at the center point in the design space. It is found from this figure that the ultimate tensile strength

Fig. 16. Optimization results of ultimate tensile strength.

308

N. Kumar et al. / Optics and Laser Technology 94 (2017) 296–309

Table 8 Obtained single objective optimization results. Optimum condition for UTS, MPa

Power, W Welding speed, mm/s Incident angle, deg

Experimental result

Parametric condition

UTS, MPa

UTS, MPa

455.52 4.95 85.96

580.50

596.36 602.80 587.45

first increases then decreases with increasing value of process parameter. It is apparent from the figure that UTS increases with laser power up to a threshold value and thereafter it starts decreasing. The threshold value of laser power is related to the decomposition temperature of the base metal. This indicates that increase in the laser power, the UTS increases until the critical temperature of decomposition (i.e., the temperature of a substance at which the substance chemically decomposes) is reached. With increases of incident angle, UTS first increases then decreases. It may be due to the fact that near Brewster’s angle, absorption of laser light by the plate surface will be maximum and that angle lies between 0 and 10 deg from the normal to the surface (i.e., lies between 80 and 90 deg from surface of the base metal) where the maximum UTS will be achieved [46]. At low scanning speed heat input to the plates will be high (overlapping of laser spots will be more) and this may lead to decomposition of base metal, resulting in low joint strength but at high scanning speed overlapping of laser spots will be less and lack of heat input may occur resulting in low joint strength. The results indicate that it is not recommended to use very high or low level value of process parameters. In term of interaction between the power and speed, as shown in Fig. 13(a) and (b), the ultimate tensile strength is maximum at medium values of laser power and scanning speed. This is due to the fact that at low power, low power density is generated, resulting in lack of penetration and thus, causing a weak joint strength. Further increasing the laser power towards the center value, the ultimate tensile strength is improved as power density start increasing and maximum strength is obtained at center value. While increasing the laser power above the center limit, it results in high power density. This causes decomposition of material (overheating) and low joint strength is achieved. Similarly at lower level value of scanning speed, overheating and decomposition of material takes place. Hence low joint strength is measured and at higher value of scanning speed lack of penetration, resulting low ultimate tensile strength is obtained. Fig. 14(a) and (b) presents the combined effect of laser power and incident angle on ultimate tensile strength. It is evident from figure that joint strength is maximum at center value of laser power and incident angle. It is clear that by increasing the incident angle up to the center limit with laser power, the resulting ultimate tensile strength tends to increase. A further increase in incident angle decreases the joint strength. With very high power, overheating occurs and with very low power lack of penetration occurs. As a result low UTS will be achieved. Fig. 15(a) and (b) represent the interaction of welding speed and incident angle on joint strength. In relation to interaction of these two process parameters, the results indicate that increase in welding speed up to center limit, ultimate tensile strength increases. Further increase of welding speed decreases the joint strength, while the incident angle does not show such a significant effect. 5. Optimization of ultimate tensile strength using desirability function analysis The desirability function analysis is one of the most widely used methods in industry for the optimization of responses processes.

jError%j

2.68 3.69 1.18

Single objective optimization analysis for laser welding of 304 SS has been carried out and optimized results of ultimate tensile strength are shown in Fig. 16. In optimization goal is set to maximize ultimate tensile strength, which is desired for good quality weld. In order to get the desired responses, equal importance has been given to the lower, the target and the upper bound of the linear desirability function. For linear desirability function (d), the value of the weight is considered to be 1. In Fig. 16 each row of the graph corresponds to a response variable (Eq. (7)) and each column corresponds to one of the process parameter. Each cell of the graph shows how one of the response variables changes as a function of the of the process parameters, keeping other parameters constant. The vertical line inside the graph indicates the optimum parameter setting and horizontal dotted line represents the optimized response values. The numbers displayed at the top of the column show the upper and lower limit of process parameters setting and the optimum parameter level setting. At the left side of each row, goal of responses, predicted response (y) at the optimum parametric setting, and individual desirability value (=1) are given. MINITAB v17 software has been used for optimization of laser welding process, the optimum ultimate tensile strength (580.50 MPa) has been obtained at laser power of 455.52 W, scanning speed of 4.95 mm/s and incident angle of 85.96. The value of composite desirability factor (D) is 1. 5.1. Final validation experiments The results of optimization obtained by desirability function analysis, has been validated by conducting confirmatory tests. Three confirmatory experiments have been conducted at optimum parametric setting. The tested results of experiments at optimum conditions are presented in Table 8. It is obtained from Table 8 that there is a small error percentage between predicted and the experimental values, which validate the applied optimization technique. 6. Conclusions Based on the above experiments and analyses the following conclusions are drawn: 1. Weld pool of 85.5 deg incident angle has elongated elliptical shape with fine columnar dendritic structure. Whereas 89.7 deg and 83 deg incident angle yield nearly spherical and tear shaped weld pool respectively along with planar and coarse columnar dendritic structure. 2. Weld metal of 85.5 deg incident angle has higher fraction of carbide and d-ferrite precipitation in the austenitic matrix compare to other weld metals. 3. Weld metal of 85.5 deg incident angle has highest avg. microhardness (280 Hv) and tensile strength (579.26 MPa) followed by 89.7 deg and 83 deg incident angle weld metals. 4. The developed empirical model has been tested by analysis-ofvariance with a confidence level of 95%. 5. The error percent between the predicted and the confirmatory experiment are found to be approximately 3.6%, which validate the desirability function analysis. The results of ANOVA

N. Kumar et al. / Optics and Laser Technology 94 (2017) 296–309

show that laser power is most significant factor affecting the ultimate tensile strength followed by incident angle and welding speed. 6. It can be observed that the ultimate tensile strength increases with the welding parameters until it reaches its center value, then start to decrease with increase in welding parameters.

Acknowledgement The authors are grateful to the CoE-TEQIP (Phase-II), Jadavpur University (Grant no. P-1/RS/137/16), Kolkata, India for providing the financial assistance for this research works. References [1] S.V. Kuryntsev, A.K. Gilmutdinov, Welding of stainless steel using defocused laser beam, J. Constr. Steel Res. 114 (2015) 305–313. [2] J. Yan, M. Gao, X. Zeng, Study on microstructure and mechanical properties of 304 stainless steel joints by TIG, laser and laser-TIG hybrid welding, Opt. Laser Eng. 48 (2010) 512–517. [3] S.H. Baghjari, S.A.A. AkbariMousavi, Experimental investigation on dissimilar pulse Nd:YAG laser welding of AISI 420 stainless steel to kover alloy, Mater. Des. 57 (2014) 128–134. [4] Y. Ai, P. Jiang, X. Shao, C. Wang, P. Li, G. Mi, Y. Liu, W. Liu, A defect-responsive optimization method for the fiber laser butt welding of dissimilar materials, Mater. Des. 90 (2016) 669–681. [5] I. Tomashchuk, D. Grevey, P. Sallamand, Dissimilar laser welding of AISI 316L stainless steel to Ti6-Al4-6V alloy via pure vanadium interlayer, Mater. Sci. Eng., A 622 (2015) 37–45. [6] M.M.A. Khan, L. Romoli, M. Fiaschi, G. Dini, F. Sarri, Laser beam welding of dissimilar stainless steels in a fillet joint configuration, J. Mater. Process. Technol. 212 (2012) 856–867. [7] M. Keskitalo, K. Mantyjarvi, J. Sundqvist, J. Powell, A.F.H. Kaplan, Laser welding of duplex stainless steel with nitrogen as shielding gas, J. Mater. Process. Technol. 216 (2015) 381–384. [8] J.R. Chukkan, M. Vasudavan, S. Muthukumaran, R.R. Kumar, N. Chandrasekhar, Simulation of laser butt welding of AISI 316Lstainless steel sheet using various heat sources and experimental validation, J. Mater. Process. Technol. 219 (2015) 48–59. [9] S. Chen, J. Huang, J. Xia, X. Zhao, S. Lin, Influence of processing parameters on the characteristics of stainless steel/copper laser welding, J. Mater. Process. Technol. 222 (2015) 43–51. [10] K. Hao, G. Li, M. Gao, X. Zeng, Weld formation mechanism of fiber laser oscillating welding of austenitic stainless steel, J. Mater. Process. Technol. 225 (2015) 77–83. [11] C.H. Ng, E.S.H. Mok, H.C. Man, Effect of Ta interlayer on laser welding of NiTi to AISI 316L stainless steel, J. Mater. Process. Technol. 226 (2015) 69–77. [12] H.C. Chen, G. Bi, B.Y. Lee, C.K. Cheng, Laser welding of CPTi to stainless steel with different temporal pulse shapes, J. Mater. Process. Technol. 231 (2016) 58–65. [13] W. Tan, Y.C. Shin, Multi-scale modeling of solidification and microstructure development in laser keyhole welding process for austenitic stainless steel, Comput. Mater. Sci. 98 (2015) 446–458. [14] U. Reisgen, M. Schleser, O. Mokrov, E. Ahmed, Optimization of laser welding of DP/TRIP steel sheets using statistical approach, Opt. Laser Technol. 44 (2012) 255–262. [15] A. Matsunawa, J.D. KimD, N. Seto, M. Mizutani, S. Katayama, Dynamics of keyhole and molten pool in laser welding, J. Laser Appl. 10 (1998) 247–254. [16] W.I. Cho, S.J. Na, C. Thomy, F. Vollertsen, Numerical simulation of molten pool dynamics in high power disk laser welding, J. Mater. Process. Technol. 212 (2012) 262–275. [17] K. Abderrazak, S. Bannour, H. Mhiri, G. Lepalec, M. Autric, Numerical and experimental study of molten pool formation during continuous laser welding of AZ91 magnesium alloy, Comput. Mater. Sci. 44 (2009) 858–866. [18] J. Zhou, H.L. Tsai, P.C. Wang, Transport phenomena and keyhole dynamics during pulsed laser welding, J. Heat Transfer 128 (2005) 680–690.

309

[19] M. Akbari, S. Saedodin, D. Toghraie, R.S. Razavi, F. Kowsari, Experimental and numerical investigation of temperature distribution and melt pool geometry during pulsed laser welding of Ti6Al4V alloy, Opt. Laser Technol. 59 (2014) 52– 59. [20] B. Shanmugarajan, J.N. Chary, G. Padmanabham, B. Arivazhagan, K.A. Shaju, A. K. Bhaduri, Studies on autogenous laser welding of type 304B4 borated stainless steel, Opt. Laser Eng. 51 (2013) 1272–1277. [21] M.J. Torkamany, F.M. Ghaini, R. Poursalehi, A.F.H. Kaplan, Combination of laser keyhole and conduction welding: dissimilar laser welding of niobium and Ti6Al-4V, Opt. Laser Eng. 79 (2015) 9–15. [22] S.H. Baghjari, S.A.A.A. Mousavi, Effects of pulsed Nd:YAG laser welding parameters and subsequent post-weld heat treatment on microstructure and hardness of AISI 420 stainless steel, Mater. Des. 43 (2012) 1–9. [23] S. Chen, M. Zhang, J. Huang, C. Cui, H. Zhang, X. Zhao, Microstructures and mechanical property of laser butt welding of Titanium alloy to stainless steel, Mater. Des. 53 (2013) 504–511. [24] M.M. Atabaki, J. Ma, G. Yang, R. Kovacevic, Hybrid laser/arc welding of advanced high strength steel in different butt joint configurations, Mater. Des. 64 (2014) 573–587. [25] J. Sun, Q. Yan, W. Gao, J. Huang, Investigation of laser welding on butt joints of Al/steel dissimilar materials, Mater. Des. 83 (2015) 120–128. [26] B. Acharjee, A.S. Kuar, S. Mitra, D. Misra, Empirical modeling and multiresponse optimization of laser transmission welding of polycarbonate to ABS, Laser Manuf. Mater. Proces. 2 (2015) 103–123. [27] E.M. Anawa, A.G. Olabi, Using Taguchi method to optimize welding pool of dissimilar laser-welded components, Opt. Laser Technol. 40 (2008) 379–388. [28] J.C. Lippold, Solidification behaviour and cracking susceptibility of pulsed-laser welds in austenitic stainless steel, Weld. J. 7 (1994) 129s–139s. [29] W. Kurz, D.J. Fisher, Dendrite growth at the limit of stability: tip radius and spacing, Acta Met. 29 (1981) 11–20. [30] K. Pal, S.K. Pal, J. Mater. Engg. Perform. 20 (2011) 918. [31] T. Debroy, S.A. David, Physical processes in fusion welding, Rev. Mod. Phys. 67 (1995) 85–116. [32] S.A. David, T. Debroy, Current issues and problems in welding science, Science 257 (1992) 497–502. [33] R. Rai, S.M. Kelly, R.P. Martukanitz, T. Debroy, A convective heat-transfer model for partial and full penetration keyhole mode laser welding of a structural steel, Metall. Mater. Trans. A 39A (2008) 98–112. [34] M. Mukherjee, S. Saha, T.K. Pal, P. Kanjilal, Influence of modes of metal transfer on grain structure and direction of grain growth in low nickel austenitic stainless steel weld metals, Mater. Charact. 102 (2015) 9–18. [35] R. Rai, T.A. Palmer, J.W. Elmer, T. Debroy, Heat transfer and fluid flow during electron beam welding of 304L stainless steel alloy, Weld. J. 8 (2009) 54s–61s. [36] T. Klein, M. Vicanek, J. Kroos, I. Decker, G. Simon, Oscillations of the keyhole in penetration laser-beam welding, J. Phys. D-Appl. Phys. 27 (1994) 2023–2030. [37] N.H. Pryds, X. Huang, The effect of cooling rate on the microstructure formed during solidification of ferritic steel, Metall. Mater. Trans. A 31A (2000) 3155– 3166. [38] J.H. Lee, H.C. Kim, C.Y. Jo, S.K. Kim, J.H. Shim, S. Liu, R. Trivedi, Mater. Sci. Eng., A 413 (2005) 306. [39] M. Mukherjee, T.K. Pal, Effect of modes of metal transfer and microstructure on corrosion behaviour of welded modified ferritic stainless steel in acidic environments, J. Appl. Electrochem. 43 (2013) 347–365. [40] M. Mukherjee, T.K. Pal, Influence of mode of metal transfer on microstructure and mechanical properties of gas metal arc welded modified ferritic stainless steel, Metall. Mater. Trans. A 43 (2012) 1791–1808. [41] J.Y. Li, S. Sugiyama, J. Yanagimoto, Microstructural evolution and flow stress of semi-solid type 304 stainless steel, J. Matter. Proc. Technol. 11 (2005) 396– 406. [42] H.W. Kerr, W. Kurz, Solidification of peritectic alloys, Int. Mater. Rev. 41 (1996) 129–164. [43] S.A. David, J.M. Vitek, Principles of weld solidification and microstructures, Int. Trends Weld. Sci. Technol. ASM (1992) 147–157. [44] P. Ludwik, Elemente der Technologichen Mechanik, Springer-Verlag OHG, Berlin, 1909. [45] J.W. Simmons, Strain hardening and plastic flow properties of nitrogen-alloyed Fe-17Cr-(8–10) Mn-5Ni austenitic stainless steel, Acta Mater. 45 (1997) 2467– 2475. [46] D. Petring, High Power Diode Laser: Technology and Applications, Friedrich Bachmann, Springer, Peter Loosen & Reinhart Poprawe, 2007, p. 391.