1-s2.0-S0924013613001982-main

Journal of Materials Processing Technology 213 (2013) 2145 2151

Contents lists available at ScienceDirect

Journal of Materials Processing Technology

jou rn al hom epage: www.elsev ier .com/ l

A parametric study of Inconel 625 wire laser dep

T.E. Abio

Manufacturing

a r t i c l

Article history:Received 14 FeReceived in reAccepted 10 JuAvailable onlin

Keywords:LaserDepositionInconel 625WireDilutionProcess characteristics

ng ponomsition

compg parng thferen

and c micio (%)

and energy per unit length of track were found to be key parameters inuencing both the process andtrack geometrical characteristics. Aspect ratio and dilution ratio showed positive dependency whereascontact angle showed negative dependency on energy per unit length of track. Conversely, material depo-sition volume per unit length of track varied directly with contact angle but inversely with aspect ratioand dilution ratio (ranging from 0% to 24%). Processing conditions at which a combination of favourablesingle track properties including low contact angle (

2146 T.E. Abioye et al. / Journal of Materials Processing Technology 213 (2013) 2145 2151

Table 1Chemical compositions of Inconel 625 wire and 304 stainless steel in wt.%.

Element Ni Cr Mn Si Al Ti Fe C Mo Nb P S

Inconel 625 Bal 22.46 0.26 0.26 0.14 0.02 8.84 3.46304 stainles Bal 0.08 0.10 0.03

is of paramvantageouset al., 2013)

Due to itemperaturity and goonickel baserange of aphardware aMore imporcoupled wiextended thindustry. Dstainless stcorrosion reoffered by suffer localchloride ionfact, the liferosive envithe surfaceof studies hpowder, for2009), optiwall manufcoatings (TuInconel 625not been un

Laser clwith powdeincreased msurface quaing the wirdeposition are highly sis importanparametersfeed directiand traversbe achieved

Three wedges, in thWhen the wLi (2005) diacterised bbeads wherthe centre establisheddeposition ewith front processing ZE41A-T5, eter designinclude theparametersprocess chafrom the litethe investigangle of las

his study, single tracks of Inconel 625 wire were depositeding processing parameters via laser cladding. The primaryves are of two folds. Firstly, a process map which pre-conel 625 wire bre laser deposition process characteristics

ying processing parameters will be developed. Secondly,sing parameters at which a combination of favourable sin-ck properties including low contact angle (

T.E. Abioye et al. / Journal of Materials Processing Technology 213 (2013) 2145 2151 2147

Fig. 2. Track characterisation (a) track metrics (b) SEM micrograph o

of experimental runs required. Real-time observation of the depo-sition runs laser head. used for thenations of lrates ranginwire tip cofor all possiselected x125 tracks)to provide a

2.3. Track c

Fig. 2a sThe tracks points alon(Talysurf Cgation, tracconductingsurface nisusing 70% The tracks tron microsQuantitativin determinconductingthe height o

2.4. Denit

Energy peffect of lasand it is detrack pass (Drate (WFR) in Eq. (2) is

EL =P

V

Table 2Processing par

Parameters

Laser powerWire feed raTraverse speSeparation dArgon gas oWire diametWire feeding

DVL =A W

V

rctan

(Xs coneighwhiln of q. (4.44

103

t of Fhile

ults

ire d

he w lasers chaire dical o

behs. Iningly

(Xu etion sion ap vam foom 1ped. d in tsmootively

mapby diwas recorded by a camera attached to precitec YW50Table 2 gives the details of the processing parameters

tracks deposition.Tracks were deposited with combi-aser power and traverse speeds with varying wire feedg from 6.7 mm s1 until stubbing was observed (i.e. thelliding with the substrate). This process was repeatedble combinations of processing parameters within theed range of laser power and traverse speed (totalling. Two tracks were deposited at each cladding condition

degree of verication.

haracterisation

hows a schematic diagram of typical track geometry.heights and widths were measured at three differentg their lengths using a Taylor Hobson surface prolerLI 1000). For the purpose of microstructural investi-k samples were transversely sectioned, mounted in

resin and sequentially ground and polished to a 1 mh. Then, the track samples were electrolytically etchedorthophosphuric acid in water (typically 6 V for 3 s).microstructures were examined under scanning elec-copy (SEM) using a back scattered electron (BSE) signal.e energy dispersive X-ray analysis (EDXA) was utiliseding the elemental composition of Fe in the tracks by

area scan (200 m 200 m) analysis along and acrossf each cross-sectioned track, as shown in Fig. 2b.

ion of terms

er unit length of track (EL) in J mm1 is a combineder power (P) in Watts and traverse speed (V) in mm s1,ned by Eq. (1). The deposition volume per unit length ofVL) in mm3 mm1 has the traverse speed and wire feed

in mm s1 as the determining variables. The constant A the cross-sectional area of the wire in mm2.

(1)

= 2 a

=s

Thetrack h2005) positios in Ewire; 88.04 percentively w

3. Res

3.1. W

As ttion ofproceswere ware typsystemprocesincreasrosiondeposidimencess m1.2 mging frdevelodeneoccur, respecon thefound ameters.

Value Unit

1.01.8 kWte 6.723.3 mm s1

ed 1.78.5 mm s1

istance between two consecutive tracks 10 mmw rate 0.42 l s1

er 1.2 mm angle 42 1 degrees

Wire deposof wire feedwire at a gi

Fig. 4 pdeposition transfer), smof wire. Thtrack of higwhen thereobserved wf a transversely sectioned sample.

FR(2)

(2HW

)(3)

c(Xc+s Xc) Xc+s) + c(Xc+s Xc) (4)

tact angle () was calculated from the values of thet (H) and width (W) using Eq. (3) (de Oliveira et al.,e the dilution ratio () was determined from the com-the track using Eq. (4) (Toyserkani et al., 2005). c and) are the densities of the feed material (Inconel 625

103 g mm3) and substrate (AISI 304 stainless steel;g mm3), respectively. Xc+s and Xs are the mean weighte in total surface of track region and substrate, respec-

Xc is the weight percent of Fe in the additive material.

and discussion

eposition characteristics

ire feed rate was sequentially varied for each combina- power and traverse speed, entirely different depositionracteristics were observed. The observed characteristicsripping, smooth wire transfer and wire stubbing. Thesef wire laser deposition process. However, each materialaved differently at every condition of laser depositionconel 625 coatings on oil and gas pipelines is being

utilised because of its excellent protection against cor-t al., 2013). As a result, optimising Inconel 625 wire laserprocess to producing continuous track with acceptableis essential to its application in this industry. A pro-lid for the bre laser deposition of Inconel 625 wire ofr laser power range of 1.01.8 kW, traverse speed ran-.7 to 5.0 mm s1 and wire feed rate 6.723.3 mm s1 wasAs shown in Fig. 3, ve different regions were clearlyhe map with 15 representing dripping, dripping mayth wire ow, stubbing may occur and stubbing regions,. Each processing condition was represented by a point

. The combined parameter on the y-axis of the map wasviding laser power by the traverse speed (see Eq. (1)).

ition volume per unit length on the x-axis is a function

rate, traverse speed and cross-sectional area of the feedven processing condition (see Eq. (2)).resents the typical examples of tracks deposited byprocess characterised with dripping of wire (dropletooth transfer of wire (smooth deposition) and stubbing

e ideal scenario (i.e. cladding conditions that produceh surface quality and acceptable dimensions) is found

is smooth transfer of wire into the meltpool. This washenever the wire tip melted at the point or close to the


Fig. 3

point of intthat produc3 of the pro

When thnation of trwire was obFig. 4. At a vwith the lasits melting.the meltpooas the substhe energy deposition

At a cladhigh for a feed wire inentered theof the wire tcollision withe centre omelted by tan irregularremained uvolume perper unit len

It can be observed from the map that region 1 (i.e. wire drippingregion) widens whereas region 5 (wire stubbing region) becomesnarrower with increasing energy per unit length of track. Thisconrms the fact that increased energy produced higher heat input,hence, quicmeltpool. Tobserved.

Also, asincreased iistics transithen wire sinated, as eof track or of track. Aleither increthe wire de

Region boundary b

gion.onde eacooth

difh depion 4e wiarac

wirnt trhat semen

J mm

ad ch

Widt wid

obta Sinc

sign(Pinkerefopped

5 sh. A process map for Inconel 625 wire laser deposited tracks.

ersection with the meltpool. The processing conditionsed smooth transfer of wire are contained in the regioncess map.e wire feed rate was excessively low for a xed combi-averse speed and laser power, intermittent dripping ofserved. This produced discontinuous tracks as shown inery low wire feed rate, the wire tip interacted too longer beam such that it absorbed heat energy sufcient for

As a result, the wire tip melted before intersecting withl. This produced intermittent dripping of molten wire

trate traversed. Similar effect was observed wheneverper unit length of track was excessive for a xed wirevolume per unit length of track.ding condition when the wire feed rate was excessivelyxed combination of traverse speed and laser power, theteracted briey with the laser beam. As result, the wire

meltpool in a nearly solid form resulting in the collisionip with the solid substrate at the base of the meltpool. A

tion recorrespbecausand smit wassmoot

Regand thcess chof bothdifferedicts tarrangof 200

3.2. Cl

3.2.1. The

resultstracks.ratio istracks can thoverla

Fig.

th the substrate caused the wire tip to move away fromf the meltpool. In the process, the wire was eventuallyhe energy in the meltpool and solidied as tracks with

shape. At an extremely high wire feed rate, the wirenmelted. This was due to the fact that wire deposition

unit length of track was too high for the given energygth of track.

Fig. 4. Typical deposits of Inconel 625 wire.

ing the travfeed rate, pwere foundAISI 304 stthe aspect per unit leincreasing per unit lenHowever, ttion volumeratio. Highenounced atrack geomany changeincrease th

3.2.2. DilutGeneral

the substrasubstrate. Timum (abodense bondker melting of the feed wire before its intersection of theherefore, more dripping and less stubbing of wire were

the wire deposition volume per unit length of trackn the process map, the deposition process character-ted from wire dripping to smooth wire deposition andtubbing. This shows that the wire dripping can be elim-xpected, by either reducing the energy per unit lengthincreasing the wire deposition volume per unit lengthso, wire stubbing can be reversed to ideal scenario byasing the energy per unit length of track or decreasingposition volume per unit length of the track.2 (i.e. wire dripping may occur region) formed theetween the dripping region and smooth wire deposi-

Laser deposition processes performed at the conditionsing to this region produced inconsistent characteristicsh of these processes were characterised with dripping

wire transfer effects after at least two trials. As a result,cult to correctly classify them either into dripping orosition region.

is the boundary between the smooth deposition regionre stubbing region. Due to inconsistence in their pro-teristics, cladding conditions that gave deposition rune stubbing and smooth wire transfer after at least twoials were grouped in this region. Finally, the map pre-mooth wire transfer may not be practicable, with thist, when cladding below energy per unit length of track1.

aracteristics

hheight aspect ratiothheight (WH) aspect ratio was calculated from theined from the height and width measurements of thee, past works have clearly indicated that track aspecticant to depositing inter-run porosity free overlappederton and Li, 2008), aspect ratio of single track depositsre be used as a quality criterion for the deposition of

tracks.ows that the track aspect ratio increases with increas-erse speed and laser power but with decreasing the wirerovided, other factors remain constant. Similar results

for bre laser micro-cladding of Co-based alloys onainless steel (Lusquinos et al., 2009). The response ofratio can be explained by the wire volume depositedngth of track. Decreasing the wire feed rate and/orthe traverse speed reduced the wire deposited volumegth of track. This resulted in track of reduced height.he track width was invariant with the wire deposi-

(mm3 mm1). These effects produced decreased aspectr aspect ratio obtained at higher power is due to pro-ttening effect (increase in width) of laser power on theetry. Since width is the numerator in the ratio therefore

that increases it and/or decreases H (denominator) wille value of the ratio.

ion ratioly, dilution is the percentage of the total volume ofte material in the track contributed by melting of thehough it is undesirable in cladding processes, some min-ut 38%) (Qian et al., 1997) is required before a fully

is achievable in a track.


Fig. 5. Variation of track aspect ratio as a function of the main processing parameters (a) P = 1.8 kW and (b) P = 1.4 kW.

Dilution ratio analysis was carried out from the composition ofthe tracks as described in Section 2.4. As shown in Fig. 6, elementalcomposition analysis (i.e. EDXA) established that Fe content, hence,the percentage dilution of the examined track samples increasedwith increasing laser power and traverse speed but with decreasingwire feed rate. As the laser power was increased, a higher volumeof the substrate was melted due to increased energy input. Also,there was sufcient mixing and vigorous meltpool movement. Thiscaused an increased percentage of the molten substrate (mainlyFe) mixing with the track layer thus producing higher dilutionratio.

At increased wire feed rate, more wire volume (due to increaseddeposition rate) was deposited into the meltpool producing biggertrack. Also, there was increased laser energy interruption by thefeed wire creaching thstrate was oto be highe

mixing in the meltpool. Eventually, there was reduced dilution ofFe from the substrate.

When the traverse speed was increased, two things becameapparent. Firstly, the deposited wire volume per unit length of trackreduced, producing smaller track. Secondly, the energy per unitlength of track also decreased causing reduced melted depth intothe substrate. However, as shown in Fig. 7, the change in melteddepth into the substrate is relatively insignicant compared withthe change in the track volume. As a result, increase in dilution ratiowas observed in the track as the speed increased.

3.2.3. Contact angleAs shown in Fig. 6, contact angle was found to increase with

increasing wire feed rate but decreased with increasing traverseand lnsiont angrease

Fig. 6. VariatioWFR = 13.3 mmausing signicant reduction in the fraction of energye substrate. As a result, low melted depth into the sub-bserved. Also, the viscosity of the meltpool is believedr at higher WFR therefore reducing the vigour, hence,

speed face tecontacan incn of dilution ratio and contact angle with the main processing parameters. (a) P = 1.8 kW, W s1.aser power. The results show that apart from the sur-, processing parameters also inuence the wetting (i.e.le) of laser deposited tracks. Fig. 8 clearly revealed that

in wire feed rate and/or decrease in traverse speedFR = 13.3 mm s1, (b) P = 1.8 kW, V = 1.7 mm s1 and (c) V = 1.7 mm s1,


Fig. 7. Etched optical macro-photographs of laser track cross-sections for Inconel 625 wire at laser power of 1800 W and wire feed rate of 10 mm s1.

(i.e. increasing the wire deposited volume per unit length of track)resulted in tracks becoming more spherical. At a low wire depo-sition volume per unit length of track, a parabolic shaped trackof contact angle lower than 90 was formed. However, as thewire feed rate increased or traverse speed decreased, more wirewas deposited per unit length of track until it was sufcient toform a track strip with a more spherical shape in cross-section.Further increases to wire deposition volume resulted in swollenanks of the track thus producing track with an obtuse contactangle.

High energy per unit length of track resulting from increasedlaser power produced hotter meltpool. The high energy meltpoolexpedited the melting of the wire, increased the uidity and vigourof the meltpool. This caused the molten pool to spread, instead ofbuilding height, away from its centre. As a result, the solid substrateat meltpool boundary melted and the meltpool size increased.Eventually, wider tracks with low contact angles were formed.

3.2.4. Overlapping tracksGeometrical characterisation of all the deposited tracks revealed

that there is a sharp contrast in the growth trends of contact angleand dilution ratio with the processing parameters.

The concern was to determine suitable processing condition(s)that give good surface quality tracks with low contact angle (


Table 3Processing conditions at which a combination of favourable single track properties including low contact angle (

Documents

1-s2.0-S0924013613001982-main