materials

Article A Primary Study of Variable Polarity Arc Using a Pulsed Plasma Gas

Zhenyang Lu 1, Wang Zhang 1, Fan Jiang 1,2,*, Shujun Chen 1 and Zhaoyang Yan 1,3

1 Engineering Research Center of Advanced Manufacturing Technology for Automotive Components, Ministry of Education, Beijing University of Technology, Beijing 100124, China; [email protected] (Z.L.); [email protected] (W.Z.); [email protected] (S.C.); [email protected] (Z.Y.) 2 Beijing Engineering Researching Center of laser Technology, Beijing University of Technology, Beijing 100124, China 3 Chemical and Materials Engineering, University of Alberta, Edmonton, AB T6G 2R3, Canada * Correspondence: [email protected]; Tel.: +86-137-2008-7645

 Received: 10 April 2019; Accepted: 20 May 2019; Published: 22 May 2019 

Abstract: A process variant of variable polarity plasma (VPPAW), that is, the pulsed plasma gas VPPAW process, was developed. The pulsed plasma gas was transmitted into the variable polarity plasma arc through a high-frequency solenoid valve to modify the output of the plasma arc. The collection of arc electrical characteristics, arc shapes, and weld formation from VPPAW, double-pulsed VPPAW (DP-VPPAW), and pulsed plasma gas VPPAW (PPG-VPPAW) was carried out to examine if the pulsed plasma gas was able to play a positive role in improving the stability and quality of the VPPAW process. The arc voltage shows that the pulsed plasma gas had a greater influence on the positive polarity voltage. The lower the plasma gas frequency was, the lower the arc voltage fluctuation frequency was and the greater the arc voltage fluctuation amplitude was. From the arc image, it could be observed that the arc core length had a short decrease during the general rising trend after plasma gas was turned on. The arc core width only had a slight change due to the restriction of the torch orifice. Compared with pulsed current wave, the pulsed plasma gas could better enhance the fluidity of the molten pool to reduce porosity during aluminum keyhole welding.

Keywords: variable polarity plasma arc welding (VPPAW); weld formation; pulsed plasma gas; arc voltage

1. Introduction Variable polarity plasma arc welding (VPPAW) has been widely used in aeronautics, astronautics, and the automobile industry to produce high-quality and high-precision weld joints of aluminum alloy [1,2]. The constrained process in the plasma arc torch leads to high energy density and arc stiffness of the plasma arc, but makes the output of the arc coupled. This characteristic results in a smaller weld lobe curve than for other processes, and leads the keyhole molten pool easily affected by environmental changes in complex welding environments [3]. All of these cause a poor dynamic stability of the keyhole and weld defects, which restrict the application of VPPAW. Increasing arc energy density is a common method to improve welding quality. The laser hybrid plasma arc welding process [4,5] can improve the energy density and stability of the plasma arc by benefiting from interactions with laser beams. The gas-focusing plasma arc welding process [6], with arc column secondly constricted by focusing gas, can improve the arc restraint degree and stability significantly. The increase in arc stability effectively reduces the disturbance from welding arc to molten pool, but it is difficult to eliminate the other influence on molten pool stability.

Materials 2019, 12, 1666; doi:10.3390/ma12101666 www.mdpi.com/journal/materials Materials 2019, 12, 1666 2 of 14

Increasing the robustness of the keyhole molten pool is more helpful for the interference resisting. The controlled pulsed plasma arc welding process [7,8], which makes the holes in periodic opening and closing state by adjusting the current output, could effectively improve the robustness of the molten pool. The double-pulsed VPPAW (DP-VPPAW) [9,10] adds additional high-frequency pulsed current into electrode negative (EN) and electrode positive (EP), which makes the molten pool oscillate periodically to improve stability of the welding process and quality. However, the additional current parameters lead the thermal-force synchronous change and cause difficulties of process control. The vibration-assisted plasma arc welding process [11,12], which uses mechanical vibration to drive molten pool vibration, is an efficient method to decrease the attendant heat fluctuation of the plasma arc, compared with the current wave. These methods produced satisfactory process effectiveness, but the mechanical coupling of vibration system and welding torch reduced the stability and precision of the plasma welding torch structure. The previous studies [13,14] have shown that the plasma gas flow rate significantly affects the pressure output, but for the heat output, it is insignificant. Based on this, a novel welding process named pulsed plasma gas VPPAW (PPG-VPPAW) was proposed in this study. A specially designed plasma arc torch was used to control plasma gas flow, and the pulsed plasma gas VPPAW system was developed. This paper focuses on the arc behavior and welding process with PPG-VPPAW. The reason for the periodic variation of arc voltage and arc profile during PPG-VPPAW is discussed based on the arc electric signals and arc image acquired from experimental results. Furthermore, the weld- experiments were carried out to explore the reason for the improving the fluidity of the molten pool and reducing porosity during aluminum keyhole PPG-VPPAW.

2. Experimental Procedure

2.1. Experimental System Figure1 shows the schematic diagram of the PPG-VPPAW system, which included three parts: the pulsed plasma gas control unit, the VPPAW system, and the data acquisition system. The plasma gas control unit consisted of a S7-200 series PLC and a 35A series high-frequency solenoid valve (rated voltage of DC24V, rated power of 5.4 W, conduction response time of 6 ms, outages response time of 2 ms, and the maximum atmospheric flux of 16.2 L/min for argon). The solenoid valve was installed in the plasma gas tube of the plasma arc torch and as close to the torch nozzle as possible to generate the pulsed plasma gas flow. The VPPAW system consisted of a welding power source, a modified plasma arc torch, and the needed accessories. The data acquisition system mainly consisted of a voltage sensor, a current sensor, a high-speed camera, a data acquisition card, and an industrial computer. During the welding process, the welding current and voltage were collected by the current and voltage sensors. The plasma arc was imaged by the high-speed camera (IDT Y4 series), which was set to focus on the fixed region around the nozzle of the unmovable plasma arc torch. The experimental data was displayed and recorded by the data acquisition card in the industrial control computer. MaterialsMaterials 20192019, 12, 12, x, FOR 1666 PEER REVIEW 3 of3 of14 14 Materials 2019, 12, x FOR PEER REVIEW 3 of 14

Figure 1. Schematic diagram of the PPG-VPPAW system. Figure 1. SchematicSchematic diagram diagram of of the the PPG-VPPAW PPG-VPPAW system. 2.2. Experimental Design 2.2. Experimental Experimental Design The principle of the PPG-VPPAW process is shown in Figure 2. When the solenoid valve was The principle of the PPG-VPPAW process is shown in Figure2. When the solenoid valve was closed,The the principleplasma gas of theflow PPG-VPPAW was blocked processby the solenoid is shown valve in Figure and gathered2. When atthe the solenoid entrance valve of the was closed, the plasma gas flow was blocked by the solenoid valve and gathered at the entrance of the solenoidclosed, valve.the plasma When gas the flow solenoid was valveblocked was by opened, the solenoid the blocked valve andplasma gathered gas flow at thewas entrance released ofand the solenoid valve. When the solenoid valve was opened, the blocked plasma gas flow was released and inputsolenoid into thevalve. plasma When arc the at solenoid a velocity valve greater was thanopened, the theset blockedvalue at plasmathat moment. gas flow Then, was thereleased plasma and input into the plasma arc at a velocity greater than the set value at that moment. Then, the plasma gas gasinput flow into returned the plasma to the arcset valueat a velocity and waited greater the than next theclosure set value of the at solenoid that moment. valve. Then, the plasma gasflow flow returned returned to the to setthe valueset value and and waited waited the nextthe next closure closure of the of solenoidthe solenoid valve. valve.

Figure 2. The principle of PPG-VPPAW process. Figure 2. The principle of PPG-VPPAW process. In order to study the effectFigure of pulsed2. The principle plasma gasof PPG-VPPAW acting on the process. plasma arc, the arc electric signals andIn arc order image to study were acquiredthe effect from of pulsed PPG-VPPAW plasma gas and acting compared on the with plasma those arc, in VPPAW the arc electric and DP-VPPAW. signals andThe arc currentIn orderimage and to were study plasma acquired the gas effect flow fromof waveforms pulsed PPG-VPPAW plasma of the gas VPPAW,and acting compared on DP-VPPAW, the plasma with and arc,those PPG-VPPAW the inarc VPPAW electric processes signals and DP-VPPAW.andare shown arc image in The Figure werecurrent3. Theacquired and VPPAW plasma from process PPG-VPPAWgas has flow the variablewaveforms and polaritycompared of the VPPAW,with current those DP-VPPAW, wave in VPPAW with unequal and and PPG-VPPAWDP-VPPAW.straight and processes reverseThe current polarity are shown and time plasma in intervals, Figure gas 3. and Theflow the VPPAW plasmawaveforms process gas flowof hasthe rate theVPPAW, remains variable constant,DP-VPPAW, polarity as square shown and currentPPG-VPPAWin Figure wave3a. with For processes theunequal DP-VPPAW are straight shown process, and in Figure reverse the 3. current polarity The VPPAW wave time has interval process beens, decreasedhas and the the variable plasma periodically, polarity gas flow the square rate total remainscurrent constant, wave with presents as unequal shown two periodicallyinstraight Figure and 3a. varying reverseFor the polarity pulses DP-VPPAW (i.e., time a high-frequencyintervalprocess,s, theand current the variable plasma wave polarity gas has flow been pulse rate decreasedremainsand a low-frequency constant,periodically, as shown pulse),the total in and Figurecurrent the plasma3a. wave For gas thepresents flowDP-VPPAW rate two is periodically the process, same asthe varying for current the VPPAW pulseswave has(i.e., process, been a high-frequencydecreasedas shown inperiodically, Figure variable3b. polarity Inthe the total PPG-VPPAWpulse current and awave low-frequency process, presents the current twopulse), periodically waveformand the plasma varying is the gas same pulses flow as rate for(i.e., theis a theVPPAWhigh-frequency same as process for the variable while VPPAW the polarity plasmaprocess, pulse gas as flowshown and rate a in low-frequency is Figure input 3b. intermittently In thepulse), PPG-VPPAW and in the the form plasma process, of a pulse,gas the flow ascurrent shownrate is waveformthein Figure same 3asisc. forthe Based the same VPPAW on aas large for process, numberthe VPPAW as ofshown trials, proce in the Figuress process while 3b. In parametersthe the plasmaPPG-VPPAW in gas this flow study process, rate were theis selected currentinput waveform is the same as for the VPPAW process while the plasma gas flow rate is input intermittentlyin Table1 to makein the the form comparison of a pulse, clear. as shown In Table in 1Figure, IEN was3c. Based the electrode on a large negative number current; of trials,IEP thewas intermittently in the form of a pulse, as shown in Figure 3c. Based on a large number of trials, the processthe electrode parameters positive in this current; study IwereB was selected the basic in Table current 1 to in make DP-VPPAW; the comparisonIP was theclear. pick In currentTable 1, in process parameters in this study were selected in Table 1 to make the comparison clear. In Table 1, 𝐼DP-VPPAW. was the electrode The arc imagenegative capture current; rate 𝐼 was was 3000 the fps electrode and the positive signal sampling current; rate𝐼 was was the 10,000 basic Hz. 𝐼 was the electrode negative current; 𝐼 was the electrode positive current; 𝐼 was the basic currentThe plasma in DP-VPPAW; gas and the 𝐼 shielding was the pick gas bothcurrent were in pureDP-VPPAW. argon. The arc image capture rate was 3000 fpscurrent and the in signalDP-VPPAW; sampling 𝐼 rate was was the 10000pick current Hz. The in plasma DP-VPPAW. gas and The the arc shielding image capture gas both rate were was pure 3000 argon.fps and the signal sampling rate was 10000 Hz. The plasma gas and the both were pure argon.

Materials 2019, 12, 1666 4 of 14 Materials 2019, 12, x FOR PEER REVIEW 4 of 14

(a)

(b)

(c)

FigureFigure 3. SchematicSchematic diagram diagram of of current–plasma current–plasma gas gas flow flow waveform. waveform. (a) (VPPAW;a) VPPAW; (b) (DP-VPPAW;b) DP-VPPAW; (c) (PPG-VPPAW.c) PPG-VPPAW.

Table 1. Parameters for arc image and electrical signal comparison.

Shielding Plasma Gas Arc Length Gas Flow 𝑰 :𝑰 Number Setback Flow Rate 𝑬𝑵 𝑬𝑷 (mm) Rate (A) (mm) (L/min) （L/min） 1-1 6 4 10 3.0 100:100

Materials 2019, 12, 1666 5 of 14

Table 1. Parameters for arc image and electrical signal comparison.

Shielding Gas Plasma Gas Arc Length Tungsten Number Flow Rate Flow Rate I :I (A) (mm) Setback (mm) EN EP (L/min) (L/min) 1-1 6 4 10 3.0 100:100 1-2 6 4 10 3.0 100:100 1-3 6 4 10 3.0 100:100 1-4 6 4 10 3.0 100:100 Electric Pulse Solenoid Valve Electric Pulse Solenoid Valve Number I :I (A) Frequency Frequency B P Duty Cycle Duty Cycle (Hz) (Hz) 1-1 - - - - - 1-2 80% 3 5:3 - - 1-3 - - - 4 4:1 1-4 - - - 20 4:1

In order to study the effect of pulsed plasma gas on the molten pool in the keyhole welding process, the weld-forming experiments were carried out. The weld bead geometries and porosity distribution from the VPPAW, DP-VPPAW, and PPG-VPPAW processes were investigated. In order to study the influence of the pulsed plasma gas on the molten pool, the filling wire was not applied for avoiding the effect of filling material on molten pool behavior and simplifying the experimental model. In these experiments, 5 mm thick 5A06 aluminum alloy was selected as the work piece. The plasma gas and the shielding gas both were pure argon. Based on a large number of trials, the parameters in this part were selected to obtain a good weld bead geometry, as shown in Table2.

Table 2. Parameters for weld-forming experiments.

Shielding Plasma Gas Tungsten Number Arc Length (mm) Gas Flow Flow Rate I :I (A) Setback (mm) EN EP Rate (L/min) (L/min) 2-1 5 4 10 3.0 130:150 2-2 5 4 10 3.0 130:150 2-3 5 4 10 3.0 130:150 2-4 5 4 10 3.0 130:150 Welding Electric Pulse Electric Solenoid Solenoid Valve Number I :I (A) Speed Frequency Pulse Duty Valve Duty B P Frequency (Hz) (mm/s) (Hz) Cycle Cycle 2-1 - 1.83 - - - - 2-2 80% 1.83 3 5:3 - - 2-3 - 1.83 - - 20 4:1 2-4 - 1.83 - - 40 4:1

3. Results and Discussion

3.1. Variation in Arc Electrical Signal The welding electrical signals of VPPAW (Experiment 1-1) and DP-VPPAW (Experiment 1-2) are shown in Figure4a,b, respectively. It can be observed that the current waves well fit the preset parameters. The root mean square (RMS) and absolute mean (AM) values of current are shown in Table3. In comparison with traditional VPPAW, the current of PPG-VPPAW was almost unchanged. Figure4c,d show the welding electrical signals of PPG-VPPAW under di fferent plasma gas flow pulse frequencies with the same plasma flow rate; the plasma gas flow pulse frequencies were 4 Hz (Experiment 1-3) and 20 Hz (Experiment 1-4), respectively. Compared with Figure4a,c, it can be observed that the current wave had a negligible effect on the pulsed plasma gas, while the arc voltage Materials 2019, 12, 1666 6 of 14

fluctuatedMaterials 2019 periodically, 12, x FOR PEER with REVIEW the pulsed plasma gas frequency. The electrode negative period6 of 14 voltage (UEN) in the VPPAW process and solenoid valve on state of the PPG-VPPAW process was 24.97 V on average,24.97 V and on average, the electrode and the positive electrode period positive voltage period (U voltageEP) was (𝑈 35.49) was V on35.49 average. V on average. When theWhen solenoid the solenoid valve was off state and the gas flow pulse frequency was 4 Hz, the 𝑈 and 𝑈 valve was off state and the gas flow pulse frequency was 4 Hz, the UEN and UEP decreased by 5.8 V and decreased by 5.8 V and 9.8 V, respectively. When the gas flow pulse frequency was 20 Hz, the 𝑈 9.8 V, respectively. When the gas flow pulse frequency was 20 Hz, the UEN and UEP decreases were and 𝑈 decreases were 5.1 V and 8.9 V, respectively. The above results showed that the lower the 5.1 V and 8.9 V, respectively. The above results showed that the lower the frequency of plasma gas, frequency of plasma gas, the more distinct the influence on the arc voltage waveform. At a certain the more distinct the influence on the arc voltage waveform. At a certain plasma gas pulse frequency, plasma gas pulse frequency, the pulsed plasma gas has a greater influence on the electrode positive thepolarity pulsed voltage. plasma gas has a greater influence on the electrode positive polarity voltage. ） A （ Current

U EN UEN ） V （ Voltage

UEP UEP

(a) (b)

Current

Voltage

ΔUEN U ENmin ΔUEN =5.8V U ENmin =19.8V

UEN=24.97V UEN

UEP=35.49V UEP

ΔUEP =9.8V U EPmin =28.1V U ΔU EP EPmin

（ ） Time ms

(c) (d)

FigureFigure 4. 4. Welding current–voltage current–voltage synchronizing synchronizing wave wave form. form.(a) VPPAW; (a) VPPAW; (b) DP-VPPAW; (b) DP-VPPAW; (c) (c)PPG-VPPAW PPG-VPPAW (4 (4 Hz); Hz); (d () dPPG-VPPAW) PPG-VPPAW (20 (20Hz). Hz).

Materials 2019, 12, x FOR PEER REVIEW 7 of 14 Materials 2019, 12, 1666 7 of 14 Table 3. The root mean square (RMS) and absolute mean (AM) values of current.

Welding Methods RMS（A） AM （A） Table 3. The root mean square (RMS) and absolute mean (AM) values of current. VPPAW 100.95 100.66 DP-VPPAWWelding Methods93.56 RMS (A) AM (A) 92.39 PPG-VPPAW(4VPPAW Hz) 100.98 100.95 100.66 100.66 PPG-VPPAW(20DP-VPPAW Hz) 100.84 93.56 92.39 100.08 PPG-VPPAW(4 Hz) 100.98 100.66 PPG-VPPAW(20 Hz) 100.84 100.08 When the plasma gas injects into the arc column, it requires more energy to keep enough ionized particles that let the current through. When the plasma gas is shut off, the required energy whichWhen was theused plasma to ionize gas the injects gas intodecreases, the arc and column, the arc it requiresvoltage morealso decreases. energy to keepThis decrease enough ionized relates particlesto the shut-off that let time the currentof plasma through. gas and When has the a maximum plasma gas value. is shut With off, the the required frequency energy of plasma which wasgas usedincreased, to ionize the the shut-off gas decreases, time decreases and the arcwith voltage lack alsoof time decreases. to let Thisthe decreasearc voltage relates decrease to the shut-o to theff timemaximum of plasma value. gas andIn the has EP a maximum phase of value. the PPG-VPP With the frequencyAW process, of plasma due to gas the increased, cathodic the cleaning shut-off timephenomenon decreases [15], with the lack size of of time the to arc let profile the arc is voltage larger decreasethan that to in the the maximum EN phase, value. which In lets the it EP be phasemore ofaffected the PPG-VPPAW by the plasma process, gas shut-off. due to the cathodic cleaning phenomenon [15], the size of the arc profile is largerFigure than that5 displays in the ENmore phase, details which about lets the it bearc more voltage affected of PPG-VPPAW by the plasma with gas the shut-o plasmaff. gas flow pulseFigure frequencies5 displays of 4 more Hz and details 20 Hz. about The the pulse arc voltage signals ofof PPG-VPPAW“1” and “0” withindicate the that plasma the gassolenoid flow pulsevalve frequenciesis in the on of state 4 Hz and and off 20 st Hz.ate, The respectively. pulse signals It can of “1” clearly and “0”be indicateseen that that with the the solenoid plasma valve gas isflow in thepulse on statefrequency and o ffofstate, 4 Hz, respectively. the arc voltage It can is clearly dropped be seen immediately that with theto the plasma minimum gas flow voltage pulse frequencywhen the ofplasma 4 Hz, gas the arcis shut voltage off, iswhich dropped costs immediately 14 ms, and to the the voltage minimum decrease voltage rate when is the0.44 plasma V/ms. gasWhen is shutthe plasma off, which gas costs is turned 14 ms, on, and it thetakes voltage 28 ms decrease to return rate the is average 0.44 V/ms. voltage, When and the plasmathe voltage gas isincrease turned rate on, is it takes0.20 V/ms 28 ms accordingly. to return the Once average the plasma voltage, gas and flow the pulse voltage frequency increase increases rate is 0.20 to 20 V/ msHz, accordingly.the voltage decrease Once the rate plasma is nearly gas flow same pulse as frequencythat in 4 Hz increases (0.46 V/ms), to 20 Hz, but the it is voltage hard to decrease find a ratestable is nearlyminimum same voltage as that due in 4 to Hz the (0.46 lack V /ofms), shut-off but it istime hard of tothe find plasma a stable gas. minimum When the voltage plasma due gas to isthe turned lack ofon shut-oin 20 Hzff time situation, of the plasmathe voltage gas. recovery When the time plasma is 22 gas ms, is and turned the voltage on in 20 increase Hz situation, rate is the 0.23 voltage V/ms recoveryaccordingly. time The is 22 above ms, and results the voltageshow that increase the plasma rate is 0.23 arc Vneeds/ms accordingly. more time to The recover above resultsthe effect show of thatplasma the gas plasma closure, arc needsand the more plasma time gas to recoverflow pulse the frequency effect of plasma has less gas effect closure, on the and voltage the plasma decrease gas flowrate. pulse frequency has less effect on the voltage decrease rate.

(a) (b)

FigureFigure 5. 5. TheThe arc arc voltage voltage of of PPG-VPPAW. PPG-VPPAW. ( a) PPG-VPPAW (4(4 Hz);Hz); ((bb)) PPG-VPPAWPPG-VPPAW (20 (20 Hz). Hz).

When the plasma gas is shut ooff,ff, there is nono plasmaplasma gas injecting immediately and the arc zone could be considered considered as as a a relative relative closed closed environm environmentent and and establish establish balance balance ea easily.sily. Compared Compared with with it, it,when when the the plasma plasma gas gas is isturned turned on, on, the the plasma plasma gas gas continuously continuously injects injects into into the the arc arc column and makes it harder to establishestablish balance, so the arc voltage needs moremore time to stabilize after the plasma gas is turned on. Furthermore, Furthermore, the the higher higher the the pl plasmaasma gas gas frequency frequency is, is, th thee lower the plasma gas accumulation rate is, soso thethe arc voltagevoltage needs less timetime to stabilize after the plasma gas is turned on. on. The voltage decrease rate is correlated with arc characteristic, but there is no significantsignificant correlation

Materials 2019, 12, 1666 8 of 14 Materials 2019, 12, x FOR PEER REVIEW 8 of 14 with the frequency of plasma gas. However, However, when when the the frequency of of plasma gas is above a certain value, thethe plasmaplasma gas gas is is turned turned on on again again when when the arcthe voltagearc voltage is not is yet not decreased yet decreased to the stableto the voltage stable voltagewithout without plasma gas,plasma so thegas, recovery so the recovery speed of speed the voltage of the becomesvoltage becomes faster. faster.

3.2. Variation Variation in in Arc Arc Profile Profile In order to better observe and analyze the variat variationion in arc profile profile collected by the high-speed video camera, it is necessary to divide the arc region for regionalizationregionalization and measurement. The arc image processing procedure procedure is is shown in Figure 66.. TheThe originaloriginal arcarc imageimage isis shownshown inin FigureFigure6 6a.a. Figure 66bb isis thethe coloredcolored gray-scalegray-scale imageimage thatthat isis transformedtransformed fromfrom thethe originaloriginal arc.arc. AccordingAccording toto thethe intensity of of the arc light, the arc cancan bebe divideddivided intointo 256256 scalescale levels.levels. Yang etet al.al. illustrated that the gray scale was larger with higher intensity in the arc gray-scale image, with 90% of the arc intensity as the arc core region [[16,17].16,17]. The gray-scale im imageage was colored by RGB for the convenience of arc core region [18,19], [18,19], the the red red and green green region region of of processed arc arc image were were the the core core and edge region, region, respectively. As As a a result, result, the the arc arc profile profile would be be measured measured and and analyzed analyzed using using the the arc arc core core region, region, as presented in Figure 66b.b. TheThe arc profile,profile, whichwhich waswas defineddefined byby arcarc corecore lengthlength LL andand arcarc corecore widthwidth under the nozzle D, was accurately measured by computerized measurement technique.technique.

Arc core D region Arc edge region

L

(a) (b)

Figure 6. EdgeEdge extraction extraction and and regiona regionalizationlization of of welding welding arc. arc. (a) (Originala) Original arc arcimage; image; (b) processed (b) processed arc image.arc image.

Figure 77aa displaysdisplays thethe variationvariation ofof arcarc profileprofile inin aa completecomplete plasmaplasma gasgas cyclecycle ofof PPG-VPPAWPPG-VPPAW with the 4 4 Hz Hz plasma plasma gas gas pulse pulse frequency frequency in in EN EN ph phase.ase. Figure Figure 8a8 ashows shows the the corresponding corresponding L Land and D, whichD, which were were measured measured using using the the method method mentioned mentioned above. above. From From 0 0to to 30 30 ms, ms, the the arc arc images images were relatively stable, the L and D were about 3.76 mm and 1.05 mm, respectively. When the plasma gas was shut off, off, from 30 ms to 60 ms, the arc constricted rapidly with with the the L L and and D D decreased decreased to to 0 0 mm. mm. From 60 ms to 80 ms, the core region of the arcarc almost disappeared because there was no plasma gas supply. When When the the plasma plasma gas gas was was turned turned on, on, fr fromom 80 80 ms ms to to 100 100 ms, ms, the the core region of the arc significantlysignificantly increasedincreased withwith LL and and D D increased increased to to 3.18 3.18 mm mm and and 1.27 1.27 mm, mm, respectively. respectively. Then, Then, the Lthe had L hada decrease a decrease to 2.07 to mm2.07 at mm 110 at ms 110 and ms increased and increased again to again 6 mm to in 6 140 mm ms. in After140 ms. that, After the Lthat, gradually the L graduallydecreased decreased to 4.09 mm to from 4.09 140 mm ms from to 170 140 ms ms and to then170 ms maintained and then inmaintained a stable state. in a In stable this period,state. In the this D period,only had the a slightD only change had a dueslight to change the restriction due to the of therestriction torch orifice. of the Thetorch above orifice. results The showabove that results the showarc core that length the arc had core a short length decrease had a short during decrease the general during rising the trendgeneral after rising plasma trend gas after was plasma turned gas on. wasThese turned two peaks on. These of arc two core peaks length of were arc core due length to the initialwere due overshoot to the phenomenon.initial overshoot Once phenomenon. the plasma Oncegas is the turned plasma on, thegas overshoot is turned velocity on, the isovershoot higher than velocity the setting is higher value, than which the causes setting the value, subsequent which causesplasma the gas subsequent to not keep plasma up, leading gas to the no Lt keep to decrease. up, leading the L to decrease. Figure 77bb shows shows the the variation variation of of arc arc profile profile in ain complete a complete plasma plasma gas cyclegas cycle of PPG-VPPAW of PPG-VPPAW with withthe 20 the Hz 20 plasma Hz plasma gas pulsegas pulse frequency frequency in the in ENthe phase.EN phase. The The corresponding corresponding L and L and D areD are shown shown in in Figure 8b. It can be seen that the change of PPG-VPPAW arc profile with 4 Hz and 20 Hz plasma gas pulse frequencies have the same tendency. As mentioned before, due to the increase of plasma

Materials 2019, 12, 1666 9 of 14

FigureMaterials8 b.2019 It, can12, x beFOR seen PEER that REVIEW thechange of PPG-VPPAW arc profile with 4 Hz and 20 Hz plasma9 of gas 14 Materials 2019, 12, x FOR PEER REVIEW 9 of 14 pulse frequencies have the same tendency. As mentioned before, due to the increase of plasma gas pulsegasgas pulsepulse frequency, frequency,frequency, there therethere is not isis enough notnot enoughenough shut-o shut-offshut-offff time timetime of plasma ofof plasmaplasma gas to gasgas aff toectto affectaffect the arc thethe profile arcarc profileprofile fully, fully, thefully, L thecannotthe LL cannotcannot decrease decreasedecrease to 0 mm toto 00 from mmmm fromfrom 8 ms 8 to8 msms 18 toto ms, 1818 and ms,ms, theandand first thethe first peakfirst peakpeak value valuevalue is also isis alsoalso less less thanless thanthan that thatthat in 4 inin Hz. 44 Hz.InHz. the InIn whole thethe wholewhole cycle cyclecycle with withwith 20 Hz 2020 plasma HzHz plasmaplasma gas pulse gasgas pulspuls frequency,ee frequency,frequency, the core thethe region corecore regionregion of the of arcof thethe always arcarc alwaysalways exists, andexists, the and D keeps the D balancedkeeps balanced at 1.02 at 0.171.02 mm.± 0.17 mm. exists, and the D keeps balanced at± 1.02 ± 0.17 mm.

((aa))

((bb)) Figure 7. Comparison of arc images at negative polarity stage. (a) PPG-VPPAW (4 Hz); (b) Figure 7.7.Comparison Comparison of arcof arc images images at negative at negative polarity polarity stage. (a )stage. PPG-VPPAW (a) PPG-VPPAW (4 Hz); (b) PPG-VPPAW(4 Hz); (b) PPG-VPPAW (20 Hz). (The color bar of this figure is same as Figure 6). (20PPG-VPPAW Hz). (The color(20 Hz). bar (The of this color figure bar is of same this figure as Figure is same6). as Figure 6).

((aa)) ((bb)) FigureFigure 8. 8. ArcArc core core region region size size of of PPG-VPPAW. PPG-VPPAW. (a)) PPG-VPPAWPPG-VPPAW (4(4 Hz);Hz); ((bb)) PPG-VPPAWPPG-VPPAW (20 (20 Hz). Hz). Figure 8. Arc core region size of PPG-VPPAW. (a) PPG-VPPAW (4 Hz); (b) PPG-VPPAW (20 Hz).

FigureFigure 99 displaysdisplays thethe arcarc profileprofile ofof EPEP phasephase inin differentdifferent VPPAWVPPAW processesprocesses andand stages.stages. FigureFigure 9a9a showsshows thethe arcarc profileprofile inin thethe VPPAWVPPAW process.process. FigureFigure 9b,c9b,c showshow shutting-offshutting-off andand turning-onturning-on stages,stages, respectively,respectively, withwith 4Hz4Hz PPG-VPPAW.PPG-VPPAW. FigureFigure 9d,e9d,e showshow shutting-offshutting-off andand turning-onturning-on stages,stages,

Materials 2019, 12, 1666 10 of 14

MaterialsFigure 2019,9 12 displays, x FOR PEER the REVIEW arc profile of EP phase in di fferent VPPAW processes and stages. Figure10 of9 14a showsMaterials the2019 arc, 12, profilex FOR PEER in theREVIEW VPPAW process. Figure9b,c show shutting-o ff and turning-on10 stages, of 14 respectively, withwith 4Hz20 Hz PPG-VPPAW. PPG-VPPAW. Figure Compared9d,e show with shutting-o these arcff profiles,and turning-on it is easy stages, to find respectively, that only withinrespectively, the 20 shut-off Hz PPG-VPPAW. with stage 20 withHz PPG-VPPAW. Compared4 Hz PPG-VPPAW with Compared these process, arc with profiles, the th esearc it arc iscore easy profiles, length to find Lit significantly thatis easy only to infind thedecreased. that shut-o onlyff Thisstagein the phenomenon with shut-off 4 Hz stage PPG-VPPAW cannot with 4be Hz process,found PPG-VPPAW when the arc the core process,plasma length gasthe L pulse significantlyarc core frequency length decreased. L increased significantly This to phenomenon 20 decreased. Hz. That meanscannotThis phenomenon bethe found influence when cannot of the plasma plasmabe found gas gas whenshutting pulse the frequency offplasma in the increasedgas EP pulse phase to frequency 20 of Hz. theThat VPPAW increased means process theto 20 influence Hz.is much That of plasmasmallermeans the gasthan influence shutting in the EN o offf inphase,plasma the EP especially gas phase shutting of in the a VPPAWhighoff in plasma the process EP gas phase is pulse much of frequency.the smaller VPPAW than This process in is the because EN is phase,much the arcespeciallysmaller core thanregion in ain high isthe directly plasmaEN phase, gasrelated pulseespecially to frequency.the temperatin a high Thisure plasma is becausefield, gas which the pulse arc has corefrequency. such region a great isThis directly inertia is because related that the to temperaturethearc temperaturecore region change is field, directly could which relatednot has quickly such to the arespond great temperat inertia to urethe that disturbance.field, the which temperature hasWhen such the change a plasma great could inertia gas not frequency that quickly the increases,respondtemperature to the the change effect disturbance. of could pulsed not When plasma quickly the gas plasma respond on the gas toEP frequency the phase disturbance. is increases,harder toWhen theobserve. e theffect plasma of pulsed gas plasma frequency gas onincreases, the EP phasethe effect is harder of pulsed to observe. plasma gas on the EP phase is harder to observe.

Figure 9. Comparison of arc images at positive polarity stage. ( The color bar of this figure is same as figure 6.) Figure 9. Comparison of arc images at positive polarity stage. ( (TheThe color bar of this figurefigure is same as Figurefigure 66.)). 3.3. Variation in Welding Formation 3.3. Variation Figure 10 in shows Welding the Formation front and back of the PPG-VPPA (20 Hz) weld joint. It can be clearly seen that theFigure keyhole 1010 showsshows is completely thethe front front and andpenetrated back back of of the atthe PPG-VPPAthe PPG-VPP end of (20 Athe (20 Hz) weld Hz) weld weldjoint. joint. joint.The It can weldIt can be clearlyjoint be clearly formed seen seen that as shownthethat keyhole the in keyhole Figure is completely 10is completelyis due penetratedto the penetrated lack atof themetal at end the fill of ingend the and of weld thethe joint. weldmolten Thejoint. pool weld The flow joint weld toward formed joint the formed as front shown ofas theinshown Figure weld. in 10Figure is due 1011 to isshows the due lack tothe the of appearance metal lack of filling metal of and the fill theingweld moltenand surface the pool molten with flow differentpool toward flow welding the toward front methods. ofthe the front weld. ofA smoothFigurethe weld. 11 weld showsFigure joint the 11 profile appearanceshows ofthe VPPAW appearance of the weld is shown of surface the inweld with Figure surface diff erent11a. with As welding showndifferent methods. in welding Figure A smooth 11b–d,methods. weldboth A DP-VPPAWjointsmooth profile weld ofand joint VPPAW PPG-VPPAW profile is shownof VPPAW form in Figurea fish-scaleis shown 11a. pattern Asin shownFigure on the in11a. Figure surface As shown 11 ofb–d, the in bothweld Figure DP-VPPAW joint 11b–d, due to both andthe stirringPPG-VPPAWDP-VPPAW action and form on PPG-VPPAW the a fish-scale welding patternformpool acaused fish-scale on the by surface thpatterne periodical of theon the weld oscillationsurface joint dueof theof to the theweld arc stirring joint pressure. due action to The the on ripplesthestirring welding formedaction pool onon caused the the welding weld by the joint periodicalpool surface caused oscillation show by th thate periodical of the the fluidity arc pressure.oscillation of the The ofwelding ripplesthe arc pool formedpressure. has on been The the weldenhanced,ripples joint formed and surface benefits on showthe theweld that welding thejoint fluidity surfacequality. of showCompared the welding that thewith pool fluidity Figure has been 11b,c,of the enhanced, it weldingis clearly and pool demonstrated benefits has been the weldingthatenhanced, the quality.plasma and benefits Comparedgas frequency the withwelding has Figure aquality. stronger 11b,c, Compared it effect is clearly acting with demonstrated on Figure the welding 11b,c, that it the formationis clearly plasma demonstrated gasthan frequency current frequencyhasthat athe stronger plasma under e ffgas ectthe frequencyactingpremise on of thehas acceptable welding a stronger formation welding effect actingformation. than currenton the Compared welding frequency formationwith under Figure the than premise11c,d, current the of densenessacceptablefrequency ofunder welding ripple the profile formation. premise on theof Compared acceptableweld joint with weldingsurface Figure is formation. directly11c,d, the proportional Compared denseness oftowith ripplethe Figurefrequency profile 11c,d, onof the weldplasmadenseness joint gas. surfaceof ripple is profile directly on proportional the weld joint to the surface frequency is directly of the proportional plasma gas. to the frequency of the plasma gas.

Figure 10. TheThe front front and and back back of of the PPG PPG-VPPA-VPPA (20 Hz) weld joint.

Figure 10. The front and back of the PPG-VPPA (20 Hz) weld joint.

Materials 2019, 12, 1666 11 of 14 Materials 2019, 12, x FOR PEER REVIEW 11 of 14

(a) (b)

(c) (d)

Figure 11. MorphologyMorphology of of weld weld surface. surface. (a) (VPPAW;a) VPPAW; (b) (DP-VPPAW;b) DP-VPPAW; (c) PPG-VPPAW (c) PPG-VPPAW (20 (20Hz); Hz); (d) (PPG-VPPAWd) PPG-VPPAW (40 (40Hz). Hz).

The preparation method of the samplesample isis shownshown inin FigureFigure 1212.. Three samples were selected in every weld joint. We We observed that the characterist characteristicsics of of the three samples of the same weld joint were basicallybasically thethe same.same. Therefore, a sample of each weld joint was randomly selected for further analysis. The cross-sectionscross-sections of of the the weld weld joint joint with with di differentfferent welding welding methods methods are are exhibited exhibited in Figure in Figure 13a. The13a. averageThe average weld weld reinforcement reinforcement and width and width are measured are measured as shown as shown in Table in4 .Table The VPPAW 4. The VPPAW process andprocess PPG-VPPAW and PPG-VPPAW process process have nearly have thenearly same the current same current wave, wave, so the so weld the weld width width from from these these two processestwo processes are similar. are similar. For the For DP-VPPAW the DP-VPPAW process, process, due to thedue decrease to the decrease of current of in current the low in frequency, the low thefrequency, weld width the weld also decreases. width also The decreases. distribution The ofdistribu porositytion in of the porosity weld fusion in the line weld region fusion with line di regionfferent weldingwith different methods welding is shown methods in Figure is shown 13b, in where Figure WZ 13b, is where weld zone, WZ is FZ weld is fusion zone, zone,FZ is fusion and HAZ zone, is heat-aand HAZffected is heat-affected zone. It shows zone. clearly It shows that the clearly porosity that has the appeared porosity bothhas appeared in fusion zoneboth andin fusion weld zone fromand weld the VPPAW zone from process, the VPPAW and the porosity process, could and the be observedporosity incould the fusionbe observed zone from in the the fusion DP-VPPAW zone process.from theIn DP-VPPAW contrast, there process. is no observableIn contrast, porosity there is inno the observable cross-section porosity from in the the PPG-VPPAW cross-section process. from Thethe resultsPPG-VPPAW of arc electrical process. characteristics The results of and arc arc electrical profile show characteristics that PPG-VPPA and arc periodically profile show fluctuates that duePPG-VPPA to the periodic periodically variation fluctuates of plasma due gas to the flow periodic rate. The variation molten poolof plasma oscillates gas periodicallyflow rate. The under molten the oscillatingpool oscillates arc. Therefore,periodically the un pulsedder the plasma oscillating gas has arc. a stronger Therefor effe,ect the than pulsed the pulsed plasma current gas has wave a whichstronger enhances effect than the the fluidity pulsed of thecurrent molten wave pool which and theenhances spillover the probability fluidity of the of the molten porosity pool from and the moltenspillover pool probability increases of [20 the,21 porosity]. from the molten pool increases [20,21].

Figure 12. The preparation method of the sample.

Materials 2019, 12, 1666 12 of 14 Materials 2019, 12, x FOR PEER REVIEW 12 of 14

(a) (b)

Figure 13. MacrographsMacrographs of of cross-section cross-section and and the the distributi distributionon of ofpores pores in weld in weld fusion fusion line line region. region. (a) (Macrographsa) Macrographs of cross-section; of cross-section; (b) (theb) thedisStrib distributionution of ofpores pores in in weld weld fusion fusion line line region. region.

Table 4. The weld reinforcement and width. Table 4. The weld reinforcement and width.

Welding Methods WeldWeld Width Width (mm) WeldWeld Reinforcement Reinforcement (mm) Welding Methods VPPAW 10.18（mm） （ 0.64mm） DP-VPPAWVPPAW 9.6410.18 0.82 0.64 PPG-VPPAW 10.01 0.73 DP-VPPAW 9.64 0.82 PPG-VPPAW 10.01 0.73 4. Conclusions 4. ConclusionsThis study proposed a novel plasma arc welding process, named pulsed plasma gas variable polarityThis plasma study proposed arc welding a novel process, plasma and arc investigated welding process, the effect named of pulsed pulsed plasma plasma gas gas on variable the arc electricalpolarity plasma signal, arcarc profile,welding and process, welding and formation. investigat Theed resultsthe effect can of be pulsed summarized plasma as gas follows: on the arc (1)electricalThe signal, shut-off arctime profile, is the and key welding factor aff formation.ecting arc behaviorThe results under can be the summarized pulsed plasma as follows: gas, and this (1)eff ectThe is shut-off stronger time in the is EPthephase key factor of the affecting variable arc polarity behavior plasma under arc the process. pulsed plasma gas, and (2)this effectIn the is EN stronger phase, in the the length EP phase of arc of core the variable region is polarity also affected plasma by arc the process. shut-off time of plasma gas. (2)Due In tothe the EN overshoot, phase, the two length peak of valuesarc core of region arc core is also length affected appear by in the the shut-off return time of plasma of plasma gas. gas. DueThe to pulsed the overshoot, plasma gas two has peak less values effect onof arc profilecore length in the appear EP phase. in the return of plasma gas. The pulsed plasma gas has less effect on arc profile in the EP phase. (3) Compared with pulsed current wave, the pulsed plasma gas could better enhance the fluidity of (3) Compared with pulsed current wave, the pulsed plasma gas could better enhance the the molten pool to reduce porosity during aluminum keyhole welding. fluidity of the molten pool to reduce porosity during aluminum keyhole welding.

Materials 2019, 12, 1666 13 of 14

Author Contributions: Z.L., S.C., and F.J. contributed to the project design; W.Z. and Z.Y. performed all experiments and data processing; F.J. wrote this manuscript; all authors participated in the discussion on the results and guided the writing of the article. Funding: This work is Supported by National Natural Science Foundation of China (No.51875004), Beijing Natural Science Foundation (No.3172004), the China Scholarship Council (CSC) for one-year study in University of Alberta for Fan Jiang, International Research Cooperation Seed Fund of Beijing University of Technology (No.2018B16), the open project of Beijing Engineering Researching Center of Laser Technology (BG0046-2018-09) and start-up funding from Beijing University of Technology. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Chen, S.J.; Zhang, R.Y.; Jiang, F.; Dong, S.W. Experimental study on electrical property of arc column in plasma arc welding. J. Manuf. Process. 2018, 31, 823–832. [CrossRef] 2. Jiang, F. Arc Behavior and Weld Pool Stability in Keyhole Plasma arc Welding. Ph.D. Thesis, Beijing University of Technology, Beijing, China, 2014. 3. Cong, B.Q.; Su, Y.; Qi, B.J. Development of pulsed arc welding technology for aluminum Alloy. Adv. Aviat. Weld. Technol. 2016, 11, 41–46. 4. Mahrle, A.; Schnick, M.; Rose, S. Process characteristics of fibre-laser-assisted plasma arc welding. J. Phys. D-Appl. Phys. 2011, 44, 345502. [CrossRef] 5. Mahrle, A.; Rose, S.; Schnick, M. Stabilisation of plasma welding arcs by low power laser beams. Sci. Technol. Weld. Join. 2013, 18, 323–328. [CrossRef] 6. Vredeveldt, H.L. Increased Power Density Plasma arc Welding: The Effect of an Added Radial Gas Flow around the arc Root. Master’s Thesis, Delft University of Technology, Delft, The Netherlands, 2014. 7. Wu, C.S.; Jia, C.B.; Chen, M.A. A Control system for keyhole plasma arc welding of plates with medium thickness. Weld. J. 2010, 89, 225–231. 8. Jia, C.B.; Wu, C.S.; Zhang, Y.M. Sensing controlled pulse key-holing condition in plasma arc welding. Trans. Nonferrous Met. Soc. China 2009, 19, 341–346. [CrossRef] 9. Wang, Y.; Qi, B.; Cong, B. Arc Characteristics in Double-Pulsed VP-GTAW for Aluminum Alloy. J. Mater. Process. Technol. 2017, 249, 89–95. [CrossRef] 10. Yang, M.; Li, L.; Qi, B. Arc force and shapes with high-frequency pulsed-arc welding. Sci. Technol. Weld. J. 2017, 22, 580–586. [CrossRef] 11. Chen, S.J.; Yang, X.; Yu, Y. Study on Low Frequency Vibration Pulse Plasma arc Welding of Ti-Alloy. Weld. J. 2011, 10, 1–6. 12. Wu, C.S.; Zhao, C.Y.; Zhang, C. Ultrasonic Vibration-Assisted Keyholing Plasma arc Welding. Weld. J. 2017, 96, 279–286. 13. Chen, S.J.; Jiang, F.; Zhang, J.L. Measurement and Analysis of Plasma arc Components. J. Manuf. Sci. Eng. 2015, 137, 011006. [CrossRef] 14. Chen, S.J.; Zhang, R.Y.; Jiang, F. Measurement and Application of arc Separability in Plasma arc. Weld. J. 2016, 95, 219–228. 15. Sarrafi, R.; Kovacevic, R. Cathodic cleaning of oxides from aluminum surface by variable-polarity arc. Weld. J. 2010, 89, 1–10. 16. Yang, M.Y.; Zheng, H.; Li, L. Arc shapes characteristics with ultra-high-frequency pulsed arc welding. Appl. Sci. 2017, 7, 45. [CrossRef] 17. Yang, M.Y.; Zheng, H.; Qi, B.J.; Yang, Z. Effect of arc behavior on Ti-6Al-4V welds during high frequency pulsed arc welding. J. Mater. Process. Technol. 2017, 243, 9–15. [CrossRef] 18. Gu, W.P.; Xiong, Z.Y.; Wan, W. Autonomous seam acquisition and tracking system for multi-pass welding based on vision sensor. Int. J. Adv. Manuf. Technol. 2013, 69, 451–460. [CrossRef] 19. Nele, L.; Sarno, E.; Keshari, A. An image acquisition system for real-time seam tracking. Int. J. Adv. Manuf. Technol. 2013, 69, 2099–2110. [CrossRef] Materials 2019, 12, 1666 14 of 14

20. Cong, B.; Ouyang, R.; Qi, B. Influence of Cold Metal Transfer Process and Its Heat Input on Weld Bead Geometry and Porosity of Aluminum-Copper Alloy Welds. Rare Metal Mater. Eng. 2016, 45, 606–611. [CrossRef] 21. Liu, A.; Tang, X.; Lu, F. Study on welding process and prosperities of AA5754 Al-alloy welded by double pulsed . Mater. Des. 2013, 50, 149–155. [CrossRef]

© 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/).