Stability and Heat Input Controllability of Two Different ... - MDPI

applied sciences Article

Stability and Heat Input Controllability of Two Different Modulations for Double-Pulse MIG Welding Jiaxiang Xue *, Min Xu, Wenjin Huang, Zhanhui Zhang , Wei Wu and Li Jin School of Mechanical and Automotive Engineering, South China University of Technology, Guangzhou 510641, China; [email protected] (M.X.); [email protected] (W.H.); [email protected] (Z.Z.); [email protected] (W.W.); [email protected] (L.J.) * Correspondence: [email protected]; Tel.: +86-020-2223-6360 Received: 2 November 2018; Accepted: 21 December 2018; Published: 1 January 2019







Abstract: Aluminum alloy welding frequently experiences difficulties such as heat input control, poor weld formation, and susceptibility to pore generation. We compared the use of two different modulations for double-pulse metal inert gas (MIG) welding to reduce the heat input required to generate oscillations in the weld pool. The stabilities of rectangular wave-modulated and trapezoidal wave-modulated double-pulse MIG welding (DP-MIG and TP-MIG) were analyzed by examining their welding processes and weld profiles. We found that the transitional pulse in TP-MIG welding results in smoother current transitions, softer welding arc sounds, and a highly uniform fish-scale pattern. Therefore, TP-MIG welding is more stable than DP-MIG welding. The effects of these double-pulse modulation schemes on welding input energy are presented. We propose methods for reducing welding input energy by varying the number of pulses or the pulse base time of low-energy pulse train while keeping the welding current and welding arc stable and unchanged. Compared to DP-MIG welding, TP-MIG welding reduces the input energy by 12% and produces finer grain sizes, which increases weld hardness. Therefore, TP-MIG welding offers a new approach for heat input control in DP-MIG welding of aluminum alloys. The results of this work are significant for aluminum alloy welding. Keywords: double-pulse MIG welding; rectangular waves; trapezoidal waves

1. Introduction Despite the constant evolution of welding materials and structures, aluminum is still considered to be an extremely promising welding material owing to its abundance, lightness, corrosion resistance, and ductility [1]. Approximately 91% of all industries in China use aluminum products, including container production, extrusion profile manufacturing, rail vehicle manufacturing, and aerospace. Therefore, the development of the aluminum industry is closely related to GDP growth in China [2,3]. Today, the rapid development of the economy and intense market competition has led to a constant requirement for a higher level of productivity. Consequently, the developments in welding technology have been dominated by increases in welding speed [4]. High-speed gas metal arc welding is performed at speeds higher than 1 m/min [5]. However, performing high-speed welding using aluminum materials is challenging, owing to the physical properties of aluminum. For example, the electrical conductivity of aluminum is 62% IACS (International Annealed Copper Standard), which is approximately 6 times that of low-carbon steel, whereas its thermal conductivity is 222 [W/(m·K)] at 25 ◦ C, which is approximately 5 times that of low-carbon steel, making high-speed welding of aluminum more difficult [6–9]. Nonetheless, the continuous development of welding techniques has been accompanied by improvements in high-speed aluminum welding techniques. Appl. Sci. 2019, 9, 127; doi:10.3390/app9010127

www.mdpi.com/journal/applsci

Appl. Sci. 2019, 9, 127

2 of 18

The welding speed and welding current can affect heat input, which can also affect the microstructure and mechanical properties [10–13]. Pulsed gas metal arc welding (P-GMAW) is a method based on electric pulses that was invented in the 1960s by English researchers. In this method, the melting of the welding wire and base metal is controlled by peak pulse current, whereas the pulse base current maintains the welding arc. This technique has significantly improved the range of adjustment for welding current. In addition, this technique prevents weld orientation from having a significant impact on welding processes during droplet transfer. However, the welds formed using the P-GMAW technique are frequently accompanied by pores, which impairs the finish and thus limits the applicability of this technique [14]. Double-pulse GMAW (DP-GMAW) is based on the P-GMAW technique; it involves the modulation of high-frequency pulses by a low-frequency waveform [15]. The advantages of DP-GMAW are lower costs and higher suitability for outdoor applications. The DP-GMAW technique has various implementations. One approach is the low-frequency switching of wire feed speed and welding current (e.g., the Kemppi Pro Evolution) [16], while another approach is the low-frequency switching of the melting rate at a constant wire feed speed (e.g., OTC welding systems). At present, the low-frequency modulation of welding current provides a better level of welding quality [17]. In this mode of modulation, low frequencies are used to modulate two large-amplitude high-frequency pulse trains. High-frequency current facilitates droplet transfer, whereas low-frequency modulation oscillates the weld pool using strong and weak pulse trains. This helps in rapidly ejecting hydrogen gas from the weld pool and thus reduces the formation of pores in the weld [15]. In the design of current pulse parameters for double-pulse welding, different transfer modes can be used to switch the currents of strong and weak pulses. In this work, we investigated the welding stability and heat input controllability of two methods for the switching of strong and weak pulses, namely, rectangular-pulse metal inert gas (DP-MIG) welding and trapezoidal-pulse MIG (TP-MIG) welding. The findings of this work will be useful for the design of process parameters in double-pulse MIG welding. 2. Design of Current Waveform Parameters for Double-Pulse MIG Welding with Two Different Modulations In double-pulse MIG welding, waveform control refers to the modulation of high-frequency pulses using a low-frequency rectangular waveform. Strong and weak pulse trains are continuously alternated during this process, which forces gases formed by the welding processes out of the weld. Therefore, double-pulse MIG welding reduces the effects of weld pores, reduces the sensitivity of the joint to cracks, and improves the finish and mechanical properties of the weld [18]. Pulsed MIG welding is an important breakthrough in welding technology, as it has eliminated numerous constraints of constant-current welding techniques. Moreover, the advent of this technology has provided a new direction for research and development in the field of arc welding [19]. The energy used by pulsed MIG welding can be controlled by adjusting time and current parameters. This ensures that low current will be used during the welding of aluminum alloy plates, in accordance with the requirements of aluminum alloy processing [20]. The relationship between current and the transfer mode of double-pulsed MIG welding is illustrated in Figure 1. The globular transfer mode is essentially open-arc welding, and it has lower levels of current compared to spray transfer. As shown in Figure 2, the welding current of the spray transfer mode increases in proportion with the surface tension of molten metal, and this mode uses higher currents than several other modes of transfer. This mode of transfer is suitable for settings with low base voltages and high peak currents [21]. The streaming transfer mode leads to higher currents, lower droplet volumes, and formation of axial droplet columns. If the current exceeds 300 A, the electrical resistivity of the transfer process increases. The control of current in pulsed transfer is performed by switching between globular transfer and spray transfer [22].

Appl. Sci. 2019, 9, x FOR PEER REVIEW

3 of 19

Appl. Sci. 2019, 9, 127 Appl. Sci. 2019, 9, x FOR PEER REVIEW

3 of 18 3 of 19 Fluidic transition

high high current current low

Jet transition Fluidic transition Jet transition Large drop transition Large drop transition Short circuiting transition

type

low

Short circuiting transition

Figure 1. Relationship between transfer mode type and current. Figure 1. Relationship between transfer mode and current.

Figure 2 illustrates the1.current waveform oftransfer single-pulse MIG (P-MIG) welding. This process Figure Relationship between mode and current. only involves a few factors, and the most important parameters are the peak current This of theprocess pulse, Ip, Figure 2 illustrates the current waveform of single-pulse MIG (P-MIG) welding. Figure 2 illustrates the current waveform of single-pulse MIG (P-MIG) welding. This process and the base current of the pulse, I b . It should be noted that I p and I b are the extreme values that only involves a few factors, and the most important parameters are the peak current of the pulse, only involves a few factors, and the most important parameters are the peak current of the pulse, I p, during the adjustment period the pulses. As Ithat b is typically small, it is sufficient only Ip , manifest and the base current of the pulse, Ib . It of should be noted Ip and Ib very are the extreme values that and base current of the pulse, b. It should be noted Ip and Ib are theaitextreme values forthe powering idleadjustment operation and Imaintaining theAs welding arc. Ib very also plays role in the preheating manifest during the period of the pulses. Ib isthat typically small, is sufficient onlythat for manifest during the adjustment period of the pulses. As I b is typically very small, it is sufficient only process.idle operation and maintaining the welding arc. Ib also plays a role in the preheating process. powering for powering idle operation and maintaining the welding arc. Ib also plays a role in the preheating current (A) process. Ip

pulse transition current pulse transition current average current

current (A) Ip

Ib tp

Ib

0 tp

0

tb

tb

T

average current

time (ms)

Figure 2. Sketch of the P-MIG scheme. Figure 2. Sketch of the P-MIG scheme. T time (ms)

Double-pulse MIG welding effectively decreases crack sensitivity pore generation generation rates. rates.In Double-pulse MIG welding effectively decreases crack sensitivity and and pore Figure 2. Sketch of the P-MIG scheme. In addition, addition,the thewelding welding patterns formed by this method are aesthetically pleasing. As shown patterns formed by this method are aesthetically pleasing. As shown in Figure in 3, Figure 3, low-energy and high-energy pulse trains alternate with each other in double-pulse Double-pulse MIG welding effectively decreases pore generation rates. In low-energy and high-energy pulse trains alternatecrack withsensitivity each otherand in double-pulse MIG welding. MIG The of theshown parameters shown Figure 3 are as Ibw ) and tbwthe addition, the welding formed by this method pleasing. in Figure Thewelding. meanings of meanings thepatterns parameters in Figure 3 are areinaesthetically as follows: Ibw (Ipwfollows: ) andAs tbwshown (tpw(I) pw represent (tpw ) represent the high-energy base (peak) current and of base time the pulses in thetrain, low-energy pulse Ibs 3, low-energy alternate with other in double-pulse MIG welding. base (peak) and current and base pulse (peak)trains time the(peak) pulses ineach theoflow-energy pulse respectively; train, respectively; I (I ) and t (t ) represent the base (peak) current and base (peak) time of ps the base ps in current The parameters shown Figure 3and are base as follows: bw (Ipw tbw (tpw) represent bs bs (peak) (Ipsmeanings ) and tbs (tof ps) the represent (peak) Itime of) and the high-energy pulse the train, therespectively; high-energy train, respectively; is pulses the pulse period oflow-energy theand low-energy and n and base (peak) current and base (peak) time ofTwthe in the pulse train,train; respectively; Ibs of Tpulse w is the period of the low-energy train; n and m pulse represent the number mpshigh number of high energy pulses and low-energy pulses per unit welding current (I )represent andenergy tbs (tpsthe ) represent the base (peak) current and base (peak) time of the high-energy pulse train, pulses and low-energy pulses per unit welding current cycle, respectively. cycle, respectively. respectively; Tw is the period of the low-energy pulse pulse train; trains and nover and the m represent theT number The alternation of high-energy and low-energy duration of (i.e., the of time The alternation of high-energy and low-energy pulse trains over the duration of T (i.e., the time high energy pulses and low-energy pulses per unit welding current cycle, respectively. required to complete the processes of double-pulse MIG welding) causes the welder to alternate required to hot complete the processes of double-pulse MIG welding) the welder to alternate The alternation high-energy and low-energy pulse trains over causes duration of Tpoint (i.e., the between andofcold states [23]. As aluminum alloys have athelow melting andtime high between hot andthese cold states [23].are As easily aluminum alloys awelding) low melting point and high required to complete the processes of double-pulse MIG causes the welder to alternate conductivity, materials affected by have temperature, which could lead to conductivity, hot cracks and these materials are easily affected temperature, which could lead hot melting cracks and weldand damage. between hot and cold states [23]. As film aluminum alloys a to low point high weld damage. The possibility ofby oxide breakdown andhave the weakening of base material strength The possibility of oxide film breakdown and the weakening of base material strength in the presence conductivity, these of materials are easily affected temperature, lead to input hot cracks and in the presence high energy levels make by it necessary to which strictlycould control the of welding of energy. high energy The levels make it necessary to strictly controland thethe input of welding energy. weld damage. possibility of oxide film breakdown weakening of base material strength in the presence of high energy levels make it necessary to strictly control the input of welding energy.

Appl. Sci. 2019, 9, x FOR PEER REVIEW Appl. Sci. 2019, 9, 127 Appl. Sci. 2019, 9, x FOR PEER REVIEW

current (A) current (A) tps

Ips、Ipw Ips、Ipw

4 of 19 4 of 18 4 of 19

n Tsn Ts

tbs tps tbs

m Twm Tpw TbwTw Tpw Tbw

Ibs Ibs Ibw Ibw 0 0 Figure 3. DP-MIG current waveform diagram.

time (ms) time (ms)

Figure 3. DP-MIG current waveform diagram.

Owing to welding behaviors and environmental effects, the arc length during double-pulse to behaviors and environmental the the arc length during double-pulse MIG MIG Owing welding iswelding always in a varying state. Thus,effects, double-pulse MIG welding current has a towelding behaviors and environmental effects, arc length during double-pulse welding is waveform. always in aWhen varying Thus, double-pulse MIG welding current hasto alow rectangular rectangular pulse train undergoes transition from high energy energy, MIG welding is always in the a state. varying state. Thus, adouble-pulse MIG welding current has a waveform. When thewire pulse train a transition from highthey energy tohigh low energy, the arcvalues, length the arc length and extension continue to increase until reach their maximum rectangular waveform. When theundergoes pulse train undergoes a transition from energy to low energy, and wirethe extension todecreases. increase until reach their maximum values, whereas thereduced current whereas current parameter Thisthey results in considerable in energy. The the arc length andcontinue wire extension continue to increase until they change reach their maximum values, parameter decreases. This results in considerable change in energy. The reduced current eventually current becomes insufficient to This sustain the in arc, which causes the arc to breakdown. The whereaseventually the current parameter decreases. results considerable change in energy. The reduced becomes insufficient sustain the causesthe thearc, arcwhich to breakdown. The of this type of effects this type becomes oftodamage on arc, the which base metal are particularly significant intolow-temperature currentofeventually insufficient to sustain causes the arceffects breakdown. The damage theIn base are particularly significant in low-temperature environments. In this environments. thismetal work, trapezoidal investigatedsignificant to address this issue. Ourwork, new effects ofonthis type of damage on the waveforms base metalwere are particularly in low-temperature trapezoidal were investigated to address thisinvestigated issue. Our new approach referred as approach is waveforms referred as TP-MIG welding; the current waveform ofaddress this scheme is shown in environments. In this to work, trapezoidal waveforms were to thisisissue. Ourto new TP-MIG the current waveform of this scheme is shown in Figure Figure 4. welding; approach is referred to as TP-MIG welding; the current waveform of 4.this scheme is shown in Figure 4.

Ip Ip

current (A) current (A) Ts Td Ts Td

Tw Tw

Tu Tu

Ibs Ibs Ibw Ibw 0 0

time (ms) time (ms) Figure 4. TP-MIG low-frequency modulation waveform.

Figure 4. TP-MIG low-frequency modulation waveform.

4. TP-MIG low-frequency modulation waveform. In TP-MIG welding,Figure the transition of a high-energy pulse train to a low-energy pulse train In TP-MIG welding, the transition of a high-energy pulse train to a low-energy pulse train in in DP-MIG welding is supplemented by a transitional high- to low-energy pulse train, whereas the DP-MIG weldingwelding, is supplemented by a of transitional high-pulse to low-energy pulse train, pulse whereas the In TP-MIG the transition a high-energy train to a low-energy train in opposite transition (low-energy pulse train to high-energy pulse train) is supplemented by a transitional opposite pulse to high-energy train) pulse is supplemented bythe a DP-MIG transition welding is(low-energy supplemented by atrain transitional high- to pulse low-energy train, whereas low- to high-energy pulse train. The rectangular waveform exhibits trapezoidal characteristics owing transitional low- to (low-energy high-energy pulse pulse train train.toThe rectangular waveform trapezoidal opposite transition high-energy pulse train) is exhibits supplemented by a to the transitional pulse trains. When the arc extends to its maximum length during the high-energy characteristics owing the transitional pulse trains. the arc extends to its maximum length transitional lowto tohigh-energy pulse train. TheWhen rectangular waveform exhibits trapezoidal to low-energy transition of the pulse train, the current gradually decreases because of the pulse during the high-energy low-energy transition of theWhen pulsethe train, current decreases characteristics owing totothe transitional pulse trains. arcthe extends to gradually its maximum length train, which is conducive for controlling arc breakdown. This subsequently improves the efficacy because of the pulse train, which is conducive arc breakdown. This subsequently during the high-energy to low-energy transitionfor of controlling the pulse train, the current gradually decreases of aluminum alloy welding and ensures the construction of an energy gradient that improves the improves thethe efficacy aluminum welding and ensures the arc construction of an energy gradient because of pulseoftrain, which alloy is conducive for controlling breakdown. This subsequently controllability and stability of the welding process. that improves the controllability andalloy stability of the welding process. improves the efficacy of aluminum welding and ensures the construction of an energy gradient that improves the controllability and stability of the welding process.

Appl. Sci. 2019, 9, 127

5 of 18

The analysis of low-frequency modulation signals includes two aspects: peak values and base values. If the peak values are stable, low-frequency modulation may be performed using Ip . The change in base current along the four aforementioned modes of transfer is shown in Figure 4. Td and Tu are the low- to high-energy and high- to low-energy pulse transition times, respectively, and Td = Tu . In Figure 4, Tl is the time required to complete the processes of low-frequency modulation, and the corresponding frequency is f l = T1 . Hence, the time, t, at which the interval [0, Tl ] lies in the l low-frequency period is as follows: et1 = t − I NT

  t × T1 . T

(1)

Therefore, the base current can be expressed as follows for all intervals of the low-frequency signal:   Ibs    0 ≤ etl ≺ Ts I − Ibw   (etl − Ts )  Ibs − bs Ts ≤ etl ≺ Ts + Td Td Ibl = (2)  Ibw Ts + Td ≤ etl ≺ Ts + Td + Tw    I − Ibw e  e   Ibw + bs (tl − Ts − Td − Tw ) Ts + Td + Tw ≤ tl ≺ T Td The analysis of high-frequency modulation signals involves four different stages. The first stage is the high-energy pulse train. In this stage, the energy is highly concentrated, and the melting depth of the weld pool is the most important regulatory factor. The next stage is the transitional highto low-energy pulse, followed by the low-energy pulse train. The behaviors in this stage include preheating of the material and stabilization of the weld pool. The final stage is the transitional lowto high-energy pulse. During this stage, the peak currents of the high-energy and low-energy pulse Appl. Sci. 2019, 9, x FOR PEER REVIEW 6 of 19 trains are always the same, but there are marked differences in their base currents (Figure 5). current (A) Ts Ip

current (A) Td

Tw

Tu

Ts

Ths

Ip

Ibs

Ibs

Ibw

Ibw

0

time (ms)

Td

Tw Thw

0

(a)

Tu

time (ms)

(b)

Figure 5. 5.High-frequency signal energy energy pulse pulsegroup groupgraph: graph:(a)(a) High-energy pulse Figure High-frequency modulation modulation signal High-energy pulse train train of the high-frequencymodulation modulationsignal; signal;(b) (b) Low-energy Low-energy pulse train of the high-frequency train ofofthe thehigh-frequency high-frequency modulation signal. modulation signal.

Suppose thattohigh-frequency modulation occurs over a periodpulse of Thtrain , andisthe the According the same principles, the period of the low-energy Thw,period and its of interval high-energy pulse train is T , corresponding to a frequency of f = 1/T . The peak of the high-energy hs current hs values corresponding to the is [Ts + Td, Ts + Td + Ths hw]. In this interval, the time and pulse train is the same as that of the low-frequency modulation signal, whereas its base value is Ibs . high-frequency signal are: The duty cycle is Ds . Based on these parameters, the corresponding time of the high-frequency signal ~ tl  Ts  Td  in [0, Ths ] is: ~ ~    Thw thw  tl  Ts  Td INT (5) etl  T eths = etl − I NT × T . (3) hw   hs Ths

 I p  I bs I hs    2

  I p  I bs      2

  ~ t  Ts  Td  sign ~ tl  Ts  Td  INT  l Ths   

    Ths  Ds Thw  . (6)  

It should be noted that the following relation must be satisfied to ensure the transfer of one droplet for each pulse:

DT D T .

(7)

Appl. Sci. 2019, 9, 127

6 of 18

The current values corresponding to this period can be expressed as:  Ihs =

I p + Ibs 2





−

     etl I p − Ibs e sign tl − I NT × Ths − Ds Ths . 2 Ths

(4)

According to the same principles, the period of the low-energy pulse train is Thw , and its interval is [Ts + Td , Ts + Td + Thw ]. In this interval, the time and current values corresponding to the high-frequency signal are:   et − Ts − Td ethw = etl − Ts − Td − I NT l × Thw (5) Thw         etl − Ts − Td I p − Ibs I p + Ibs e − sign tl − Ts − Td − I NT × Ths − Ds Thw . (6) Ihs = 2 2 Ths It should be noted that the following relation must be satisfied to ensure the transfer of one droplet for each pulse: Ds Ths = Dw Thw . (7) This can be rewritten as:

Ds T f = hw = hs . Dw Ths f hw

(8)

The high- to low-energy and low- to high-energy transitions have a fixed time span, and the corresponding energy variation plots are shown in Figure 6. In the interval [Ts , Ts + Td ] of the high- to low-energy transition, the base current can be expressed as: Ib = Ibs − Appl. Sci. 2019, 9, x FOR PEER REVIEW

Ibs − Ibw e (t1 − Ts ). Td

current (A) Ts Ip

(9) 7 of 19

current (A)

Td Ths

Tw

Tu

Ts

Td

Tw

Ip

Ibs

Ibs

Ibw

Ibw

0

Tu Ths

0

time (ms)

time (ms)

(a)

(b)

Figure 6. 6. Transition Transition pulse pulse group group of of high-frequency high-frequency modulation modulation signal: signal: (a) (a)HighHigh- and and low-energy low-energy Figure transitionpulse pulsegroup; group;(b) (b)Low-energy Low-energy transition transition pulse pulse group. group. transition

Theperiod periodin inFigure Figure 6a 6a that that corresponds corresponds to to the the interval interval [T [Tss,,TTss++TThshs ] is: The ] is:  ~  Ts s  ~ = etl ~− I NT etlt− l T ethd ×TThdhd −TTss. . thd  tl  INT  T hs

(10) (10)

 Ths 

The corresponding currents are given as: The corresponding currents are given as:    I −I I −I     − bs bwI bs(et1−I bwTs )~  I p − Ibs + bs I bsbw(etI1bw− T~s )    et − Ts  I p + Ibs T T ( ) (    I I t T I I t Ts )  e     p bsd d 1 − 1  s   p bs  = sign t − I NT 1 ~ × Ths − Ts − Ds Ths   T T t T      ..   sign ~ d d 2 2 I  t  INTThs1 s   T  T  D T 

Ihd

hd

  

2

     

2

  

 

 T   hs 

hs

s

s hs

(11)

 (11) 

Similarly, the base current corresponding to the interval [Ts + Td + Tw, T] of the transitional lowto high-energy pulse train can be determined. In Figure 6b, the period corresponding to the interval [Ts + Td + Tw, T] is:

~ t T T T 

Appl. Sci. 2019, 9, 127

7 of 18

Similarly, the base current corresponding to the interval [Ts + Td + Tw , T] of the transitional lowto high-energy pulse train can be determined. In Figure 6b, the period corresponding to the interval [Ts + Td + Tw , T] is:  ethl = etl − I NT

etl − Ts − Td − Tw Ths



× Thd − Ts − Td − Tw .

(12)

The corresponding current values can be expressed as: 

Ihl

 I p + Ibw + = 

   Ibs − Ibw I − Ibw (et1 − Ts − Td − Tw )   I p − Ibw − bs (et1 − Ts − Td − Tw )  Td Td −     2 2

(13)

    et1 − Ts − Td − Tw e ×sign t1 − I NT × Ths − Ts − Ds Ths Ths

The calculations show that the total number of high- (low-) energy pulse trains will decrease with increases in tbw and tbs , if Tw and Ts are fixed. Therefore, if the current is to be limited to lower values, it is necessary to reduce n and m while appropriately increasing tbw and tbs . In addition, the dependence of current on the number of pulses must be closely monitored to minimize the latter’s impact on current. In conventional DP-MIG welding, the energy of the welding processes undergoes rapid switching, which is detrimental to welding stability. In addition, this increases the probability of screw-like wires being formed the arc is started in “cold plate” conditions. Appl. Sci. 2019, 9,when x FOR PEER REVIEW 8 of 19 3.3.Comparison Comparisonof ofStabilities Stabilities of of DP-MIG DP-MIG Welding Welding and and TP-MIG TP-MIG Welding Welding The stabilities of TP-MIG welding and DP-MIG welding were compared by analyzing the sound The stabilities of TP-MIG welding and DP-MIG welding were compared by analyzing the ofsound their of arcs, their transient welding current waveforms, and their weld morphologies. their arcs, their transient welding current waveforms, and their weld morphologies. An welding system system and and aa 1.2-mm 1.2-mm(ER4043) (ER4043) Anexperiment experiment was was conducted conducted using using an an automated automated welding aluminum alloy welding wire. The weld was formed on an AA6061 alloy plate (with thicknesses aluminum alloy welding wire. The weld was formed on an AA6061 alloy plate (with thicknesses inin the of of 1–51–5 mm) at a constant wire feed standardized current levels in this experiment therange range mm) at a constant wirespeed. feed The speed. The standardized current levels in this were 60 A, 80 A, 100 A, 120 A, 140 A, 160 A, 180 A, and 200 A. The DP-GMAW long-step experiment were 60 A, 80 A, 100 A, 120 A, 140 A, 160 A, 180 A, and 200 A. The DP-GMAWcalibration long-step curve for 1.2curve mm ER4043 alloy weldingalloy wires (with tbswires and t(with 8 ms, being s and n bw being calibration for 1.2 aluminum mm ER4043 aluminum welding tbs and tbwnbeing 8 wms, ns 15, Ipsnbeing 470 being 330 Ibs 330 andA, Ipwand being 45 A) constructed expert data and w being 15,A,IpsIpw being 470 A,A, Ipwand being Ibs and Ipwwas being 45 A) wasusing constructed usingis shown Figure 7. expertin data is shown in Figure 7. Peak time of strong pluse Peak time of weak pluse

7

Time (ms)

6 5 4 3 2 1 0

0

50

100

150

200

250

Welding current (A)

Figure aluminum alloy alloy welding weldingwire. wire. Figure7.7.1.2-mm 1.2-mm calibration calibration curve step distance of aluminum

3.1. Process Stability of DP-MIG Welding The current waveform and welding morphology corresponding to a welding current of 80 A are shown in Figure 8. The figure shows that DP-GMAW creates fish-scale patterns that are more aesthetically pleasing, and sharper welding-arc sounds. Welding was performed using numerical values generated by the Newton’s interpolation algorithm. The results of this experiment are shown

1 0

0

50

100

150

200

250

Welding current (A) Appl. Sci. 2019, 9, 127

8 of 18

Figure 7. 1.2-mm calibration curve step distance of aluminum alloy welding wire.

3.1. Process Process Stability Stability of of DP-MIG DP-MIG Welding Welding 3.1.

600

600

450

450

Current (A)

Current (A)

The current current waveform waveform and and welding welding morphology morphology corresponding corresponding to to aa welding welding current current of of 80 80 A A The are shown shown in Figure 8. The The figure figure shows shows that that DP-GMAW DP-GMAW creates creates fish-scale fish-scale patterns patterns that that are are more are aesthetically pleasing, pleasing, and and sharper sharper welding-arc welding-arc sounds. sounds. Welding Welding was was performed performed using numerical aesthetically areare shown in values generated generated by by the the Newton’s Newton’sinterpolation interpolationalgorithm. algorithm.The Theresults resultsofofthis thisexperiment experiment shown Figure 9. The figure shows that fish-scale patterns of higher produced using the numerical in Figure 9. The figure shows that fish-scale patterns of quality higher were quality were produced using the values generated the Newton–Raphson method in welding result highlights the numerical values by generated by the Newton–Raphson methodoperations. in weldingThis operations. This result superioritythe of computational algorithms in autoregulation of welding operations. highlights superiority of computational algorithms in autoregulation of welding operations.

300

150

0 0

200

400

600

800

300

150

0

1000

0

Time (ms) Appl. Sci. 2019, 9, x FOR PEER REVIEW

200

400

600

Time (ms)

Appl. Sci. 2019, 9, x FOR PEER REVIEW

(a)

(b)

(c) (c)

(d) (d)

800

1000

99 of of 19 19

Figure 8. 80 A A welding welding waveform; (b) Figure 8. Waveform and and weld weld formation formation at at 80 80 A A and and 160 160 A: A: (a) (a) 80 (b) 160 160 A A 8. Waveform welding waveform; welding waveform; (c) 80 A weld formation; (d) 160 A weld formation. welding welding waveform; waveform; (c) (c) 80 80 A A weld weld formation; formation; (d) (d) 160 160 A A weld weld formation. formation.

(a) (a)

(b) (b)

(c) (c) byby parameter self-adjusting welding test; test; (a) 83(a) A 83 weld (b) 92 Figure 9. Weld seam obtained parameter self-adjusting welding A weld Figure 9. 9. Weld Weldseam seamobtained obtained by parameter self-adjusting welding test; (a) 83 A formation; weld formation; formation; A weld formation; (c) 135 A weld formation. (b) (b) 92 92 A A weld weld formation; formation; (c) (c) 135 135 A A weld weld formation. formation.

3.2. Process Stability of TP-MIG Welding 3.2. 3.2. Process Process Stability Stability of of TP-MIG TP-MIG Welding Welding A TP-MIG welding experiment was conducted using a fixed current of 120 A and a range A welding experiment was conducted using aa fixed current of 120 range of A TP-MIG TP-MIG and welding experiment conducted current 120 A A and andtoaa the range of of frequencies, the results of thiswas experiment areusing shownfixed in Figure 10.ofAccording weld frequencies, and the results of this experiment are shown in Figure 10. According to the weld frequencies, and the results of this experiment are shown in Figure 10. According to the weld morphologies shown in this figure, a frequency of 2 Hz results in extremely clearly defined fish-scale morphologies shown in this aa frequency of in clearly defined morphologies shown this figure, figure, frequency of 22 Hz Hz results results in extremely extremely clearly defined fish-scale fish-scale patterns and large gapsinbetween the fish scales. Increasing the frequency to 5 Hz significantly improves patterns and large gaps between the fish scales. Increasing the frequency to 5 Hz significantly patterns large betweenofthe scales.patterns. Increasing the frequency to welding 5 Hz significantly the weld and quality andgaps the density the fish fish-scale A further increase in frequency improves the weld quality and the density of the fish-scale patterns. A further increase in improves the weld quality and the density of the fish-scale patterns. A further increase in welding welding to 15 Hz results in a corresponding increase in the density of the fish scales and the manifestation frequency to 15 Hz results in a corresponding increase in the density of the fish scales and frequency to 15features. Hz results in a corresponding increase in the density fish scales and the the of single-pulse In addition, it is found that the transitional pulsesofinthe TP-MIG help stabilize manifestation of single-pulse features. In addition, it is found that the transitional pulses in TP-MIG manifestation of single-pulse features. In addition, it is found that the transitional pulses in TP-MIG help help stabilize stabilize current current transitions transitions and and reduce reduce the the sharpness sharpness of of welding-arc welding-arc sounds. sounds. In In summary, summary, TP-MIG welding is highly stable and produces high-quality welds that are aesthetically pleasing. TP-MIG welding is highly stable and produces high-quality welds that are aesthetically pleasing.

A TP-MIG welding experiment was conducted using a fixed current of 120 A and a range of frequencies, and the results of this experiment are shown in Figure 10. According to the weld morphologies shown in this figure, a frequency of 2 Hz results in extremely clearly defined fish-scale patterns and large gaps between the fish scales. Increasing the frequency to 5 Hz significantly improves the weld quality and the density of the fish-scale patterns. A further increase in welding Appl. Sci. 2019, 9, 127 9 of 18 frequency to 15 Hz results in a corresponding increase in the density of the fish scales and the manifestation of single-pulse features. In addition, it is found that the transitional pulses in TP-MIG current transitions and transitions reduce the sharpness of the welding-arc sounds. In summary, TP-MIG is help stabilize current and reduce sharpness of welding-arc sounds. In welding summary, highly stable and produces high-quality welds that are aesthetically pleasing. TP-MIG welding is highly stable and produces high-quality welds that are aesthetically pleasing.

(a)

(b)

(c) Figure 10. 120 A waveform and weld formation at different frequencies: (a) 2 Hz; (b) 5 Hz; (c) 15 Hz.

4. Process Control Control Experiment Experiment on on Reduction 4. Process Reduction of of Heat Heat Input Input in in TP-MIG TP-MIG Welding Welding and and DP-MIG DP-MIG Welding Welding A comparative experiment was conducted using TP-MIG welding and DP-MIG welding, where A comparative experiment was conducted using TP-MIG welding and DP-MIG welding, where the input energy decreased by varying the number of pulses in the pulse train, the frequency of the input energy decreased by varying the number of pulses in the pulse train, the frequency of the the low-energy pulse train, and the welding speed. The effects of rectangular and trapezoidal low-energy pulse train, and the welding speed. The effects of rectangular and trapezoidal modulation on internal weld quality were analyzed by comparing the resulting weld morphologies modulation on internal weld quality were analyzed by comparing the resulting weld morphologies and metallurgical structures. Table 1 presents the chemical compositions of the base metals and and metallurgical structures. Table 1 presents the chemical compositions of the base metals and welding wires used in the MIG welding processes. welding wires used in the MIG welding processes. Table 1. Chemical composition of AA6061 material and ER4043 wire (mass fraction%). Material

Mg

Si

Fe

Cu

Mn

Ti

Al

AA6061 ER4043

0.80~1.2 0.05

0.4~0.8 5.60

0.7 0.80

0.15~0.40 0.30

0.15 0.05

0.15 –

Bal. Bal.

4.1. Reducing Heat Input by Varying Base Time by Adjusting the Number of Pulses in the Low-Energy Pulse Train The energy of the welding arc is transferred to the base metal, and part of this energy increases the temperature of the welding wire. The total energy of the welding arc is given by P = IU, and the effective welding power is Pλ = λIU. The effective power of these energy sources is in the range of 0.70–0.80; in this work, Pλ is considered to be 0.70. Based on the principles of DP-MIG welding and TP-MIG welding, it can be inferred that if current is to be reduced in these processes, the corresponding number of pulses should also be reduced. If the frequency is unchanged, decreasing the number of pulses will increase the base time. Based on the input energy equation, q = Pλ /v, decreasing the current will result in a proportionate decrease in the effective power, if the input frequency is fixed. A comparative experiment was conducted using DP-MIG welding and TP-MIG welding at a welding speed of 0.6 m/min and constant current. The other parameters in this experiment are listed in Tables 2 and 3. The appearances of the welds formed by TP-MIG welding and DP-MIG welding at each level of average current are shown in Figures 11 and 12, whereas Figures 13 and 14 are the corresponding waveform plots. Table 2. DP-MIG test data at different number of pulses in the low-energy pulse train. Ips (A)

tps (ms)

Ibs (A)

tbs (ms)

n (number)

Ipw (A)

tpw (ms)

Ibw (A)

tbw (ms)

m (number)

Average Current (A)

Average Voltage (V)

Input Energy (J)

210 210

2 2

72 72

6 6

16 16

210 210

2 2

48 48

15 26

8 5

88 80

22 21

135,520 117,600

frequency is fixed. frequency is fixed. A comparative experiment was conducted using DP-MIG welding and TP-MIG welding at a A comparative experiment was conducted using DP-MIG welding and TP-MIG welding at a welding speed of 0.6 m/min and constant current. The other parameters in this experiment are listed welding speed of 0.6 m/min and constant current. The other parameters in this experiment are listed in Tables 2 and 3. The appearances of the welds formed by TP-MIG welding and DP-MIG welding at in Tables 2 and 3. The appearances of the welds formed by TP-MIG welding and DP-MIG welding at each level of average current are shown in Figures 11 and 12, whereas Figures 13 and 14 are the eachSci. level current are shown in Figures 11 and 12, whereas Figures 13 and 14 are Appl. 2019,of9, average 127 10 ofthe 18 corresponding waveform plots. corresponding waveform plots. Table 2. DP-MIG test data at different different number of of pulses in in the low-energy low-energy pulsetrain. train. Table 2. 3. DP-MIG TP-MIG test Table data at different number number of pulses pulses in the the low-energy pulse pulse train. tps, Ips Ips , Ipsw tps Ips Ipw , Ipws (A) (ms) (A) (A) (ms) 210 210 2 210 2 210 2 210 210 2

tps Ibs, tpsw , tbs I tnbs Ibs, tpws tbs bs n tpw (A) (number) (ms) (A) (ms) (ms) (ms) (A) (number)

72 2 72 2 72 72

6 6 6 6

72 72

16 6 16 6 16 16

Ipw (A) (A) 15210 210 15210 210 n Ipw

tpw tbwIbw Ibw tpw Ibw (A) (ms) (ms) (A) (ms)

48 2 2 48 2

2

(A)

2548 48 5048

48

Average Average Input Average tbw m Average Average Average Input Input Current Voltage Voltage Energy Energy m nws Current (ms) (number) Current Voltage (J)Energy (V) (V) (A)(A) (J) (ms) (number) (A) (V) (J) 84 135,520 6 15 4 88 88 22 22 135,520 15 4 84 88 22 135,520 3 26 80 80 21 21 117,600 5 117,600 26 5 80 21 117,600

m tbw nsw

(a) (a)

(b) (b) Figure 11. 11. DP-MIG DP-MIG weld currents: (a) Average current Figure weld under under different different average average currents: currents: (a) (a) Average Average current current of of 88 88 A; A; (b) (b) Average Average current of of 80 80 A. A. current

(a) (a)

Appl. Sci. 2019, 9, x FOR PEER REVIEW

11 of 19

(b) (b)

300

300

250

250

200

200

Current (A)

Current (A)

Figure TP-MIG weldunder underdifferent different average average currents: current of 88 A; (b) Average Figure 12. 12. TP-MIG weld currents:(a) (a)Average Average current of 88 A; (b) Average current of 80 A. current of 80 A.

150 100

150 100 50

50 0

0

200

400

600

800

1000

0 0

Time (ms)

200

400

600

800

1000

Time (ms)

(a)

(b)

Figure DP-MIG currentwaveform waveformdiagram diagram of different in in thethe low-energy pulse Figure 13.13. DP-MIG current differentnumber numberofofpulses pulses low-energy pulse train: (a) Average current of 88 A; (b) Average current of 80 A. train: (a) Average current of 88 A; (b) Average current of 80 A. Table demonstrate 3. TP-MIG test data different number of pulses in the low-energy pulse train. Figures 11–14 thatatthe average welding current can be reduced by decreasing the number in, the low-energy pulse train if the parameters of the high-energy train Input are kept Ips, Ipswof , pulses tps, tpsw Average pulse Average Ibs tbs Ibw tbw pws Ipw, Ipws tpw, tinput Current Voltage The Energy n m nsw 14% nwsusing unchanged. The energy can be reduced by approximately this approach. welds (A) (ms) (A) (ms) (ms) (A) (J) in Figures 11 and 12 are normal. However, the fish scales of the TP-MIG(A) weld are(V)more uniform. 210 2 72 6 15 48 25 6 4 4 88 22 135,520 Micrographs of 2the welds by currents with 210 72 produced 6 15 different 48 50 3 in TP-MIG 4 4 welding 80 were compared 21 117,600

300

300

250

250

C

C

100

100 50

50

0

0

0

200

Appl. Sci. 2019, 9, 127

400

600

800

1000

0

200

Time (ms)

400

600

800

Time (ms)

(a)

1000

11 of 18

(b)

300

300

250

250

200

200

Current (A)

Current (A)

Figure 13. DP-MIG current waveform diagram of different number of pulses in the low-energy pulse those produced by DP-MIG welding, as shown in Figures 15 and 16. Grain growth is more severe in train: (a) Average current of 88 A; (b) Average current of 80 A. Figures 15a and 16a compared to Figures 15b and 16b, and DP-MIG welds have fewer second-phase particles. These results indicate that a high energy input induces grain growth and reduces the number Table 3. TP-MIG test data at different number of pulses in the low-energy pulse train. of second phase particles. Therefore, reducing the number of pulses in the low-energy pulse train Ips, Ipsw, tps, tpsw, Average Average Input Ibs tbs Ibw tabwcomparison of Figures 15 and 16 shows that the effectively reduces the input energy. Furthermore, Ipw, Ipws tpw, tpws Current Voltage Energy n m nsw nws (A) (ms) (A) (ms) TP-MIG in the presence (ms)finer grain sizes. This shows that the molten metals scour and (A) weld has (A) shear (V) (J) 210 2 72 weak 6 welding 15 48 which 25 prevents 6 4 stable 4 grain88growth.22Furthermore, 135,520 of alternating strong and arcs, 210 2 72 6 15 48 50 3 4 4 80 21 117,600 the fragmented grains induce the appearance of heterogeneous nucleation points, which is conducive to grain refinement.

150 100 50

150 100 50

Appl. Sci. 2019, 9, x FOR PEER REVIEW Appl. Sci. 2019, 9, x FOR PEER REVIEW 0 0

200

400

600

800

1000

12 of 19 12 of 19

0 0

200

400

600

800

1000

15 and and 16 16 shows shows that that the theTime TP-MIG weld has has finer finer grain grain sizes. sizes. This This shows shows that that the molten metals metals scour scour (ms) weld Timethe (ms) 15 TP-MIG molten and shear shear in in the the presence presence of alternating alternating strong strong and and weak weak welding welding arcs, arcs, which which prevents stable stable grain grain (a)of (b) and prevents growth. Furthermore, the fragmented grains induce the appearance of heterogeneous nucleation growth. Furthermore, the fragmented grains induce thenumber appearance ofinin heterogeneous nucleation Figure 14.14. TP-MIG waveform different number pulses the low-energy pulse Figure TP-MIGcurrent current waveform diagram diagram of different ofofpulses the low-energy pulse points, which which is is conducive conducive to to grain grain refinement. refinement. points, train: Averagecurrent currentofof88 88A; A;(b) (b)Average Average current current of 80 A. train: (a)(a) Average A. Figures 11–14 demonstrate that the average welding current can be reduced by decreasing the number of pulses in the low-energy pulse train if the parameters of the high-energy pulse train are kept unchanged. The input energy can be reduced by approximately 14% using this approach. The welds in Figures 11 and 12 are normal. However, the fish scales of the TP-MIG weld are more uniform. Micrographs of the welds produced by different currents in TP-MIG welding were compared with those produced by DP-MIG welding, as shown in Figures 15 and 16. Grain growth is more severe in Figures 15a and 16a compared to Figures 15b and 16b, and DP-MIG welds have fewer second-phase particles. These results indicate that a high energy input induces grain growth and reduces the number of second phase particles. Therefore, reducing the number of pulses in the low-energy pulse train effectively reduces the input energy. Furthermore, a comparison of Figures

(a) (a)

(b) (b)

Figure 15. 15. Micrograph Micrograph of of the the welds welds of of DP-MIG DP-MIG under under different different average average currents: currents: (a) (a) Average Average current current Figure of 88 88 A; A; (b) (b) Average Average current current of of 80 80 A. A. of Average current of 80 A.

(a) (a)

(b) (b)

Figure 16. Micrograph currents: (a)(a) Average current of Micrograph of of the thewelds weldsof ofTP-MIG TP-MIGunder underdifferent differentaverage average currents: Average current Figure 16. Micrograph of the welds of TP-MIG under different average currents: (a) Average current 88 A; (b) Average current of 80 A. of 88 88 A; A; (b) (b) Average Average current current of of 80 80 A. A. of

4.2. Reducing Reducing Heat Heat Input Input by by Adjusting Adjusting the the Frequency Frequency of of Low-Energy Low-Energy Pulse Pulse Train Train to to Vary Vary the the Base Base Time Time 4.2. If the the parameters parameters of of the the high-energy high-energy pulse pulse train train and and the the peak peak current current and and peak peak times times of of the the If low-energy pulse train are fixed, the heat input can be reduced by increasing the duration of the base low-energy pulse train are fixed, the heat input can be reduced by increasing the duration of the base

Appl. Sci. 2019, 9, 127

12 of 18

4.2. Reducing Heat Input by Adjusting the Frequency of Low-Energy Pulse Train to Vary the Base Time If the parameters of the high-energy pulse train and the peak current and peak times of the low-energy pulse train are fixed, the heat input can be reduced by increasing the duration of the base current in the low-energy pulse train because this effectively reduces the average current. This approach is known as variation in the base time. Two DP-MIG welding experiments were conducted at a welding speed of 0.6 m/min, and the parameters are listed in Table 4. The appearance and morphology of the welds formed in these experiments and their corresponding current waveform plots are illustrated in Figure 17. Appl. Sci. 2019, 9, x FOR PEER REVIEW

13 of 19

Table 4. Comparison of DP-MIG test data of low-energy pulse group at different frequencies. Table 4. Comparison of DP-MIG test data of low-energy pulse group at different frequencies. IpsIps (A) (A)

tpstps (ms) (ms)

IIbsbs (A) (A)

ttbs bs (ms) (ms)

n n

Ipw Ipw (A) (A)

210210 210210

22 22

77 77 77 77

66 66

16 16 16 16

210 210 210 210

Average Average Average Average tpwtpw Ibw Ibw tbw tbw Voltage m Current Current Voltage m (ms) (A) (ms) (ms) (A) (ms) (V) (A) (A) (V) 2 2 50 50 13 13 9 9 90 90 22 22 2 2 50 50 22 22 9 9 80 80 22 22

Input Input Energy Energy (J) (J) 138,600 138,600 123,200 123,200

300

Current (A)

250 200 150 100 50 0 0

100

200

300

400

500

400

500

Time (ms)

(a) 300

Current (A)

250 200 150 100 50 0 0

100

200

300

Time (ms)

(b) Figure of of 90 90 A, A, frequency of 3.8 Hz;Hz; (b) Figure 17. 17. DP-MIGW DP-MIGW atatdifferent differentfrequencies: frequencies:(a) (a)Average Averagecurrent current frequency of 3.8 Average current of 80ofA,80frequency of 3 of Hz.3 Hz. (b) Average current A, frequency

Two TP-MIG welding of experiments were also conducted pulse at thegroup sameatwelding (0.6 m/min), Table 5. Comparison TP-MIGW test data of low-energy differentspeed frequencies. and the parameters are listed in Table 5. The superficial morphologies of the welds formed in these Ips, Ipsw, tps, tpsw, Average Average Input Ibs tbs Ibw tbw are shown in Figure 18. experiments and their corresponding current waveforms Ipw, Ipws tpw, tpws Current Voltage Energy n m nsw nws (ms) 2 2

(A)

(ms)

77 77

6 6

15 15

(A)

(ms)

50 50

20 30 300 250

rrent (A)

(A) 210 210

200 150

7 7

4 4

4 4

(A) 90 80

(V) 22 22

(J) 138,600 123,200

0 0

100

200

300

400

500

Time (ms)

(b) Figure DP-MIGW at different frequencies: (a) Average current of 90 A, frequency of 3.8 Hz; Appl. Sci. 2019, 17. 9, 127 13 of 18 (b) Average current of 80 A, frequency of 3 Hz. Table 5. Table 5. Comparison of of TP-MIGW TP-MIGW test test data data of of low-energy low-energy pulse pulse group group at at different differentfrequencies. frequencies. IpsI,psI,psw , , Ipsw Ipw Ipws , I,pws Ipw (A) (A) 210 210 210 210

tpsw, , tpstps , t, psw tpw , t,pws tpws tpw (ms) (ms) 22 22

Ibs Ibs tbs tbs n (A) (A) (ms) (ms)

Ibw n (A)

Ibwtbw tbwm (ms) (ms) (A)

77 77 77 77

15 50 15 50

50 20 50 30

6 6

6 15 6 15

207 307

AverageAverage AverageInputInput Average CurrentVoltage VoltageEnergy Energy mnsw nsw nws nws Current (J) (J) (V) (V) (A) (A) 90 90 22 22 138,600 7 4 4 4 4 138,600 80 80 22 22 123,200 7 4 4 4 4 123,200

300 250

Current (A)

200 150 100 50 0 0

Appl. Sci. 2019, 9, x FOR PEER REVIEW

100

200

300

400

Time (ms)

500

14 of 19

(a) 300

Current (A)

250 200 150 100 50 0 0

100

200

300

400

500

Time (ms)

(b) Figure TP-MIGWatatdifferent differentfrequencies; frequencies;(a)(a) Average current frequency 3 Hz; Figure 18. 18. TP-MIGW Average current of of 90 90 A, A, frequency of 3ofHz; (b) (b) Average current of 80 frequency Average current of 80 A, A, frequency of of 2.42.4 Hz.Hz.

This thethe input energy by approximately 12%. The average frequency Thisapproach approachreduces reduces input energy by approximately 12%. The current averageand current and decrease as the base time of the low-energy pulse train increases, and the fish scales formed at the frequency decrease as the base time of the low-energy pulse train increases, and the fish scales welded formed joint at thebecome weldeddenser. joint become denser. 4.3. 4.3. Effects Effects of of Welding Welding Speed Speed on on Input Input Energy Energy The welding arc generates heat, andand a heat-affected zone zone is created close toclose the joint. Anjoint. increase The welding arc generates heat, a heat-affected is created to the An in the welding speed leads to a proportionate decrease in heat transfer and a significant decrease in increase in the welding speed leads to a proportionate decrease in heat transfer and a significant input energy if the effective power is unchanged. This concentrates the energy of the arc and reduces decrease in input energy if the effective power is unchanged. This concentrates the energy of the arc the energy into the weld pool, which reduces heat reduces input to heat the sides andtotal reduces thefed total energy fed into the weldfurther pool, which further inputoftothe thewelding sides of track. As a result, in the base metal becomes narrow. Hence,narrow. the welding speed the welding track.the Asheat-affected a result, the zone heat-affected zone in the base metal becomes Hence, the negatively correlates with the input heat. It should be noted that the heat-affected zone significantly welding speed negatively correlates with the input heat. It should be noted that the heat-affected affects the mechanical properties of the joint, and a large heat-affected lead to joint zone significantly affects the mechanical properties of the joint, and a zone largecould heat-affected zonefailure could or fractures. heat-affected zones were analyzed byzones carrying out hardnessby tests on the out welded joint. lead to jointThe failure or fractures. The heat-affected were analyzed carrying hardness The results of these tests are shown in Figure 19. tests on the welded joint. The results of these tests are shown in Figure 19.

DP-MIG welding and TP-MIG welding were performed at a constant current of 88 A, and the welding speed was gradually increased from 0.6 m/min to 1.0 m/min. The resulting joints are shown in Figures 20 and 21. The Vickers hardnesses of the welded joints formed at 0.6 m/min and 1.0 m/min were determined with a load of 0.98 N and loading time of 15 s. Hardness tests were first conducted on the base metal to perform detailed analysis of changes in the welded joint hardness; it was found that the hardness of the base metal was 38 HV. The hardness of each zone of the weld is shown in

the welding track. As a result, the heat-affected zone in the base metal becomes narrow. Hence, the the welding track. As a result, the heat-affected zoneheat. in the metal becomes narrow. Hence, the welding speed negatively correlates with the input It base should be noted that the heat-affected welding speed negatively correlates with the input heat. It should be noted that the heat-affected zone significantly affects the mechanical properties of the joint, and a large heat-affected zone could zoneto significantly the mechanical properties zones of the joint, a largebyheat-affected could lead joint failureaffects or fractures. The heat-affected were and analyzed carrying outzone hardness lead on to the jointwelded failurejoint. or fractures. Theofheat-affected were analyzed tests The results these tests arezones shown in Figure 19. by carrying out hardness Appl. Sci. 2019, 9, 127 14 of 18 testsDP-MIG on the welded joint. resultswelding of thesewere tests performed are shown in 19. current of 88 A, and welding andThe TP-MIG at Figure a constant the DP-MIG and TP-MIG welding were performed at a constant current joints of 88 are A, and the welding speedwelding was gradually increased from 0.6 m/min to 1.0 m/min. The resulting shown welding speed was gradually increased from 0.6 m/min to 1.0 m/min. The resulting joints are shown in Figures 20 and 21. The Vickers hardnesses of the welded joints formed at 0.6 m/min and 1.0 m/min DP-MIG welding and TP-MIG welding were performed at a constant current of 88 A, and the in Figures 20 and 21. The Vickers hardnesses of the welded joints formed 0.6 m/min andconducted 1.0shown m/min were determined a load of 0.98 N and timetoof1.0 15 s. Hardness werejoints first welding speed waswith gradually increased fromloading 0.6 m/min m/min. Theattests resulting are were with loaddetailed of 0.98 Nanalysis and loading time of 15 Hardness tests were and first conducted on thedetermined base20metal to perform of changes in thes.formed welded joint hardness; it 1.0 was found in Figures and 21. Thea Vickers hardnesses of the welded joints at 0.6 m/min m/min on the base metal to perform detailed analysis of changes in the welded joint hardness; it was found that the hardnesswith of the baseofmetal HV. The hardness of each zone the first weldconducted is shown in were determined a load 0.98 Nwas and38 loading time of 15 s. Hardness testsof were on thatbase the22. hardness of the base metal was 38ofHV. The hardness of each zone of the it weld shown in Figure the metal to perform detailed analysis changes in the welded joint hardness; wasisfound that Figure 22. the hardness of the base metal was 38 HV. The hardness of each zone of the weld is shown in Figure 22.

Figure 19. 19. Sample Sample extraction extraction of of welded welded joint. joint. Figure Figure 19. Sample extraction of welded joint.

Appl. 2019, x FOR PEERREVIEW REVIEW Appl. Sci.Sci. 2019, 9, x9,FOR PEER

15 of 1519 of 19

Figure m/min. Figure20. 20.DP-MIGW DP-MIGWwelding weldingjoint jointatat1.0 1.0m/min. Figure 20. DP-MIGW welding joint at 1.0m/min.

Figure 21. TP-MIGW welding joint at 1.0m/min. Figure m/min. Figure21. 21. TP-MIGW TP-MIGWwelding weldingjoint jointat at1.0 1.0m/min. 65

70

60

65

70

0.6 m/min 0.6 1.0m/min m/min 1.0 m/min

Microhardness (HV 0.2) Microhardness (HV 0.2)

60 55

55

Microhardness (HV Microhardness (HV 0.2)0.2)

65

50

50 45 40 35 30

45 40 35 30 -10

-10

-8

-6

-4

-2

0

2

4

6

8

(mm)8 -8 Distance -6 -4 from -2 the 0 weld 2 center 4 6

Distance from (a) the weld center (mm)

10

10

65

0.6 m/min m/min 1.00.6 m/min

60

60

1.0 m/min

55

55 50 50 45 45

40

40

35

35 30 30

-10

-10

-8

-6

-4

-2

0

2

4

6

8

Distance the weld center -8 -6 from -4 -2 0 2 4 (mm) 6

10

8

10

(b)the weld center (mm) Distance from

(a) (b)of DP-MIGW Figure 22.22. Hardness distribution of welded jointsjoints at two (a) Comparison hardness Figure Hardness distribution of welded atspeeds: two speeds: (a) Comparison of DP-MIGW hardness of welded joints; (b) Comparison of TP-MIGW hardness of welded joints. of welded joints; (b) Comparison of TP-MIGW hardness of welded joints. Figure 22. Hardness distribution of welded joints at two speeds: (a) Comparison of DP-MIGW hardness of welded joints; (b) Comparison of TP-MIGW hardness of welded joints.

The hardness related to to that thatof ofthe thewelding weldingwire. wire.Figure Figure The hardnessofofthe theweld weldatatits itscenter center is is closely closely related 23 23 indicates that welded joints formed at 1.0 are generally stronger thanthan jointsjoints formed at 0.6at m/min. indicates that welded joints formed at m/min 1.0 m/min are generally stronger formed 0.6 The hardness of the weld at its center is closely related to that of the welding wire. Figure 23 In m/min. DP-MIGInwelding, increasean inincrease the welding speed increases hardness 19% andby decreases DP-MIGan welding, in the welding speedthe increases thebyhardness 19% andthe indicates that welded joints formed at 1.0 m/min are generally stronger than joints formed at 0.6 decreases the area thatthe is softer than the base metal (i.e., softened zone) by 21%. In TP-MIG an welding, area that is softer than base metal (i.e., softened zone) by 21%. In TP-MIG welding, increase m/min. In DP-MIG welding, an increase in the welding speed increases the hardness by 19% and the welding speed increases the hardness byofan average of 17% and in an theincrease weldinginspeed increases the hardness by an average 17% and decreases thedecreases softenedthe zone decreases the area that is softer than the base metal (i.e., softened zone) by 21%. In TP-MIG welding, softened zone by 27%. The change in hardness with welding speed is shown in Figure 23. The figure by 27%. The change in hardness with welding speed is shown in Figure 23. The figure shows that an an increase in the welding speed increases the hardness by an average of 17% and decreases the shows that an appropriate in welding speed helps control energy input and in thin-plate appropriate increase in weldingincrease speed helps control energy input in thin-plate welding is beneficial softened zone by 27%. The change in hardness with welding speed is shown in Figure 23. figure and is spread beneficial for reducing the spread heat-affected Byofcomparing theThe results forwelding reducing the of the heat-affected zone. of Bythe comparing the zone. results DP-MIG welding and shows that an appropriate increase in welding speed that helps control energy resulted input ininthin-plate of DP-MIG welding and TP-MIG welding, it was found the latter approach harder TP-MIG welding, it was found that the latter approach resulted in harder joints. In addition, it was welding and is beneficial reducing theinput spread of the heat-affected zone. comparingHowever, the results joints.that In addition, it energy wasfor shown the can be reduced in bothBy approaches. shown the input can that be reduced inenergy both approaches. However, the heat-affected zone of the DP-MIG welding andinTP-MIG it was found that the approach resulted in harder heat-affected zone TP-MIG welding, welding was slightly smaller, andlatter the decrease in input energy due in TP-MIG welding was slightly smaller, and the decrease in input energy due to the increase in the joints. Inincrease addition, was shownspeed that was the input can be in both approaches. However, to the in it the welding more energy pronounced in reduced TP-MIG welding. welding speed was more pronounced in TP-MIG welding. the heat-affected zone in TP-MIG welding was slightly smaller, and the decrease in input energy due to the increase in the welding speed was more pronounced in TP-MIG welding.

shows that an appropriate increase in welding speed helps control energy input in thin-plate welding and is beneficial for reducing the spread of the heat-affected zone. By comparing the results of DP-MIG welding and TP-MIG welding, it was found that the latter approach resulted in harder joints. In addition, it was shown that the input energy can be reduced in both approaches. However, the heat-affected Appl. Sci. 2019, 9, 127 zone in TP-MIG welding was slightly smaller, and the decrease in input energy due 15 of 18 to the increase in the welding speed was more pronounced in TP-MIG welding.

(a)

(b)

Figure Difference hardness of welded joints at two speeds: DP-MIGW hardness of Figure 23.23. Difference in in thethe hardness of welded joints at two speeds: (a) (a) DP-MIGW hardness of welded welded joint; (b) DP-MIGW of welded joint; (b) DP-MIGW hardnesshardness of welded joint. joint. Appl. Sci. 2019, 9, x FOR PEER REVIEW

16 of 19

Figure 24 shows the microstructure of welded joints welded by DP-MIG and TP-MIG at welding 24 shows microstructure of weldedthat joints by DP-MIG and TP-MIG at the speeds of Figure 0.6 m/min and the 1.0 m/min. It is observed thewelded grain size of the fusion zone and welding speeds of 0.6 m/min and 1.0 m/min. It is observed that the grain size of the fusion zone and heat-affected zone at the welding speed of 1.0 m/min is less coarsening than that at the welding theof heat-affected the welding speed of 1.0 m/min is less coarsening than that at the welding speed 0.6 m/min.zone Theatreason for this phenomenon may be that the welded joints obtained at the speed of 0.6 m/min. The reason for this phenomenon may be that the welded joints obtained at the welding speed of 1.0 m/min have less heat input. Besides, welded joints of TP-MIG have less pores welding speed of 1.0 m/min have less heat input. Besides, welded joints of TP-MIG have less pores than that of DP-MIG, which may due to the effect of high- to low-energy transition pulse groups and than that of DP-MIG, which may due to the effect of high- to low-energy transition pulse groups and low-energy transition pulse groups in TP-MIG welding. High- to low-energy transition pulse groups low-energy transition pulse groups in TP-MIG welding. High- to low-energy transition pulse groups andand low-energy transition pulse groups can make the arc state more stable; therefore, less air is trapped low-energy transition pulse groups can make the arc state more stable; therefore, less air is intotrapped the molten pool. into the molten pool.

(a)

(b)

(c)

(d)

Figure Microstructurephotograph photograph of of welded welded joints: 0.60.6 m/min, (b) DP-MIGW, Figure 24. 24. Microstructure joints:(a) (a)DP-MIGW, DP-MIGW, m/min, (b) DP-MIGW, 1.0 m/min, TP-MIGW,0.6 0.6m/min, m/min, (d) 1.0 m/min, (c)(c) TP-MIGW, (d)TP-MIGW, TP-MIGW,1.0 1.0m/min. m/min. 180

e (N)

4000 3000

ength (MPa)

5000

160 140 120

154

165 148

168

Appl. Sci. 2019, 9, 127

16 of 18

Figure 25 shows the tensile curves and the ultimate tensile strength of the welded joints. The ultimate tensile strength of DP-MIG joints obtained at the welding speeds of 0.6 m/min and 1.0 m/min is 154 MPa and 148 MPa, respectively. The ultimate tensile strength of TP-MIG joints (c) (d) obtained at the welding speeds of 0.6 m/min and 1.0 m/min is 165 MPa and 168 MPa, respectively, Figure 24. Microstructure photograph of welded joints: (a) DP-MIGW, 0.6 m/min, (b) DP-MIGW, which is better than the tensile results of DP-MIG joints. 1.0 m/min, (c) TP-MIGW, 0.6 m/min, (d) TP-MIGW, 1.0 m/min. 180

Ultimate tensile strength (MPa)

5000

Tensile force (N)

4000 3000 2000

DP-MIGW DP-MIGW TP-MIGW TP-MIGW

1000 0 0.0

0.5

1.0

1.5

2.0

Appl. Sci. 2019, 9, x FOR PEER REVIEW

Distance (mm)

2.5

0.6m/min 1.0m/min 0.6m/min 1.6m/min

3.0

3.5

160

154

165

168

148

140 120 100 80 60 40 20 0 DP-MIGW 0.6 m/min

DP-MIGW 1.0 m/min

TP-MIGW 0.6 m/min

TP-MIGW 17 of 19 1.0 m/min

(a) tensile curves and the ultimate tensile strength(b) Figure 25 shows the of the welded joints. The ultimateFigure tensile strength of DP-MIG joints obtained at the welding speeds of 0.6 strength. m/min and 1.0 25. Tensile properties of weldedjoints: joints: (a) curves; (b) Ultimate tensile Figure 25. Tensile properties of welded (a)Tensile Tensile curves; (b) Ultimate tensile strength. m/min is 154 MPa and 148 MPa, respectively. The ultimate tensile strength of TP-MIG joints obtained at the welding speeds of 0.6 m/min and 1.0 m/min is 165 MPa and 168 MPa, respectively, Figure 26 shows photographs of fractured tensile samples welded by DP-MIG and TP-MIG at which is better than the tensile results of DP-MIG joints. the welding Figure speeds 0.6 photographs m/min andof1.0 m/min. Allsamples tensilewelded fractures revealand that there atare many 26 of shows fractured tensile by DP-MIG TP-MIG cleavagethe surfaces tearing fibersAlland steps around dimples, which that weldingformed speeds consisting of 0.6 m/minofand 1.0 m/min. tensile fractures reveal that there are indicates many cleavage surfaces consisting ofmixed tearing fibers and steps around dimples, which indicates that tensile the fracture belongs toformed a plastic-tough fracture mode. In addition, in the fractured fracture belongs a plastic-tough mixedseen, fracture mode.may In addition, in the fractured tensile samples the of DP-MIG, sometopores can be clearly which be the important reason for the low samples of DP-MIG, some pores can be clearly seen, which may be the important reason for the low tensile properties of DP-MIG joints. tensile properties of DP-MIG joints.

(a)

(b)

(c)

(d)

Figure 26. Photographs of fractured tensile tensile samples: (a) DP-MIGW, 0.6 m/min, DP-MIGW, 1.0 Figure 26. Photographs of fractured samples: (a) DP-MIGW, 0.6 (b) m/min, (b) DP-MIGW, m/min, (c) TP-MIGW, 0.6 m/min, (d) TP-MIGW, 1.0 m/min. 1.0 m/min, (c) TP-MIGW, 0.6 m/min, (d) TP-MIGW, 1.0 m/min.

5. Conclusions In this work, a comparative analysis was performed on the process stability and heat input controllability of TP-MIG welding and DP-MIG welding. The following conclusions are drawn from the experimental results: 1.

Owing to abrupt changes in current induced by pulse train transitions in DP-MIG welding,

Appl. Sci. 2019, 9, 127

17 of 18

5. Conclusions In this work, a comparative analysis was performed on the process stability and heat input controllability of TP-MIG welding and DP-MIG welding. The following conclusions are drawn from the experimental results: 1.

2.

3.

Owing to abrupt changes in current induced by pulse train transitions in DP-MIG welding, the welding arc makes a sharp sound. In addition, the probabilities of welder burn-back and the formation of screw-like wires during long welding operations are higher in DP-MIG welding than in TP-MIG welding. In TP-MIG welding, the presence of transitional pulses in the pulse train transitions helps stabilize the welding process and soften the sound of the welding arc. The weld formed by TP-MIG welding has a high level of aesthetic quality. The input energy can be reduced by adjusting the number of pulses in the low-energy pulse train or the base time of the pulses. It was found that varying the number of pulses can reduce the current by 8 A and the input energy by approximately 14%. The welds formed by TP-MIG welding have finer grain sizes than those formed by DP-MIG welding. Under the same conditions, the input energy can be reduced by 12% by adjusting the base time of the pulses. In addition, the density of the fish scales increases as the frequency decreases. The energies produced by TP-MIG welding and DP-MIG welding decrease as the welding speed increases. An increase in the welding speed from 0.6 m/min to 1.0 m/min decreases the area of the softened zone by 21% and 27% in DP-MIG welding and TP-MIG welding, respectively. In addition, the joint formed by TP-MIG welding is harder than the joint formed by DP-MIG welding. Additionally, the joint formed by TP-MIG welding has better tensile properties than the joint formed by DP-MIG welding. Therefore, TP-MIG welding is a superior technique in terms of its efficacy in reducing welding energy input.

Author Contributions: M.X. and X.J. jointly conceived and designed the experiment. M.X., Z.Z. and J.L. jointly performed the experiment and conducted data analysis. M.X. and W.W. analyzed the data and plotted the figures. M.X. and X.J. wrote this paper. Z.Z. and W.W. proofread and translated the paper. X.J. revised the paper, supervised research, and provided financial support. Funding: This paper was supported by the National Natural Science Foundation project of China (51875213) and the High-Level Leading Talent Introduction Program of Guangdong academy of sciences [grant numbers 2016-GDASRC-0106]; Natural Science Foundation of Fujian [grant numbers 2018J01503]. Acknowledgments: In this section, you can acknowledge any support given which is not covered by the author contribution or funding sections. This may include administrative and technical support, or donations in kind (e.g., materials used for experiments). Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3. 4. 5. 6. 7.

Wang, L.; Wu, C.S.; Gao, J.Q. Suppression of humping bead in high speed GMAW with external magnetic field. Sci. Technol. Weld. Join. 2016, 21, 139–141. [CrossRef] Li, Y.; Wu, C.S.; Wang, L.; Gao, J.K. Analysis of additional electromagnetic force for mitigating the humping bead in high-speed gas metal arc welding. J. Mater. Process. Technol. 2015, 229, 207–215. [CrossRef] Choi, H.W.; Farson, D.F.; Cho, M.H. Using a Hybrid Laser Plus GMAW Process for Controlling the Bead Humping Defect. Weld. J. 2006, 174-s–179-s. [CrossRef] Rosado, T.; Almeida, P.; Pires, I.; Miranda, R.; Quintino, L. Innovations in arc welding. In Proceedings of the 5th Congresso Luso-Moc, ambicano de Engenharia, Maputo, Mozambique, 2–4 September 2008. Liu, A.; Tang, X.; Lu, F. Arc profile characteristics of Al alloy in double-pulsed GMAW. Int. J. Adv. Manuf. Technol. 2013, 65, 1–7. [CrossRef] Gery, D.; Long, H.; Maropoulos, P. Effects of welding speed, energy input and heat source distribution on temperature variations in butt joint welding. J. Mater. Process. Technol. 2005, 167, 393–401. [CrossRef] Hang, Z.; Wu, D.; Zou, Y. Effect of bypass coupling on droplet transfer in twin-wire indirect arc welding. J. Mater. Process. Technol. 2018, 262, 123–130.

Appl. Sci. 2019, 9, 127

8.

9.

10. 11. 12.

13.

14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

18 of 18

Wen, C.; Wang, Z.; Deng, X.; Wang, G.; Misra, R.D.K. Effect of heat input on the microstructure and mechanical properties of low alloy ultra-high strength structural steel welded joint. Steel Res. Int. 2018, 89, 1700500. [CrossRef] Reis, R.P.; Scotti, A.; Norrish, J.; Cuiuri, D. Investigation on welding arc interruptions in the presence of magnetic fields: Arc length, torch angle and current pulsing frequency influence. IEEE Tran. Plasma Sci. 2013, 41, 133–139. [CrossRef] Armentani, E.; Esposito, R.; Sepe, R. The influence of thermal properties and preheating on residual stresses in welding. Int. J. Comput. Mater. Sci. Surf. Eng. 2007, 1, 146–162. [CrossRef] Hazvinloo, H.R.; Honarbakhsh Raouf, A. Effect of gas-shielded flux cored arc welding parameters on weld width and tensile properties of weld metal in a low carbon steel. J. Appl. Sci. 2010, 10, 658–663. Armentani, E.; Pozzi, A.; Sepe, R. Finite-element simulation of temperature fields and residual stresses in butt welded joints and comparison with experimental measurements. In Proceedings of the ASME 2014 12th Biennial Conference on Engineering Systems Design and Analysis, ESDA 2014, Copenhagen, Denmark, 25–27 June 2014. Somashekara, M.A.; Naveenkumar, M.; Kumar, A.; Viswanath, C.; Simhambhatla, S. Investigations into effect of weld-deposition pattern on residual stress evolution for metallic additive manufacturing. Int. J. Adv. Manuf. Technol. 2017, 90, 2009–2025. [CrossRef] Wang, L.L.; Jin, L.; Huang, W.J.; Xue, J. Effect of Thermal Frequency on AA6061 Aluminum Alloy Double Pulsed Gas Metal Arc Welding. Mater. Manuf. Process. 2015, 31, 2152–2157. [CrossRef] Pépe, N.; Egerland, S.; Colegrove, P.A.; Yapp, D.; Leonhartsberger, A.; Scotti, A. Measuring the process efficiency of controlled gas metal arc welding processes. Sci. Technol. Weld. Join. 2013, 16, 412–417. [CrossRef] Kah, P.; Suoranta, R.; Martikainen, J. Advanced gas metal arc welding processes. Int. J. Adv. Manuf. Technol. 2013, 67, 655–674. [CrossRef] Cruz, J.G.; Torres, E.M.; Alfaro, S.C.A. A methodology for modeling and control of weld bead width in the GMAW process. J. Braz. Soc. Mech. Sci. 2015, 37, 1529–1541. [CrossRef] Jin, L.; Xue, J.X.; Hu, Y.; Zhang, Z. Effects of Thermal Frequency on Microstructures, Mechanical and Corrosion Properties of AA6061 Joints. Appl. Sci. (Basel) 2018, 8, 540. [CrossRef] Mvola, B. Adaptive Gas Metal Arc Welding Control and Optimization of Welding Parameters Output: Influence on Welded Joints. Int. Rev. Mech. Eng. 2016, 10, 67–72. [CrossRef] Yi, L.; Yan, Z.; Xie, X.; Yang, Z. Effect of welding heat input to metal droplet transfer characterized by structure-borne acoustic emission signals detected in GMAW. Measurement 2015, 70, 75–82. [CrossRef] Narendra, K.S.; Annaswamy, A.M. Robust Adaptive Control; Springer: London, UK, 2015. Miao, B.; Li, T. A novel neural network-based adaptive control for a class of uncertain nonlinear systems in strict-feedback form. Nonlinear Dyn. 2015, 79, 1005–1013. [CrossRef] Huang, S.; Gu, X.; Jiao, X.; Zhang, Y. Effects of pulse current on droplet transfer in hyperbaric pulsed MIG welding. Trans. China Weld. Inst. 2015, 36, 25–29. © 2019 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/).