metals
Article
The Optimization of Process Parameters and
Microstructural Characterization of Fiber Laser
Welded Dissimilar HSLA and MART Steel Joints
Celalettin Yuce *, Mumin Tutar, Fatih Karpat and Nurettin Yavuz
Department of Mechanical Engineering, Uludag University, Bursa 16059, Turkey;
mumintutar@uludag.edu.tr (M.T.); karpat@uludag.edu.tr (F.K.); nyavuz@uludag.edu.tr (N.Y.)
* Correspondence: cyuce@uludag.edu.tr; Tel.: +90-224-294-1919
Academic Editor: Giuseppe Casalino
Received: 22 July 2016; Accepted: 10 October 2016; Published: 18 October 2016
Abstract: Nowadays, environmental impact, safety and fuel efficiency are fundamental issues for
the automotive industry. These objectives are met by using a combination of different types of
steels in the auto bodies. Therefore, it is important to have an understanding of how dissimilar
materials behave when they are welded. This paper presents the process parameters’ optimization
procedure of fiber laser welded dissimilar high strength low alloy (HSLA) and martensitic steel
(MART) steel using a Taguchi approach. The influence of laser power, welding speed and focal
position on the mechanical and microstructural properties of the joints was determined. The optimum
parameters for the maximum tensile load-minimum heat input were predicted, and the individual
significance of parameters on the response was evaluated by ANOVA results. The optimum levels of
the process parameters were defined. Furthermore, microstructural examination and microhardness
measurements of the selected welds were conducted. The samples of the dissimilar joints showed a
remarkable microstructural change from nearly fully martensitic in the weld bead to the unchanged
microstructure in the base metals. The heat affected zone (HAZ) region of joints was divided into
five subzones. The fusion zone resulted in an important hardness increase, but the formation of a soft
zone in the HAZ region.
Keywords: laser welding; dissimilar weld; parameter optimization; microstructural examination
1. Introduction
The automotive sector is focused on developing and manufacturing fuel saving, higher safety
vehicles with cost efficient methods. This will be achieved through proper design and using lighter and
stronger materials on the auto body parts. Therefore the utilization of advanced high-strength steels
(AHSS) is widespread. Due to the higher strength and good formability properties, AHSS can replace
conventional thicker materials used in vehicle bodies without comprising crashworthiness. Dual phase
(DP), complex phase (CP), martensitic steel (MART) and transformation-induced plasticity (TRIP)
steels are the most common types of the AHSS [1]. Among these AHSS types, MART steel is one of the
strongest cold-rolled AHSS on the market and has become the preferred material for automotive body
applications, such as side impact beams, bumpers and structural components. Although using AHSS
steels in the automobile structure is increasing, due to specific mechanical properties, high strength
low alloy (HSLA) steel is still mainly used for structural parts, such as cross members, longitudinal
beams, chassis components, etc. [2].
Welding is one of the most used and essential joining technique in the fabrication of the auto body
and plays a significant role in assessing the final mechanical and metallurgical properties of the joined
parts [3]. Due to much superiority over conventional welding methods, such as non-contact and single
Metals 2016, 6, 245; doi:10.3390/met6100245 www.mdpi.com/journal/metals
Metals 2016, 6, 245 2 of 17
side access welding, low process cost and suitability of automation, laser welding is becoming an
attractive and economically advantageous joining technique in the automotive industry [4]. Joints
of dissimilar steel combinations in auto body structures are widely utilized for several applications
requiring a special combination of properties besides cost saving and weight reduction. However,
due to different metallurgical, thermal and physical properties of the materials, dissimilar material
welding is more challenging than similar materials welding. Due to low and concentrated heat input
and high speed properties, laser welding has also advantages on joining dissimilar materials over
other conventional methods [5]. Thus, reduced distortion and a narrower heat affected zone (HAZ)
with limited microstructural changes can be obtained.
There are several studies in the literature concerning the laser welding of similar or dissimilar
DP and HSLA steels. Saha et al. [2] examined the mechanical and microstructural properties of laser
welded DP980 and HSLA steel sheets. They stated that the tensile strength of the dissimilar welds
was lower than DP welds. Xu et al. [6] investigated microstructural and mechanical properties, and
Parkes et al. [7] reported the fatigue properties of laser welded DP and HSLA joints with varying weld
geometries. Parkes et al. [8] evaluated the tensile properties of laser welded HSLA and DP steels at
cryogenic, room and elevated temperatures. They reported that with the temperature increase, the
tensile properties were decreased. In addition, several research works investigated laser welding
of higher degree DP steels and AHSS. Wang et al. [9,10] investigated the effect of energy input and
softening mechanism on the laser butt welded DP1000 steel. They found that the weld bead width and
softening zone width become narrowed at lower energy input levels. Additionally, the mechanical
properties were increased. The study of Rossini et al. [11], concerned with laser welding of dissimilar
AHSS types, has shown that a fully martensitic microstructure was present in the 22MnB5, DP and
TRIP steels close to the fusion zone (FZ), while mainly tempered martensite and ferrite zones were
close to the base metal.
Although there are many research works about laser welding of DP and HSLA steels, only limited
work has been reported on the laser welding of MART steels. Nemecek et al. [12] compared the
microstructural and mechanical properties of MART steel joints made by laser and metal active gas
(MAG) welding. They stated that the strength of the laser welded joints was higher than arc welding,
and the HAZ width and grain coarsening in the HAZ were minimal. Zhao et al. [13] investigated the
effect of welding speed on weld bead geometry and the tensile properties of the laser welded MART
steel. They observed that, due to the fast cooling rate, the FZ of the joints contained predominantly
martensite. Furthermore, the tensile load gradually increased with decreasing welding speed.
Due to welding process parameters directly affecting the quality of the weld joints, it is necessary
to work in the suitable range. However, defining the suitable parameters to obtain the required quality
welded joints is a time-consuming process. Several optimization methods are utilized in order to solve
this problem. The Taguchi method is one of the most common design of experiment (DOE) techniques
that allows the analysis of experiments with the minimum number [14,15]. In the literature, several
researchers have used DOE methods to optimize quality characteristics in laser welding parameters.
Benyounis and Olabi [16] have presented a review of the application of optimization techniques
in several welding processes. Anawa and Olabi [17] used the Taguchi method for the purpose of
increasing the productivity and decreasing the operation cost of laser welding ferritic-austenitic steel
sheets. Another study of the authors [18] analyzed the optimized shape of dissimilar laser welded
joints and fusion zone area depending the process parameters. Sathiya et al. [19] carried out the Taguchi
method and desirability analysis to relate the parameters to the weld bead dimension and the tensile
strength of the joints with various shielding gasses. Fiber laser welding has demonstrated its capability
of welding dissimilar steel joint with and without the help of a synergic power source like the arc [20].
Acherjee et al. [21] used Taguchi, response surface methodology (RSM) and desirability function
analyses in laser transmission welding, and they investigated the optimal parameter combination for
the joint quality.
Metals 2016, 6, 245 3 of 17
In addition to these studies, several researchers used other DOE methods to investigate the
effect of laser parameters on the mechanical properties and bead geometries of laser welded joints.
Benyounis et al. [22] examined the influence of process parameters on the weld bead geometry.
They stated that weld bead dimensions were affected by the level of heat input. Ruggiero et al. [23] and
Olabi et al. [24] showed the effects of the process parameters on the weld geometry and operating cost
for austenitic steel and low carbon steel. The authors developed models and stated that, in terms of
weld bead dimensions, the most influential parameter was welding speed. Reisgen et al. [25] optimized
the parameters of the laser welded DP and TRIP steels to obtain the highest mechanical strength and
minimum operation costs. Zhao et al. [26] investigated the effects of prescribed gap and laser welding
parameters on the weld bead profile of galvanized steel sheets in a lap joint format and developed
regression models. Benyounis et al. [27] reported the multi-response optimization of laser welded
austenitic stainless steel. They developed mathematical models and established relationships between
process parameters and responses, such as cost, tensile and impact strength.
As a result of the literature review, laser power, welding speed and focal position were found to be
the most important welding parameters for welded joints’ quality and mechanical performance. Due to
the mechanical properties, especially tensile strength, being dependent on the weld bead geometry,
heat input comes to the fore [19]. Although various studies examined the influence of laser parameters
on the weld quality of dissimilar HSLA and DP steel joints, the information on fiber laser welding of
dissimilar HSLA and MART steel sheets is still not quite clear. Whereas, resolving the issue of reducing
vehicle mass while improving crash safety, the use of AHSS and HSLA is increasing. It is essential
to investigate the effect of laser welding process parameters on the mechanical performance and
quality of these steel types. Therefore, the aim of this work was to evaluate the effects of laser welding
parameters of laser power, welding speed and focal position on the response, which was a proportional
combination of tensile load (TL) and heat input (HI) using the Taguchi method. In this way, we will be
able to find the optimal welding parameters that would maximize TL, while minimizing the HI of the
fiber laser welded dissimilar HSLA and MART steel joints. In addition, for the selected samples, the
microstructural and microhardness examinations were discussed.
2. Experimental Details
In this study, all experiments were carried out on 1.5 mm-thick cold rolled MART and HSLA steel
sheets. The mechanical and chemical properties of the materials are shown in Table 1 [28]. The steel
sheets were sheared into 250 mm × 80 mm coupons, which had the sheared edges placed together for
running welds in butt joint configuration to make 250 mm × 160 mm, as shown in Figure 1a.
Table 1. Mechanical properties and chemical composition of the steels.
Material C Si Mn P S Al Nb + Yield Strength Ultimate ElongationTi (MPa) Strength (MPa) (min %)
Docol
1200M 0.14 0.4 2.0 0.02 0.01 0.015 0.1 950 1200–1400 3
HSLA * 0.1 0.5 1.8 0.025 0.025 0.015 0.15 500 570–710 14
* HSLA: high strength low alloy.
The IPG ytterbium fiber laser attached to a Kuka robotic arm was used for welding experiments.
The maximum power of the laser was 3 kW, and the wavelength was 1070 nm. The laser transmitted
through the fiber optic cables and then came to a welding head. The fiber laser had a fiber core diameter
of 0.2 mm with a laser beam spot diameter of 0.6 mm. The focal length was 300 mm. During the fiber
laser welding process, no shielding gas was used.
Metals 2016, 6, 245 4 of 17
Metals 2016, 6, 245 4 of 17 
 
FFiigguurree 11.. ((aa)) SScchheemmaattiicc iilllluussttrraattiioonn ooff tthhee fifibbeerr llaasseerr wweellddeedd sstteeeell sshheeeettss;; ((bb)) ddiimmeennssiioonnss ooff tthhee tteennssiillee 
tteesstt ssppeecciimmeennss.. MMAARRTT,, mmaarrtteennssiittiicc sstteeeell.. 
ITnhteh IiPs Gst uydttye,rfboiruomp tfiimbeizr ilnagsetrh eatptarochcesds tpoa ara Kmuektaer rso,bthoetiTc aagrumc hwiams eutsheodd fwora swuesldedin. gT hexeppearriammeenttesr. 
dTehsei gmnaixs itmheukme ypsotweperin otfh tihsem laesthero dwtaos a3c hkWiev, ianngdh tihghe wquaavleitlyenwgith owuatsin 1c0r7e0a sninmg. cTohset. lFaisresrt ltyr,aansumitiattbelde 
othrrthoougohn tahlea rfribaeyrs hopoutilcd cbaeblseesl eacntedd tdhepn ecnadmineg to nat hweetlodtianlgd hegearede. oTfhfer efeibdeorm la(sDeOr Fh)a,dw ah ifcihbecra ncobre 
cdailacmuleatteerd obf y0.s2u mm iwngithth ae liansdeirv bideaumal sDpOotF doifaemaectherp roof c0e.s6s mpmar.a Tmheet efro.cTahl eleDngOtFh fworase a3c0h0 pmamra.m Deuterirnigs 
the nfiubemr blaesreorf wpealrdaimnge tperolceevsesl,s nmo isnhuiesld1.inTgh geans, twhaese uxspeedr.i ments were run based on the orthogonal
arrayI,na ntahlyisz isntgudthye, dfoatra oapndtimidieznintigfy itnhge tphreoocpetsism puamrapmareatemrse,t etrhsea nTda,gfiuncahlil y,mifentheoceds swarays,  cuosneddu.c tTinhge 
cpoanrfiamrmeatetiro dnetsrigalns iws ithet hkeeyo psttiempa ilnl etvheisls mofetheodp atroa macehtieervs.inIng thhiigshs tquudayl,ietyx pweritihmoeuntt sinwcerereasdinesgi gcnoesdt. 
uFsirisntglya, naL s2u5itoarbtlheo goortnhaolgaornrayl ,awrrhaiych smhoeualnds b25e rsoewlesctaendd dtehpreendcoinlugm onns .thFeiv teolteavl edlsegwreree ocfo nfrseidederoemd 
f(oDrOeFa)c,h wohfitche ctahnr ebe pcarolccuelsastepda rbaym seutmerms,iwngh itchhe winedrievildasuearl pDoOwFe or,f weaeclhd ipnrgocsepsese pdaarnadmfeotecra.l Tphoes iDtiOonF. 
Tfohre elaecvhe lpsaorfamthetepra irsa mthet nerusmwberre ocfh poasreanmbeatseerd leovnelps rmeviniouuss 1w. Tohrekns,i tnhteh exlpiteerriamtuernetsa wnderceo rnusnid bearsinedg 
tohne tlhaese orrstyhsotgeomnacla pararbaiyli,t iaensa. lFyuzritnhge rthmeo drea,ttar iaanlde xipdernitmifeynintsg wtheer eoapptipmliuemd t poadreatmeremteirnse atnhde,o fpinearalltyin, igf 
rnaencgeessoafrye,a chonpdroucetisnsgp caoranmfiremteartinono rtdriearlst owpirtohd tuhcee oapntiamccaelp lteavbelles qouf atlhitey pwaerladminegte.rTsh. eInl etvheisls sotuf dthye, 
pexropceersismpeanrtasm weetreers daerseigsnheodw unsiingT aabnl eL2.5A ornthegoagtoivneald aerfroacyu,s wishoicbht amineeadnsw 2h5e nrotwhes faoncda lthproeien tcoplousmitinosn. 
iFsivbe lolewvetlhse wspeercei mcoenssiduerfraecde .for each of the three process parameters, which were laser power, 
welding speed and focal position. The levels of the parameters were chosen based on previous works 
in the literature and consiTdaebrlien2g. Lthaese lrawseerld siynsgtpemroc ceassppaabrialimtieetse.r sFuanrtdhleervmelso.re, trial experiments were 
applied to determine the operating range of each process parameter in order to produce an acceptable 
quality welVdainrigab. lTehse levelsU onfi tthe pSryomcebsosl paLreavmele1ters aLreev eslh2ownL einv eTl a3ble L2e. vAe ln4egatLievvee dl 5efocus is 
obtained wLahseenr Pthoew feorcal poinWt positionP is below1 0t0h0e speci1m25e0n surfa1c5e0. 0 1750 2000
Welding Speed mm/s S 5 15 25 35 45
Focal Position Tamblme 2. LaserF welding p0rocess par−am0.2eters an−d 0l.e4vels. −0.6 −0.8
Variables Unit Symbol Level 1 Level 2 Level 3 Level 4 Level 5 
In the data analysis, in order to evaluate the effect of the selected parameters on the response, the
Laser Power W P 1000 1250 1500 1750 2000 
signal-to-noWiseel(dSin/gN S)preaetdio smarme/cs alculaSt ed. In ad5 dition, S1/5 N ratio2s5 are use3d5 to redu4c5e the response
variability. InFtohciasl wPoosritkio, tnh e lamrgmer -the-bFe tter S/N0r atio wa−s0.c2h osen−i0n.4o rder t−o0.m6 axim−iz0e.8t he responses.
The S/N ratio for the larger-the-better for the respon(ses was calculated as follows:In the data analysis, in order to evaluate the effect of the) selected parameters on the response, 
n
the signal-to-noise (S/N) ratios are caSlc/uNlat=ed−. I1n0 laod 1gdition∑, S/
1N ratios are used to reduce the respons
2 (1
e) 
variability. In this work, the larger-the-better S/N rnait=io1 ywi as chosen in order to maximize the 
responses. The S/N ratio for the larger-the-better for the responses was calculated as follows: 
where yi is the response data from the experiment for the i-th parameter and n is the number
of experiments. A higher S/N ratio indicates super1ionr c1onsideration for the optimal parameter
S N  10log  
combination, since the major signal dominates the n o
i 2  (1) i1sey. Equation (2) is used to calculate thei
parameter effects:
where yi is the response data from the experime1nt nfor the i-th parameter and n is the number of 
S/N
experiments. A higher S/N ratio indicateis,j
= S/N (2)
 supne∑rior conk sideration for the optimal parameter k=1
combination, since the major signal dominates the noise. Equation (2) is used to calculate the 
where S/Ni,j is the average S/N value of the j-th level of the i-th parameter and n is the number of theparameter effects: 
experiment, which includes the j-th level of the i-th parameter. Additionally, the S/Nk is the value of
 
Metals 2016, 6, 245 5 of 17
the k-th experiment S/N. Finally, a statistical analysis of variance (ANOVA) was used to indicate the
relative effect of each process parameter on the responses.
At the metallographic examination stage of the study, the samples were cut from the weld
cross-section using an electrical discharge cutting machine, then mounted in Bakelite, ground and
polished up to 0.25-µm diamond paste. Two different etching procedures were conducted to reveal the
grain boundaries and weld zone microstructure. In the first stage of the etching, 3% Nital solution
was used. Then, to reveal some microstructures, subsequently, tint etched using 10% Na2S2O5 was
performed. Then, samples were analyzed for microstructural changes and possible defects using an
optic microscope (OM, Nikon DIC, Tokyo, Japan) with the Clemex image analysis system and the
scanning electron microscope (SEM, Zeiss EVO 40 XVP, Oberkochen, Germany). Vickers microhardness
measurements (DUROLINE-M microhardness tester, Metkon, Turkey) were performed with a 200-g
load, and 10-s dwell time. Tensile samples were machined from the perpendicular to the welding
direction in accordance with ASTM, E8/E8M (Figure 1b). Tensile tests were performed using a
computerized tensile testing machine (UTEST-7014, Ankara, Turkey) using a constant crosshead speed
of 5 mm/min.
3. Results and Discussion
3.1. Optimization of the Process Parameters via the Taguchi Method
In this study, a Taguchi orthogonal array, which can handle five levels of the parameters with
three columns and 25 rows, was used. The parameter optimization procedure was done in order
to get a welded joint that has the maximum TL by minimizing the HI. HI plays a crucial role in the
quality of the joint and indirectly the operation cost. The weld joint quality can be defined as weld
bead geometry, mechanical properties and distortions [25]. Weld bead geometry, which means the
bead width and penetration depth, is an important physical characteristic of a weldment, especially
for dissimilar laser welding processes [19]. The appropriate weld bead geometry depends on the HI
rate [22]. A shallower and inadequate penetration depth is related to an insufficient HI rate. Thence,
the TL of the welded joint will decrease. However, a higher HI gives a slower cooling rate, and so,
in the HAZ, large grain sizes can have poor toughness and decrease in TL. Hence, HI and, consequently,
weld bead geometry affect the tensile strength of the joints [16,18]. Therefore, in this study, TL and HI
were evaluated together as a response variable. Due to the tensile strength being the most important
quality indicator of the welded joint, the effect ratio of the TL was determined to be higher, 60%.
In determining the effect ratio of the HI, operational cost and weld bead geometry were considered.
Namely, this ratio should not be too low because of the insufficient penetration, and also, it should not
be too high in terms of cost and decreased strength of the joint. Therefore, it was determined to be 40%.
In determining these effect ratios, they have also benefited from operational experience.
The TL of the laser welded joints was experimentally determined using tensile tests. At least three
different specimens’ tensile test results’ average were taken. Additionally, HI was calculated by the
laser power divided by the welding speed. Due to the scale of the values of TL and HI being different,
a normalization process was applied to these values. Equation (3) was used for the normalization of
the TL values.
X
X in = (3)Xmax
where Xn is the normalized value, Xi is the value of the relevant row and Xmax is the maximum value.
Since the objective function was a combination of the TL and HI, it is necessary to express it in the
same form. Therefore, before applying Equation (3), the reciprocals of the HI values were taken using
Equation (4) to convert the values to the larger the better form.
1
Xp = (4)Xi
Metals 2016, 6, 245 6 of 17
where Xp is a pre-normalized value, which was used in Equation (1), and Xi is the HI value of the
relevant row.
The experimental layout for the process parameters, average TL, standard deviations (SD),
HI values and normalized values are shown in Table 3. The S/N ratios for the response were calculated.
The response column represents the sum of 60% normalized TL and 40% normalized HI. The S/N
ratios of the process parameters were calculated by using Equation (2), and the effect of each parameter
level was determined. As can be seen in Table 4, welding speed was the most important parameter for
the response. Laser power and focal position followed this parameter, respectively.
Table 3. Design matrix with experimental results. TL, tensile load; HI, heat input.
Parameters Outputs and Calculations
Exp. S/N
No. Power Speed Focal TL NormalizedHI Normalized
Response
Ratio
(W) (mm/s) (mm) (kN) SD TL (J/mm) HI
1 1000 5 0 5.92 0.04 0.995 200.000 0.111 0.642 −3.849
2 1000 15 −0.2 5.49 0.05 0.923 66.667 0.333 0.687 −3.260
3 1000 25 −0.4 4.61 0.22 0.775 40.000 0.556 0.687 −3.260
4 1000 35 −0.6 3.43 0.11 0.578 28.571 0.778 0.658 −3.635
5 1000 45 −0.8 3.18 0.20 0.534 22.222 1.000 0.720 −2.853
6 1250 5 −0.2 5.88 0.02 0.990 250.000 0.089 0.629 −4.026
7 1250 15 −0.4 5.82 0.06 0.978 83.333 0.267 0.694 −3.172
8 1250 25 −0.6 5.32 0.08 0.894 50.000 0.444 0.714 −2.926
9 1250 35 −0.8 4.44 0.08 0.746 35.714 0.622 0.697 −3.135
10 1250 45 0 3.71 0.14 0.625 27.778 0.800 0.695 −3.160
11 1500 5 −0.4 5.73 0.04 0.964 300.000 0.074 0.608 −4.321
12 1500 15 −0.6 5.93 0.05 0.997 100.000 0.222 0.687 −3.260
13 1500 25 −0.8 5.93 0.06 0.997 60.000 0.370 0.746 −2.545
14 1500 35 0 5.82 0.00 0.979 42.857 0.519 0.795 −1.992
15 1500 45 −0.2 4.30 0.10 0.723 33.333 0.667 0.701 −3.085
16 1750 5 −0.6 5.52 0.06 0.929 350.000 0.063 0.583 −4.686
17 1750 15 −0.8 5.90 0.02 0.992 116.667 0.190 0.671 −3.465
18 1750 25 0 5.79 0.10 0.974 70.000 0.317 0.712 −2.950
19 1750 35 −0.2 5.87 0.02 0.987 50.000 0.444 0.770 −2.270
20 1750 45 −0.4 5.95 0.01 1.000 38.889 0.571 0.829 −1.628
21 2000 5 −0.8 5.52 0.07 0.929 400.000 0.056 0.579 −4.746
22 2000 15 0 5.87 0.10 0.987 133.333 0.167 0.659 −3.622
23 2000 25 −0.2 5.58 0.08 0.938 80.000 0.278 0.674 −3.426
24 2000 35 −0.4 5.62 0.04 0.946 57.143 0.389 0.723 −2.817
25 2000 45 −0.6 5.67 0.05 0.953 44.444 0.500 0.772 −2.247
Table 4. Response table for the S/N ratios for the objective.
Level Laser Power Welding Speed Focal Position
1 −3.372 −4.326 −3.349
2 −3.284 −3.356 −3.351
3 −3.041 −3.022 −3.040
4 −3.000 −2.770 −3.214
5 −3.372 −2.595 −3.115
Delta 0.372 1.731 0.311
Rank 2 1 3
The S/N ratios’ main effect plot showed how each process parameter affects the response
characteristic. The means of the S/N ratios exhibit a good correlation with the main effects of the mean
of means (Figure 2). This result indicates that process parameters show higher mean values resulting
in higher variability. The response seems to be mainly affected by the process parameters, as shown in
Figure 2. It can be seen that the welding speed was the most important process parameter that affected
Metals 2016, 6, 245 7 of 17 
in Figure 2. It can be seen that the welding speed was the most important process parameter that 
Metals 2016, 6, 245 7 of 17
affected the response. There was a small difference between laser power and focal position; while the 
focal position plots showed the lowest effect on the response to those parameters. 
the reInsp tohnis es.tuTdhye,r ethwe aospatimsmaal lpl adriaffmereetnerc ecobmetbwineaetniolans weraps ofwouenrda ntod bfoec 1a7l5p0o Wsit ifoonr; lwasheirl eptohweefro, c4a5l 
pmomsit/iso fnorp wloetsldsihnogw spedeetdh eanlodw −e0s.4t  emffmec tf orn ththee forecsapl opnoseititont.h Tohseisp paarraameetteerrs .combination was Sample 
20 in the orthogonal array in Table 3; thus, no additional confirmation experiments were required. 
 
(a) 
 
(b) 
FFiigguurree 22.. EEffffeeccttss pplloottss ooff ((aa)) SS//NN rraattiiooss;; aanndd ((bb)) mmeeaann ooff mmeeaannss ffoorr tthhee rreessppoonnssee.. 
3.2. Analysis of Variance 
In this study, the optimal parameter combination was found to be 1750 W for laser power, 45 mm/s
for wTehldei nogrdsepre oedf iamnpdo−rt0a.n4cme omf ftohre tphaerfaomcaeltepross oitnio tnh.eT rheissppoanrsaem weatesr dceotmerbmininaetido nuswinags SAaNmOpVleA2.0 Biny 
tchoemoprathrionggo nthael  aersrtaiymiantiToanb loef 3t;hteh uesx,pneoriamdednittiaol nearlrcoorns fiargmaiantsiot ntheex pmereiamne snqtsuawrer, eAreNqOuVireAd .tests the 
importance of all main factors and their interactions. In this study, The ANOVA shows that for the 
3re.2s.pAonnasley (smisaoxf iVmauriman cTeL and minimum HI), welding speed has the greatest effect with a contribution 
of 64T.0h1e poerrdceenrt.o Lf aimsepr oprotwanecre anofd tfhoecapl aproasmiteiotenr esffoencttsh we errees p5.o6n0s%e awnads 2d.8e2te%r,m reinspedecutisvienlgy (ATNabOleV 5A).. 
BTyhisc ormespualtr iinsg cothmepeasttiibmlea twiointh oTfatbhlee e4x, pwehriimche nist atlheer rroesrspoangsaei ntsatblteh efomr ethane Ssq/Nu arraet,ioAsN. DOuVeA tot etshtse 
tihneterimacptioorntsa nbceetwofeeanll tmhea ipnrofaccetsosress apnadratmheeitrerisn tneorat cbtieoinngs. dIenfitnheids , stthued yr,esTihdeuaAl NerOroVrA wsahso lwarsgteh aint 
fAoNr tOhVeAre. sponse (maximum TL and minimum HI), welding speed has the greatest effect with a
contribution of 64.01 percent. Laser power and focal position effects were 5.60% and 2.82%, respectively
(Table 5). This result is compatTibablelew 5i.t AhnTaalbyslees4 o, fw vhariciahnicse tthabelree fsopr omnesaentsa. ble for the S/N ratios. Due to
the interactions betDwegereenes tohf e procSeusmse osf pSqaurarmese tersAndojutstbeedi Mngeadn eSfiqnuaerdes, the residual errCoornwtriabsutliaorng e
Source F p 
in ANOVA. Freedom (DF) (Seq SS) (Adj MS) (%) 
Laser Power 4 0.004975 0.001244 0.61 0.663 5.60 
Welding Speed 4 0.056813 0.014203 6.97 0.004 64.01 
Focal Position 4 0.002506 0.000627 0.31 0.867 2.82 
Residual Error 12 0.024450 0.002038   27.55 
Total 24 0.088745     
 
Metals 2016, 6, 245 8 of 17
Table 5. Analyses of variance table for means.
Source Degrees of Sum of Squares Adjusted Mean ContributionFreedom (DF) (Seq SS) Squares (Adj MS) F p (%)
Laser Power 4 0.004975 0.001244 0.61 0.663 5.60
Welding Speed 4 0.056813 0.014203 6.97 0.004 64.01
Focal Position 4 0.002506 0.000627 0.31 0.867 2.82
MetaRlse 2s0id16u,a 6l, E2r4r5o r 12 0.024450 0.002038 27.558 of 17 
Total 24 0.088745
3.3. Effects of Process Parameters on the Response 
3.3. Effects of Process Parameters on the Response
In this study, it was observed that welding speed was a significant parameter that affects the 
respoInnsteh, iws shtiuchd yi,s itmwaxaismoubmse rTvLe danthda mt wineilmdiunmg sHpIe. eAdltwhaosugahs itghnei fiecffaenctt poaf rlaamseert eprowtheart maffaeyc tssetehme 
qreusipteo nssme,awll hinic hAiNs OmVaxAim reusmulTtsL, iatn ids mani niimpuomrtaHnIt. Aprlothcoeussg hpathraemefefteecrt dofulea steor tphoew aesrsomciaaytesdee HmI.q Tuhitee 
isnmcarellaisne AoNf lOasVeAr preoswuletrs ,ciatuisseasn mimopreo rhtaenatt pinropcuets.s Upnadraemr ethteer hdiugeht olatsheer apsoswoceira,t eifd tHheI. wTheeldiinncgr esapseeeodf 
wlaseerre pnoowt ecrhocasuense ps rmoporeerlhye, atthien wpuetl.dU bnedaedr wthoeuhlidg hbela bseror apdoewneerd,  iaf nthde twhee lsduinrfgacspe eqeudawliteyr eonf otht ech woseeldn 
pdreocpreearslye,dt.h Tehweereldfobreea, dthwe olausledr bpeobwreora danende wd ealnddintgh espsueerfda csehoquuladli tbye ocfotnhseidweereldd dtoegcreetahseerd t.oT gheetr egfooorde, 
wtheeldla sperropfiolews earnadn TdLw. eWldhienng tshpee eladsesrh opuolwdebre wcoasn skiedpetr ecdontostgaentth, ewr ittohg ientcgreoaosdinwg ewldelpdrionfigl esspeaendd, THLI. 
Wdehcreenatsheed.l aDseurep toow werewlda sbekaedp tgceoonmsteatnrty, wreiltahteindc rteoa sthineg HwIe, ldwienlgd sbpeeaedd ,wHiIddthe cwreaass eidn.cDreuaesetdo weitldh 
ibnecardeagseionmg eHtrI.y Irne alaltl eldasteor tphoewHeIr, lweveledlsb, ewahdewn itdhteh swpeaesdin wcraesa 5se mdmw/isth, tihnec rbeeaasdinsg wHerI.eI lnaragllelra dseurep toow theer 
elexvceelsss,iwveh ehneatth einsppuete d(Fwigausr5e m3am). /Os,nt htheeb eoathdesrw seidree,l awrgheernd tuhee tsopteheede wxcaess s4iv5 emhmea/ts,i ntphue tb(eFaigdus rwe 3eare). 
fOonunthde too tbhee rnsairdreo,wwehr e(nFitghueresp 3ebe)d. was 45 mm/s, the beads were found to be narrower (Figure 3b).
 
(a) 
 
(b) 
FFiigguurree 33.. TTrarannsvsveresres esescetciotinosn osf otfheth joeinjotisn utssiungsi ndgiffderifefnetr ehnetath ienaptuintsp: u(at)s :30(a0) J/3m00mJ, /Smammp, leS a1m1;p (lbe) 1414; 
(J/bm) m44, JS/ammmpl,eS 2a5m. ple 25.
As known weld bead dimensions directly affect the TL of the joints [25], at insufficient HI at the 
As known weld bead dimensions directly affect the TL of the joints [25], at insufficient HI at
low laser power levels or high welding speeds, adequate penetration did not occur, and the TL of the 
the low laser power levels or high welding speeds, adequate penetration did not occur, and the
joints was decreased. Besides, at excessive HI levels, the HAZ would be wider, and that causes a 
TL of the joints was decreased. Besides, at excessive HI levels, the HAZ would be wider, and that
decrease in TL. According to the tensile test results, the welding speed in the range between 35 mm/s 
causes a decrease in TL. According to the tensile test results, the welding speed in the range between
and 45 mm/s would lead to minimum HI and acceptable TL for the joints. The focal position 
35 mm/s and 45 mm/s would lead to minimum HI and acceptable TL for the joints. The focal position
determines the laser spot size and consequently the power density on the surface, depending on the 
optical path. In this study, the focal position has the lowest effect on the response. It is believed that 
the level range of this parameter caused this situation due to the range of the spot diameters being 
quite small.  
3.4. Microstructure and Microhardness Evolution 
The microstructural examination and microhardness evolution of the selected welds that have 
the highest (Sample 20) and lowest (Sample 21) response values were discussed. Three different 
zones, including FZ, HAZ and base metal (BM), were revealed by examining the selected sample’s 
cross-sections. The BM of the HSLA consisted of a ferrite matrix with carbides dispersed in the grains 
 
Metals 2016, 6, 245 9 of 17
determines the laser spot size and consequently the power density on the surface, depending on the
optical path. In this study, the focal position has the lowest effect on the response. It is believed that
the level range of this parameter caused this situation due to the range of the spot diameters being
quite small.
3.4. Microstructure and Microhardness Evolution
The microstructural examination and microhardness evolution of the selected welds that have
the highest (Sample 20) and lowest (Sample 21) response values were discussed. Three different
Meztaolsn 2e0s1,6i, n6,c 2lu45d ing FZ, HAZ and base metal (BM), were revealed by examining the selected sampl9e ’osf 17 
cross-sections. The BM of the HSLA consisted of a ferrite matrix with carbides dispersed in the grains
anadn adt atthteh egrgariani nbobuounnddaraireise s(F(Figiguurree 44aa)).. AAss sshown iin Fiigurree 44bb,, MMAARRTTs tseteleslsw werereco cmopmrpisreidseodf of 
mamrtaerntesnitsiict imc micricorsotsrtuructcutureress aanndd aa ssmallll prroporrttiion  off ffeerrrritiitcica annddb baianiintiicticg rgarianisn. s. 
  
(a) 
  
(b) 
FigFuigrue r4e. 4O. pOtpictaicla ml micircorgograrpaphh aanndd SSEEMM vviieewwss ooff tthhee:: ((aa)) HHSSLLAAb baasseem meetatlal( B(BMM);)(;b ()bM) MARATRBTM B.M. 
In Itnheth we ewldelidnign gprporcoecsess,s f,ifinnaal lmmicicrroossttrruuccttuurreess aarree aaffffeecctteeddb byyp peeaakkt etemmppereartautruerea nadndth tehceo coolionlginrga treate 
of othf eth reerleelveavnatn tzozonneess, ,aanndd ccaarrbboonn eeqquuiivvaalelenntt( C(CEE) v) avlauleuree sruesltueldtefdro fmrotmhe tchhee mchisetmryisotfryth oefs tteheel ss[t2e9e–ls3 2[2].9–
32]A. lAthlothuoguhgthh etrheearree anruem neurmouesrofoursm fuorlameufolarec afolcru claatlicnuglaCtEin,gY uCrEio, kYau’srfioorkma’us lafowrmasuulas ewdains uthsiesds tiund tyhis 
stubdeyca buesecaoufsiets osfu iittasb siulitiytafboirliCty- Mfonr sCte-eMlsn[ 3s3t]e.eTlsh e[3C3E]. vTahluee Cs Eof vstaeleulessw oefr estceaelclsu lwateerde ucsailncgulYauterido kuas’sing 
Yufroiromkau’lsa fgoirvmenublay gEiqvuenat bioyn E(5q)uaantidosnh (o5w) annidn sThabolwe n6 [i3n4 T,3a5b].leT 6h e[3T4i,3e5le]m. Tehnet wTia eslceomnesnidte wreads acsotnhseidNebred 
as etlheem Nenbt ebleecmauenset obfetchaeuisres iomf itlhaer ierf fseimct iolanrt hefefesctet eolsn’ thhaer dsetenealbsi’l ihtya.rdenability. 
 Si Mn Cu Ni (Cr MoNbV)
CE  C  f (C)        (5) 
24 6 15 20 5 
where f(C) is the accommodation factor and is calculated as; 
f (C)  0.750.25tanh 20C 0.12  (6) 
Table 6. The carbon equivalent (CE) values of the HSLA and MART steels. FZ, fusion zone. 
Calculated Zone HSLA MART FZ 
CE 0.330 0.453 0.391 
The microstructure of the FZ of Sample 20, with a 0.391 CE value (average of MART and HSLA 
steels), is predominantly martensite with a bainitic structure (Figure 5). With the effect of the heat 
exchange gradient, in the vicinity of the fusion boundary, grains were elongated towards the weld 
center. However, in the center of the FZ, equiaxed grains were observed (Figure 5a). Furthermore, due 
to the lack of shielding gas, as a possible result of the diffusion of some elements, i.e., oxygen and 
nitrogen from the air, it is thought to be some inclusions in the FZ, which were marked with yellow 
arrows in Figure 5b. 
 
Metals 2016, 6, 245 10 of 17
[ ]
Si Mn Cu Ni (Cr + Mo + Nb + V)
CE = C + f (C) + + + + (5)
24 6 15 20 5
where f (C) is the accommodation factor and is calculated as;
f (C) = 0.75 + 0.25tanh [20 (C − 0.12)] (6)
Table 6. The carbon equivalent (CE) values of the HSLA and MART steels. FZ, fusion zone.
Calculated Zone HSLA MART FZ
CE 0.330 0.453 0.391
The microstructure of the FZ of Sample 20, with a 0.391 CE value (average of MART and HSLA
steels), is predominantly martensite with a bainitic structure (Figure 5). With the effect of the heat
exchange gradient, in the vicinity of the fusion boundary, grains were elongated towards the weld
center. However, in the center of the FZ, equiaxed grains were observed (Figure 5a). Furthermore,
due to the lack of shielding gas, as a possible result of the diffusion of some elements, i.e., oxygen and
nitrogen from the air, it is thought to be some inclusions in the FZ, which were marked with yellow
aMrertoalws 2s01in6, F6,i 2g4u5r e 5b. 10 of 17 
  
(a) (b) 
FigFuirgeu 5re. (5a.) (Oa)pOticpatli cmalicmroicgrroagprha;p ahn; da n(bd)( bSE) MSE Mmicmroicgrroagprhap shhoswhoinwgin thget hFeZF oZf othfet hSeamSapmlep 2le0.2 0.
Weld zone microstructures of Sample 21, which have the highest heat input and, of course, 
Weld zone microstructures of Sample 21, which have the highest heat input and, of course, slowest
slowest cooling rate, are completely different from Sample 20 and not associated with the CE values 
cooling rate, are completely different from Sample 20 and not associated with the CE values due to
due to the slow cooling conditions. The FZ of Sample 21 consisted of ferritic microstructures with 
the slow cooling conditions. The FZ of Sample 21 consisted of ferritic microstructures with multiple
multiple morphologies, e.g., grain boundary, acicular and Widmanstatten (Figure 6a). Due to the 
morphologies, e.g., grain boundary, acicular and Widmanstatten (Figure 6a). Due to the oriented
oriented solidification and slow cooling rate, elongated and extremely coarse grains were revealed. In 
solidification and slow cooling rate, elongated and extremely coarse grains were revealed. In Figure 6b,
Figure 6b, grain boundaries were dashed with yellow, which contain different ferritic structures. 
grain boundaries were dashed with yellow, which contain different ferritic structures. Acicular ferritic
Acicular ferritic microstructures can also be seen in Figure 6. The yellow arrows show the inclusions 
microstructures can also be seen in Figure 6. The yellow arrows show the inclusions where acicular
where acicular ferrites nucleated (Figure 6c). 
ferrites nucleated (Figure 6c).
 
(a) 
  
(b) (c) 
Figure 6. Detailed different magnifications of FZ microstructures of Sample 21: (a) FZ at ×100 
magnification; (b) extremely coarse grains in FZ; and (c) inclusions in FZ. 
 
Metals 2016, 6, 245 10 of 17 
  
(a) (b) 
Figure 5. (a) Optical micrograph; and (b) SEM micrograph showing the FZ of the Sample 20. 
Weld zone microstructures of Sample 21, which have the highest heat input and, of course, 
slowest cooling rate, are completely different from Sample 20 and not associated with the CE values 
due to the slow cooling conditions. The FZ of Sample 21 consisted of ferritic microstructures with 
multiple morphologies, e.g., grain boundary, acicular and Widmanstatten (Figure 6a). Due to the 
oriented solidification and slow cooling rate, elongated and extremely coarse grains were revealed. In 
Figure 6b, grain boundaries were dashed with yellow, which contain different ferritic structures. 
MAectiaclsu2l0a1r6 f, e6,r2ri4t5ic microstructures can also be seen in Figure 6. The yellow arrows show the inclu11sioofn1s7 
where acicular ferrites nucleated (Figure 6c). 
 
(a) 
  
(b) (c) 
FFiigguurree 66.. DDeettaaiilleedd ddiiffffeerreenntt mmaaggnniiffiiccaattiioonnss ooff FFZZ mmiiccrroossttrruuccttuurreess ooff SSaammppllee 2211:: ((aa)) FFZZ aatt ××110000 
mmaaggnniifificcaattiioonn;; ((bb)) eexxttrreemmeellyy ccooaarrssee ggrraaiinnss iinn FFZZ;; aanndd ((cc)) iinncclluussiioonnss iinn FFZZ.. 
 The HAZ of Sample 20 can be divided into five subzones, namely partially molten zone (PMZ),
coarse-grained HAZ (CGHAZ), fine-grained (FGHAZ), inter-critical HAZ (ICHAZ) and sub-critical
HAZ (SCHAZ). Optical micrographs of these different subzones can be seen in Figures 7 and 8. In the
microstructural examinations, PMZ could not be observed. Both MART and HSLA steel, in CGHAZ,
consisted of martensitic-bainitic microstructure as a result of the transformation of coarsened austenite
grains (Figures 7a and 8a). While the CGHAZ of MART steel shows a higher proportion of martensitic
and lower proportion of bainitic microstructures, HSLA steel shows a higher proportion of bainitic
and lower proportion of martensitic microstructures. This can be attributed to the CE values of the
steels. A higher CE value promoted the formation of martensite, whereas a lower CE value promoted
bainitic structures. Although the FGHAZ of MART steel’s microstructure is similar to CGHAZ, but
consisted of finer grains, this zone could not be observed in HSLA steel. In the ICHAZ, where the
peak temperature is between A3 and A1, the partial transformation of ferrite to a mixture of ferrite and
austenite resulted in martensite islands between the fine-grained ferrite matrix and carbides in HSLA
steel (Figure 7b) [2]. Figure 7b shows a transition zone towards SCHAZ.
Metals 2016, 6, 245 11 of 17 
The HAZ of Sample 20 can be divided into five subzones, namely partially molten zone (PMZ), 
coarse-grained HAZ (CGHAZ), fine-grained (FGHAZ), inter-critical HAZ (ICHAZ) and sub-critical 
HAZ (SCHAZ). Optical micrographs of these different subzones can be seen in Figures 7 and 8. In the 
microstructural examinations, PMZ could not be observed. Both MART and HSLA steel, in CGHAZ, 
consisted of martensitic-bainitic microstructure as a result of the transformation of coarsened 
austenite grains (Figures 7a and 8a). While the CGHAZ of MART steel shows a higher proportion of 
martensitic and lower proportion of bainitic microstructures, HSLA steel shows a higher proportion 
of bainitic and lower proportion of martensitic microstructures. This can be attributed to the CE 
values of the steels. A higher CE value promoted the formation of martensite, whereas a lower CE 
value promoted bainitic structures. Although the FGHAZ of MART steel’s microstructure is similar 
to CGHAZ, but consisted of finer grains, this zone could not be observed in HSLA steel. In the 
ICHAZ, where the peak temperature is between A3 and A1, the partial transformation of ferrite to a 
Metals 2016,m6,ix2t4u5re of ferrite and austenite resulted in martensite islands between the fine-grained ferrite matrix 12 of 17
and carbides in HSLA steel (Figure 7b) [2]. Figure 7b shows a transition zone towards SCHAZ. 
 
   
(a) (b) (c) 
Figure 7. Detailed heat affected zone (HAZ) microstructures and subzones of the HSLA side of 
Figure 7. Detailed heat affected zone (HAZ) microstructures and subzones of the HSLA side of
Sample 20: (a) coarse-grained HAZ (CGHAZ); (b) inter-critical HAZ (ICHAZ); and (c) sub-critical  
Sample 20H:A(Za )(ScCoHaArsZe).- grained HAZ (CGHAZ); (b) inter-critical HAZ (ICHAZ); and (c) sub-critical
HAZ (SCHAZ).
Metals 2016, 6, 245 12 of 17 
The ICHAZ of MART steel exhibited a dual phase microstructure containing ferrite with fine 
and well-dispersed martensite. In addition, some portion of the acicular ferritic microstructures can 
be seen in Figure 8c. Since shielding gas was not used, nitrogen and oxygen absorption could promote 
titanium base nitrides, carbo-nitrides and oxide inclusions where acicular ferrites can nucleate [2,36–
38]. Furthermore, the slow cooling rate of this zone could induce ferritic structures to be formed. 
Figure 8d shows the SCHAZ of MART steel. In this zone, tempered martensite and bainite formed 
due to the lower peak temperature than A1. However, it is expected that the coarsening of the carbides 
occurs in the HSLA side, and there is no difference identified metallographically. This can be related 
to the thermal stability of the HSLA, which is greater than MART and, therefore, does not have a 
microstructure that is distinct from its BM [2]. 
 
 
  
(a) (b) 
  
(c) (d) 
Figure 8. Detailed HAZ microstructures and subzones of the MART side of Sample 20: (a) CGHAZ; 
Figure 8. Detailed HAZ microstructures and subzones of the MART side of Sample 20: (a) CGHAZ;
(b) fine-grained (FGHAZ); (c) ICHAZ; and (d) SCHAZ. 
(b) fine-grained (FGHAZ); (c) ICHAZ; and (d) SCHAZ.
For Sample 21, the whole weld zone was roughly 11 mm, so only the micrographs of specific 
zones are presented here. The CGHAZ of HSLA side of Sample 21 consisted of ferritic and bainitic 
structures and it is shown with dashed lines. The FGHAZ of the HSLA side contains similar, but finer 
grains with respect to CGHAZ (Figure 9b). Beside the FGHAZ, coarsening of the carbides occurred 
in the HSLA side. 
  
(a) (b) 
Figure 9. Microstructures of the HAZ zone for HSLA side of Sample 21: (a) CGHAZ and (b) 
FGHAZ. 
 
Metals 2016, 6, 245 12 of 17 
 
  
(a) (b) 
Metals 2016, 6, 245 13 of 17
The ICHAZ of MART steel exhibited a dual phase microstructure containing ferrite with fine and
well-dispersed martensite. In addition, some portion of the acicular ferritic microstructures can be
seen in Figure 8c. Since shielding gas was not used, nitrogen and oxygen absorption could promote
titanium base nitrides, carbo-nitrides and oxide inclusions where acicular ferrites can nucleate [2,36–38].
Furthermore, the slow cooling rate of this zone could induce ferritic structures to be formed. Figure 8d
shows the SCHAZ of MART steel. In this zone , tempered martensite and bainite for med due to
the lower peak temperature (tch)a n A (d) 1. However, it is expected that the coarsening of the carbides
occurFsiginurteh 8e. HDeStLaiAledsi HdeA,Za nmdictrhoesrtreuicstunroesd ainffde rseunbczeonidese notfi tfiheed MmAeRtaTl lsoidger aopf hSaicmaplllye. 2T0h: i(sa)c CanGHbeArZe;l ated
to th(ebt)h fienrem-garal isnteadb (iFliGtyHoAfZt)h; (ec)H ICSHLAZ, ;w ahndic (hd)i sSCgHreAatZe.r than MART and, therefore, does not have a
microstructure that is distinct from its BM [2].
For Sample 21, the whole weld zone was roughly 11 mm, so only the micrographs of speciffiic 
zones are presented here. The CGHAZ of HSLA side of Sample 21 consisted of ferritic and bainitic 
structures and it is shown with dashed lines. The FGHAZ of tthe HSLA ssiide cconttaiins ssiimiillar, but ffiiner 
grains with  resspeecctt ttoo CCGGHHAAZZ( F(Figiguurere9 9bb).).B Beseisdideet htheeF GFGHHAAZ,Zc, ocaorasresneinigngo fotfh tehcea crbaridbeidseosc couccrruerdreidn 
itnh ethHeS HLSALsAid sei.de. 
  
(a) (b) 
FiguFreig9u.rMe 9ic. rMositcrruocsttururecstuorfetsh oefH thAeZ HzAonZe zfoonreH fSoLr AHSsLidAe osifdSea omf pSlaem2p1:le( a2)1C: (GaH) CAGZHaAndZ (abn)dF G(bH) AZ.
FGHAZ. 
 
In the CGHAZ and FGHAZ of the MART side of Sample 21, as a result of the higher CE,
coarse baiting, ferritic and martensitic microstructures were identified (Figure 10a,b). The ICHAZ,
in accordance with the Fe-Fe3C equilibrium diagram, consisted of fine ferritic structures with small
portions of pearlitic structures (Figure 10c). As expected, under the influence of a relatively high
temperature, which is in the range of martensite tempering temperatures, tempered martensite formed
in SCHAZ of the MART side (Figure 10d).
Microhardness measurements were conducted in the various zones of Samples 20 and 21.
The microhardness of the BM of the HSLA and MART steels was measured as 213 and 404 Vickers,
respectively. The hardness profile of the welded joint section varies significantly because of the phase
transformations during the thermal cycle of the welding process. Figure 11 shows the microhardness
map of Sample 20. Figure 11 also presents the microhardness profile across the mid-section of
the sample. Due to the rapid cooling of FZ, each material showed an increase in hardness of FZ
relative to BM. The average microhardness value in the FZ is 480 Vickers and varies across the section.
This fluctuation is attributed to the mixed microstructure of the FZ. Different hardness of the martensitic
and bainitic microstructures could cause the fluctuation of the hardness profile. In addition, various
morphologies (i.e., columnar and equiaxed) in FZ could be a reason for the various hardness. However,
some researchers have focused to determine an empirical formula for FZ hardness using CE values;
in the present study, the measured hardness of FZ is higher than the calculated values using the
mentioned formulas [31,36]. The calculated hardness values using the formulas given in the literature
are 434 HV and 365 HV. In all compared zones, MART steel exhibited higher hardness values due to
the higher CE value, which has a significant influence on the hardenability. While the hardness of the
Metals 2016, 6, 245 13 of 17 
Metals 2016, 6, 245 14 of 17
In the CGHAZ and FGHAZ of the MART side of Sample 21, as a result of the higher CE, coarse 
baiting, ferritic and martensitic microstructures were identified (Figure 10a,b). The ICHAZ, in 
HacScLoArdasindcee ewxhitihbi ttshae sFhea-rFpe3iCnc ereqausielitbhrriuoumg hditahgerHamA,Z cuopnstiosttehde FoZf ,fitnhe MferAriRtiTc ssitdreucsthuorwess wa sitohf tesnminalgl 
zponrteioin sH oAf Zp.eTahrelitcioc nsttirnuucotusreisn c(rFeiagsuerter e1n0dc)i.n Aths eeHxpSeLcAtedsi,d uenwdaesr dtuhe tionftlhueefnecrer itoifc ma ircerloasttivrueclytu hreigohf 
HtemSLpAersatteuerl.e,T hwehtiecmh pies ring tzhoen reaanngde foerf rimtica/rtmenasritteen stietmicpdeurainl gp htaesmepsetrruactuturreess, itnemMpAeRreTds tmeealrcteanusietde 
afodrmecerdea isne SinCHhaArdZn oefs sth. e MART side (Figure 10d). 
  
(a) (b) 
  
(c) (d) 
FFiigguurree 1100.. DDeettaaiilleedd HHAAZZ mmiiccrroossttrruuccttuurreess aanndd ssuubbzzoonneess ooff tthhee MMAARRTT ssiiddee ooff SSaammppllee 2211:: ((aa)) CCGGHHAAZZ;; 
((bb)) FFGGHHAAZZ;; ((cc)) IICCHHAAZZ;; aanndd ((dd)) SSCCHHAAZZ.. 
Metals 2016, 6, 245 14 of 17 
Microhardness measurements were conducted in the various zones of Samples 20 and 21. The 
microhardness of the BM of the HSLA and MART steels was measured as 213 and 404 Vickers, 
respectively. The hardness profile of the welded joint section varies significantly because of the phase 
transformations during the thermal cycle of the welding process. Figure 11 shows the microhardness 
map of Sample 20. Figure 11 also presents the microhardness profile across the mid-section of the 
sample. Due to the rapid cooling of FZ, each material showed an increase in hardness of FZ relative 
to BM. The average microhardness value in the FZ is 480 Vickers and varies across the section. This 
fluctuation is attributed to the mixed microstructure of the FZ. Different hardness of the martensitic 
and bainitic microstructures could cause the fluctuation of the hardness profile. In addition, various 
morphologies (i.e., columnar and equiaxed) in FZ could be a reason for the various hardness. 
However, some researchers have focused to determine an empirical formula for FZ hardness using 
CE values; in the present study, the measured hardness of FZ is higher than the calculated values 
using the mentioned formulas [31,36]. The calculated hardness values using the formulas given in 
the literature are 434 HV and 365 HV. In all compared zones, MART steel exhibited higher hardness 
values due to the higher CE value, which has a significant influence on the hardenability. While the 
hardness of the HSLA side exhibits a sharp increase through the HAZ up to the FZ, the MART side 
shows a softening zone in HAZ. The continuous increase trend in the HSLA side was due to the 
ferritic microstructure of HSLA steel. The tempering zone and ferritic/martensitic d ual phase 
structures in MART steeFlF icigaguuurrseee 1d111 .a .M Mdieciccrroroehhaaasrrded ninnees sshs ammradappn aeansnsdd.  pprrooffiillee ooff SSaammppllee 2200.. 
 The microhardness map and profile of Sample 21 can be seen in Figure 12. The highest and the 
lowest microhardness values were measured in the BM of MART and HSLA, respectively. The 
highest value is related to the predominantly martensitic microstructure of the BM of MART steel. 
Among the weld zone of the MART steel, the ICHAZ showed the lowest microhardness 
corresponding with the ferritic-pearlitic microstructure. The measured microhardness values 
through the FZ showed a fluctuation, which can be a result of multiple morphologies of ferritic 
structures. In the HSLA side, microhardness values showed a decreasing trend up to the BM. 
 
Figure 12. Microhardness map and profile of Sample 21. 
4. Conclusions 
In this study, fiber laser welded dissimilar MART and HSLA steels have been evaluated with 
respect to tensile properties, microstructure and hardness profile. As the first step of this study, the 
process parameters of laser welded dissimilar steel joints have been optimized to maximize the TL 
and minimize the HI of the welded joints using the Taguchi method. The order of importance of the 
process parameters on the response was welding speed, laser power and focal position. The welding 
 
Metals 2016, 6, 245 14 of 17 
Metals 2016, 6, 245  15 of 17
Figure 11. Microhardness map and profile of Sample 20. 
TThhee miiccrroohhaarrddnneessss maapp aanndd pprroofifillee ooff SSaamppllee 2211 ccaann bbee sseeeenn iinn FFiigguurree 1122.. TThhee hhiigghheesstt aanndd tthhee 
llooweesstt micicrroohhaarrddnneessssv avlauleusews ewreerme emaseuarseudreind thine BthMe oBfMM AofR MT aAnRdTH aSnLdA ,HreSsLpAec, triveesplye.cTtihveelhyi.g hTehset 
vhaigluheesits vreallautee dis troetlahteedp rteod tohme ipnraendtolymminaarntetlnys imticarmteincsriotsictr mucitcurorestoruf cthtuerBe Mof othf eM BAMR Tofs tMeeAl.RATm stoenegl. 
tAhme wonegld tzhoen ewoefltdh ezMonAeR Tofs tethele, thMeAICRHT AsZtesehl,o wtheed tIhCeHlAowZe ssthmowicerodh atrhden elsoswceosrtr emspiocnrodhinargdwneitshs 
tchoerrfeesrpriotnicd-pinega rlwitiicthm tihcreo sfterrurcittiucr-ep.eTarhleitimc emasiucrroedstrmucictruorhe.a rTdhnee ssmveaalsuuersetdh romuigcrhohthaerdFnZesssh ovwaeludeas 
flthurcotuugahti otnh,e wFhZi cshhocawnebde aa frleuscutultaotifomn,u wltihpilcehm coanrp bheo lao grieessuoltf ofefr rmituicltsiptrluec mtuorerps.hIonlotghieesH oSfL Afersriidtiec, 
mstricurcothuarerds.n Iens sthvea HluSeLsAsh soiwdee,d maicdreochraeardsinnegsstr veanldueusp sthootwheedB aM d. ecreasing trend up to the BM. 
 
FFiigguurree 1122.. Miiccrroohhaarrddnneessss maapp aanndd pprroofifillee ooff SSaamppllee 2211.. 
44.. CCoonncclluussiioonnss 
IInn tthhiiss ssttuuddyy,, fifibbeerr llaasseerr wweellddeedd ddiissssiimmiillaarr MMAARRTT aanndd HHSSLLAA sstteeeellss hhaavvee bbeeeenn eevvaalluuaatteedd wwiitthh 
rreessppeecctt ttoo tteennssiillee pprrooppeerrttiieess,, mmiiccrroossttrruuccttuurree aanndd hhaarrddnneessss pprrooffiillee.. AAss tthhee ffiirrsstt sstteepp ooff tthhiiss ssttuuddyy,, tthhee 
pprroocceessss ppaarraammeetteerrsso offl alasesrerw weledleddedd idssisimsimilairlasrt esetel ejol ijnotisnhtsa vheavbee ebneeonp toimptiizmedizteodm toa xmimaxiziemtihzee TthLea TndL 
manidni mmiizneimthizeeH tIhoef HthIe owf tehlde ewdejlodinedts juosinintgs uthseinTga gthuec hTiamguetchhoi dm. eTthheodor. dTehreo of rimdepro orft ainmcpeoorfttahnecep roofc tehses 
pparoracmeses tpearsraomnettheersr eosnp tohnes reewspaosnwsee lwdainsg wsepledeindg,  lsapseeerdp,o lwaseerr apnodwfeorc aanl dp ofosictaiol np.oTsihtieonw. eTldhein wgeslpdeinedg 
w as found to be the most effective process parameter, and its interaction with the laser power should
be monitored for the HI and TL of the joints. It was observed that, if the HI was not sufficient due to
high speed or low laser power, the weld bead geometry was not formed appropriately. In addition,
when applying excessive HI, the HAZ would be wider, and that causes a decrease in TL. The optimum
combination of laser welding process parameters was a welding speed of 45 mm/s, a laser power of
1750 W and the focal position of −0.4 mm.
In the second step, the microstructural examination and microhardness evolution of the selected
welds that have the highest and lowest response values were discussed. Weld zone microstructures of
selected samples were completely distinct due to the different HI and consequently not associated with
CE values due to slow cooling rates. The HAZ of the samples was divided into five subzones, namely
PMZ, CGHAZ, FGHAZ, ICHAZ and SCHAZ, due to the grain transformations. Due to the phase
transformations during the thermal cycle of the process, the hardness profile of the welded sections
varies significantly. Due to the rapid cooling of FZ, each sample showed an increase in hardness of FZ
relative to BM. While the hardness of the HSLA side exhibits a sharp increase through the HAZ up to
the FZ, the MART side shows a softening zone in HAZ.
Metals 2016, 6, 245 16 of 17
Acknowledgments: The authors acknowledge the Uludağ University Commission of Scientific Research Projects
under Contract No. OUAP (MH)-2016/6 for supporting this research. Additionally, a part of this research was
supported by the Coskunoz Holding Research and Development Department at Bursa, Turkey.
Author Contributions: Celalettin Yuce and Fatih Karpat conceived of and designed the experiments.
Celalettin Yuce and Mumin Tutar performed the experiments and microstructural examination. Celalettin Yuce,
Mumin Tutar and Nurettin Yavuz analyzed the data. All of the authors discussed the results and commented on
the manuscript at all stages. All co-authors contributed to the manuscript proofing and submissions.
Conflicts of Interest: The authors declare no conflict of interest.
References
1. Kuziak, R.; Kawalla, R.; Waengler, S. Advanced high strength steels for automotive industry. Arch. Civ.
Mech. Eng. 2008, 8, 103–117. [CrossRef]
2. Saha, D.C.; Westerbaan, D.; Nayak, S.S.; Biro, E.; Gerlich, A.P.; Zhou, Y. Microstructure-properties correlation
in fiber laser welding of dual-phase and HSLA steels. Mater. Sci. Eng. A 2014, 607, 445–453. [CrossRef]
3. Liu, Y.; Dong, D.; Wang, L.; Chu, X.; Wang, P.; Jin, M. Strain rate dependent deformation and failure behavior
of laser welded DP780 steel joint under dynamic tensile loading. Mater. Sci. Eng. A 2015, 627, 296–305.
[CrossRef]
4. Benasciutti, D.; Lanzutti, A.; Rupil, G.; Haeberle, E.F. Microstructural and mechanical characterisation of
laser-welded lap joints with linear and circular beads in thin low carbon steel sheets. Mater. Des. 2014, 62,
205–216. [CrossRef]
5. Casalino, G.; Dal Maso, U.; Angelastro, A.; Campanelli, S.L. Hybrid laser welding: A review. DAAAM Int.
Sci. Book 2010, 38, 413–430.
6. Xu, W.; Westerbaan, D.; Nayak, S.S.; Chen, D.L.; Goodwin, F.; Zhou, Y. Tensile and fatigue properties of
fiber laser welded high strength low alloy and DP980 dual-phase steel joints. Mater. Des. 2013, 43, 373–383.
[CrossRef]
7. Parkes, D.; Xu, W.; Westerbaan, D.; Nayak, S.S.; Zhou, Y.; Goodwin, F.; Bhole, S.; Chen, D.L. Microstructure
and fatigue properties of fiber laser welded dissimilar joints between high strength low alloy and dual-phase
steels. Mater. Des. 2013, 51, 665–675. [CrossRef]
8. Parkes, D.; Westerbaan, D.; Nayak, S.S.; Zhou, Y.; Goodwin, F.; Bhole, S.; Chen, D.L. Tensile properties of
fiber laser welded joints of high strength low alloy and dual-phase steels at warm and low temperatures.
Mater. Des. 2014, 56, 193–199. [CrossRef]
9. Wang, J.; Yang, L.; Sun, M.; Liu, T.; Li, H. Effect of energy input on the microstructure and properties of butt
joints in DP1000 steel laser welding. Mater. Des. 2016, 90, 642–649. [CrossRef]
10. Wang, J.; Yang, L.; Sun, M.; Liu, T.; Li, H. A study of the softening mechanisms of laser-welded DP1000 steel
butt joints. Mater. Des. 2016, 97, 118–125. [CrossRef]
11. Rossini, M.; Spena, P.R.; Cortese, L.; Matteis, P.; Firrao, D. Investigation on dissimilar laser welding of
advanced high strength steel sheets for the automotive industry. Mater. Sci. Eng. A 2015, 628, 288–296.
[CrossRef]
12. Němeček, S.; Mužík, T.; Míšek, M. Differences between laser and arc welding of HSS steels. Phys. Proced.
2012, 39, 67–74. [CrossRef]
13. Zhao, Y.Y.; Zhang, Y.S.; Hu, W. Effect of welding speed on microstructure, hardness and tensile properties in
laser welding of advanced high strength steel. Sci. Technol. Weld. Join. 2013, 18, 581–590. [CrossRef]
14. Arslanoglu, N.; Yigit, A. Experimental investigation of radiation effect on human thermal comfort by Taguchi
method. Appl. Therm. Eng. 2016, 92, 18–23. [CrossRef]
15. Tutar, M.; Aydin, H.; Yuce, C.; Yavuz, N.; Bayram, A. The optimisation of process parameters for friction stir
spot-welded AA3003-H12 aluminium alloy using a Taguchi orthogonal array. Mater. Des. 2014, 63, 789–797.
[CrossRef]
16. Benyounis, K.Y.; Olabi, A.G. Optimization of different welding processes using statistical and numerical
approaches—A reference guide. Adv. Eng. Softw. 2008, 39, 483–496. [CrossRef]
17. Anawa, E.M.; Olabi, A.G. Optimization of tensile strength of ferritic/austenitic laser-welded components.
Opt. Lasers Eng. 2008, 46, 571–577. [CrossRef]
18. Anawa, E.M.; Olabi, A.G. Using Taguchi method to optimize welding pool of dissimilar laser-welded
components. Opt. Laser Technol. 2008, 40, 379–388. [CrossRef]
Metals 2016, 6, 245 17 of 17
19. Sathiya, P.; Jaleel, M.Y.A.; Katherasan, D.; Shanmugarajan, B. Optimization of laser butt welding parameters
with multiple performance characteristics. Opt. Laser Technol. 2011, 43, 660–673. [CrossRef]
20. Casalino, G.; Campanelli, S.L.; Ludovico, A.D. Laser-arc hybrid welding of wrought to selective laser molten
stainless steel. Int. J. Adv. Manuf. Technol. 2013, 68, 209–216. [CrossRef]
21. Acherjee, B.; Kuar, A.S.; Mitra, S.; Misra, D. A sequentially integrated multi-criteria optimization approach
applied to laser transmission weld quality enhancement—A case study. Int. J. Adv. Manuf. Technol. 2013, 65,
641–650. [CrossRef]
22. Benyounis, K.Y.; Olabi, A.G.; Hashmi, M.S.J. Effect of laser welding parameters on the heat input and
weld-bead profile. J. Mater. Process. Technol. 2005, 164, 978–985. [CrossRef]
23. Ruggiero, A.; Tricarico, L.; Olabi, A.G.; Benyounis, K.Y. Weld-bead profile and costs optimisation of the CO2
dissimilar laser welding process of low carbon steel and austenitic steel AISI316. Opt. Laser Technol. 2011, 43,
82–90. [CrossRef]
24. Olabi, A.G.; Alsinani, F.O.; Alabdulkarim, A.A.; Ruggiero, A.; Tricarico, L.; Benyounis, K.Y. Optimizing the
CO2 laser welding process for dissimilar materials. Opt. Lasers Eng. 2013, 51, 832–839. [CrossRef]
25. Reisgen, U.; Schleser, M.; Mokrov, O.; Ahmed, E. Optimization of laser welding of DP/TRIP steel sheets
using statistical approach. Opt. Laser Technol. 2012, 44, 255–262. [CrossRef]
26. Zhao, Y.; Zhang, Y.; Hu, W.; Lai, X. Optimization of laser welding thin-gage galvanized steel via response
surface methodology. Opt. Lasers Eng. 2012, 50, 1267–1273. [CrossRef]
27. Benyounis, K.Y.; Olabi, A.G.; Hashmi, M.S.J. Multi-response optimization of CO2 laser-welding process of
austenitic stainless steel. Opt. Laser Technol. 2008, 40, 76–87. [CrossRef]
28. SSAB Products. Available online: http://www.ssab.com/products/brands/docol (accessed on
22 August 2016).
29. Guo, W.; Crowther, D.; Francis, J.A.; Thompson, A.; Liu, Z.; Li, L. Microstructure and mechanical properties
of laser welded S960 high strength steel. Mater. Des. 2015, 85, 534–548. [CrossRef]
30. Khan, M.I.; Kuntz, M.L.; Biro, E.; Zhou, Y. Microstructure and mechanical properties of resistance spot
welded advanced high strength steels. Mater. Trans. 2008, 49, 1629–1637. [CrossRef]
31. Oyyaravelu, R.; Kuppan, P.; Arivazhagan, N. Metallurgical and Mechanical properties of Laser welded High
Strength Low Alloy Steel. J. Adv. Res. 2016, 7, 463–472. [CrossRef] [PubMed]
32. Coelho, R.S.; Corpas, M.; Moreto, J.A.; Jahn, A.; Standfuß, J.; Kaysser-Pyzalla, A.; Pinto, H. Induction-assisted
laser beam welding of a thermomechanically rolled HSLA S500MC steel: A microstructure and residual
stress assessment. Mater. Sci. Eng. A 2013, 578, 125–133. [CrossRef]
33. Talaş, Ş. The assessment of carbon equivalent formulas in predicting the properties of steel weld metals.
Mater. Des. 2010, 31, 2649–2653. [CrossRef]
34. Yurioka, N. Carbon Equivalents for Hardenability and Cold Cracking Susceptibility of Steels. In Proceedings
of the Select Conference on Hardenability of Steels, Derby, UK, 17 May 1990.
35. Santillan Esquivel, A.; Nayak, S.S.; Xia, M.S.; Zhou, Y. Microstructure, hardness and tensile properties of
fusion zone in laser welding of advanced high strength steels. Can. Metall. Q. 2012, 51, 328–335. [CrossRef]
36. Midawi, A.R.H.; Santos, E.B.F.; Huda, N.; Sinha, A.K.; Lazor, R.; Gerlich, A.P. Microstructures and mechanical
properties in two X80 weld metals produced using similar heat input. J. Mater. Process. Technol. 2015, 226,
272–279. [CrossRef]
37. Beidokhti, B.; Kokabi, A.H.; Dolati, A. A comprehensive study on the microstructure of high strength low
alloy pipeline welds. J. Alloys Compd. 2014, 597, 142–147. [CrossRef]
38. Beidokhti, B.; Dolati, A.; Koukabi, A.H. Effects of alloying elements and microstructure on the susceptibility
of the welded HSLA steel to hydrogen-induced cracking and sulfide stress cracking. Mater. Sci. Eng. A 2009,
507, 167–173. [CrossRef]
© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).