Effect of DC Micro-Pulsing on Microstructure and Mechanical Properties of TIG Welded Ti-6Al-4V

Abstract

This paper deals with the influence of micro-pulsed direct current on microstructure and mechanical properties of gas tungsten arc welding (GTAW) weldments of Ti-6Al-4V (Ti-64). Bead-on-plate GTA welds were made on the samples in the un-pulsed and micro-pulsed (125 Hz and 250 Hz) conditions. Post-weld heat treatment (PWHT) was performed on a few coupons at 700 °C for 3 h in an inert atmosphere, followed by furnace cooling. In the microstructure, the fusion zone (FZ), base metal (BM), and heat-affected zone (HAZ) can be easily distinguished. The top surface of the FZ has large columnar grains because of lower heat loss to the surrounding atmosphere, and the bottom region of the FZ has comparatively smaller equiaxed grains. The micro-pulsed samples’ FZ grain size was lower than that of those made without pulsing. This shows that high-frequency current has substantially refined prior β grains. The microstructure of the FZ is characterized by an acicular morphology composed of α, martensitic α′, and retained β phases. The FZ’s hardness was higher than the BM due to the presence of martensitic α′. Additionally, the hardness in the HAZ was elevated due to the formation of finer martensitic α′. Micro-pulsed DC welding led to improved mechanical properties, including higher hardness, ultimate tensile strength (UTS), and ductility compared to un-pulsed welding. This enhancement is attributed to the grain refinement achieved with micro-pulsed DC. After PWHT, the prior β grain size remained relatively unchanged compared to the as-welded condition. However, the hardness in the FZ decreased due to the decomposition of α′ into α and β phases. The ductility of all samples improved as a result of the widening of the diffusional α phase.

1. Introduction

The remarkable properties of Titanium (Ti), such as its lightweight, superior corrosion resistance, and outstanding strength-to-weight ratio, make it ideal for chemical, aircraft, marine, and petrochemical applications [1]. Pure-Ti has a density of 4.51 g/cm³ and melts at approximately 1670 °C [2]. Among the available grades, Ti-6Al-4V, an α-β Ti alloy, has the most significant commercial value in jet engine applications because of its superior strength and good formability [3]. Ti-alloys can be joined by many welding methods, such as solid-state, resistance welding, electron beam welding, etc. In the past, a successful joining of titanium alloys has been performed using linear friction welding [4] and laser beam welding (LBW) [5]. In friction welding, heating, and cooling rates vary continuously, plastic deformation occurs, and there will be excessive material flow. The resulting formation of defects, such as voids and flash formation, can weaken the strength of the weld nugget. Similarly, laser beam welding (LBW) may introduce issues, such as keyhole formation and porosity. In contrast, gas tungsten arc welding (GTAW) is the most commonly used method for welding titanium alloys. However, there are several problems associated with fusion welding of these alloys. If titanium is heated in air to temperatures above 650 °C, it undergoes rapid oxidation. Proper use of shielding gas can prevent contamination from oxygen [6]. However, during GTAW of Ti-6Al-4V, the base metal (BM) experiences a typical thermal cycle, which causes complex microstructural transformations. This results in an alteration of the mechanical properties of the weldment compared to BM. For instance, Wang et al. [6] have observed significantly grown prior β-grains in the heat-affected zone (HAZ) region of the GTA welded Ti-6Al-4V weldment due to the heating and cooling cycles. In addition, a lamellar α + β structure has been observed within the prior β-grains [7]. Similar observations are made by Pasang et al. [8] while studying the weld performance of Ti-alloys prepared using the LBW, GTAW, and electron beam welding process. In addition, some amount of retained β- is also observed. On the other hand, in the FZ of Ti-6Al-4V, the coarse prior β-grain, acicular α and α′ phases are formed as a result of solidification. Balasubramanian et al. [8] linked the formation of these microstructures to the steep thermal gradient and fast cooling rates characteristic of the welding process. Babu et al. [9] reported that the development of the α′-martensitic phase in the FZ of GTA-welded Ti-6Al-4V is due to the rapid cooling rates during the weldment’s solidification. This is further supported by Oh et al. [10], who compared GTAW and EBW processes. Their study found that EBW, with its higher cooling rates, resulted in a greater amount of α′ martensitic microstructure.

The microstructure of the FZ and HAZ has a substantial impact on the mechanical properties of the weldment, which is further dependent on the type of welding used. Balasubramanian et al. [11] conducted a comparative study on the microstructure, tensile, and impact properties of Ti-6Al-4V joints made by GTAW, EBW, and LBW. They concluded that the EBW joints exhibited a higher strength but lower ductility, while the GTAW joints showed a high ductility but low strength. EBW joints have higher strength due to the presence of fine serrate and regular plate-shaped α microstructure. As a consequence of the microstructural heterogeneity, a loss of mechanical properties has been observed in the Ti-6Al-4V weldments. In particular, the FZ of Ti-6Al-4V GTAW joint exhibits low ductility due to the formation of acicular microstructure, large β-grain size, and higher heat inputs [12]. A grain size reduction using appropriate techniques is often preferred to improve the welds’ ductility. Several methods are adopted for refining the prior β-grain size. Among these, the current pulsing technique is widely accepted, where the current supplied during welding varies from a maximum to minimum value at a pre-defined frequency by changing the heat supply.

In an early study, Reddy et al. [13] demonstrated the potential benefits of the current pulsing technique, such as improved mechanical properties. Their study reported an increase in the strength of titanium alloy GTA weldments due to grain refinement in the FZ caused by pulsing. Similar grain refinement has also been observed in GTAW joints of steel welds [14]. Furthermore, Sundaresan et al. [15] observed a greater grain refinement in α–β titanium alloy GTA welds using pulsed current, especially with alternating current (AC) due to increased weld pool agitation. Babu et al. [16] compared the mechanical properties of pulsed and un-pulsed GTA welded Ti-6Al-4V joints. They concluded that the current pulsing technique improved mechanical properties like microhardness, strength, and ductility. The influence of pulsing frequency on weldment properties is also widely studied. Yang et al. [17] investigated the arc shape characteristics in ultra-high-frequency pulsed arc welding and found that high-frequency pulsed current enhances heat input efficiency. They observed that pulsed arcs achieve deeper penetration, likely due to increased collisions that lower the temperature, reduce current density, and decrease energy transfer efficiency. Chai et al. [18] analyzed the effects of pulse parameters on weld bead geometry and microstructure, finding that an increased duty ratio leads to coarsening of the fusion zone microstructure but without a significant reduction in hardness. However, a higher duty ratio causes grain boundary liquation near the fusion line, which may compromise the weld bead’s performance. Therefore, excessive duty ratios should be avoided to prevent this issue. Balasubramanian et al. [19] investigated the effect of current pulsing and pulse frequency on the tensile characteristics of Ti-alloy welds. They reported that the effects of pulsing frequency and peak current on tensile properties follow a bimodal trend, initially increasing and then decreasing regardless of base current or pulse-on-time. Yuan et al. [20] studied the effect of pulsed current frequency on grain refinement during the solidification of Ti-64 alloy in plasma arc welding. They observed that higher pulsed current frequencies significantly improved grain refinement, particularly at the prior beta grain boundaries. The repeated heating from the pulsed current caused re-melting at the dendrite tips in the molten pool, promoting dendrite fragmentation. Balasubramanian et al. [21] investigated the effect of pulsed current GTA welding parameters on the weld characteristics of titanium alloys, including grain size and hardness. Their study revealed that an optimal pulse frequency of 6 Hz and a peak current of 80 A resulted in maximum hardness and minimum grain size. Pulsed current in inert gas welding proved more advantageous for titanium welding than conventional methods. Mehdi et al. [22] reported a reduction in residual stresses in the Ti-6Al-4V GTA weldments subjected to current pulsing at a frequency of 1 Hz and 5 Hz. They also reported a reduction in grain size when the pulse frequency increased from 1 Hz to 5 Hz.

DC micro-pulsing in Ti-6Al-4V (Ti-64) GTA welding significantly influences both the microstructure and width of the FZ and HAZ. It refines the grain structure in the FZ by promoting a more controlled thermal cycle, resulting in finer α + β phases instead of coarse martensitic structures. The reduced heat input from pulsing narrows the FZ, as the energy is more localized. In the HAZ, micro-pulsing suppresses grain growth and limits phase transformation, leading to a more refined microstructure and a narrower HAZ. Overall, DC micro-pulsing improves weld quality by refining microstructure and minimizing the extent of heat-affected areas [23].

Ultra-high frequency (UHF)—GTAW has also been shown to refine the grain structure in the FZ. UHF pulsing can enhance weld penetration, refine grain structure, constrict the welding arc, and reduce porosity. Additionally, it can produce smaller, more rounded weld beads, leading to improved surface quality. Yang et al. [24] found that UHF pulsing during GTAW of Ti-6Al-4V led to increased arc force and plasma core region. UHF-GTAW resulted in a 24–30% reduction in grain size compared to conventional arc welding due to arc energy constriction. This grain refinement contributed to higher basket weave formation, a type of microstructure that improves crack resistance. Additionally, in another work, they reported that the decreased FZ grain size in samples welded using UHF compared to conventional GTAW contributed to enhanced ductility [25]. Wang et al. [26] investigated the effects of high-frequency (HF) pulsing on weld penetration and grain refinement in GTA-welded Al-6061 alloy, finding a 120% increase in penetration and a 37% reduction in grain size with a 40 kHz pulsing current. Wu et al. [27] also reported similar improvements in Ti-64 alloy weldments, observing narrower bead width and deeper penetration with HF pulsed current. Wei et al. [28] have investigated the arc characteristics and weld bead microstructure of Ti-64 alloy by subjecting it to ultrahigh-frequency pulse (UHFP) current in GTAW. They observed that UHFP-GTAW significantly refines the grain size compared to the conventional GTAW process. This refinement is attributed to the concentrated heat input, which causes the fusion zone to experience rapid melting and solidification. Yang et al. [29] concluded that, compared to conventional GTAW, the UHF GTAW process results in reduced overall heat input, with the reduction becoming more pronounced as pulse frequency increases. This higher pulse frequency also influenced the microstructure of the FZ, leading to a uniform distribution of basket weave patterns and short acicular α′ martensite. These refined microstructures may enhance the mechanical properties of the weld. Arivarasu et al. [30] have investigated the effect of the frequency with respect to penetration and bead width to penetration ratio of gas tungsten arc welded AISI 304L alloy. They observed that complete penetration of the bead was achieved in the high-frequency (150 Hz) pulsing with a heat input of 652 J/mm whereas, in low-frequency pulsing (8 Hz and 10 Hz) the penetration was lower, respectively, with a heat input of 672 J/mm and 669 J/mm. Notably, the high-frequency welds demonstrated grain refinement in the fusion zone, attributed to the lower heat input and rapid solidification processes. This research underscores the critical role of frequency in enhancing weld quality and performance. Zhang et al. [31] have studied the high-frequency arc characteristics of SS304 alloy subjected to arc welding with pulsed current. They found that at 100 Hz, the weld mechanical properties hit their peak. This may be attributed to optimal frequency leading to the most effective stirring of the molten pool, which is confirmed by grain analysis. However, at higher frequencies like 500 Hz, the dynamic inertia effect lessened, and increased arc size, due to ionization and temperature rise, took precedence. There may be a resonance frequency that can further refine the grains even beyond 100 Hz. Dong et al. [32] investigated the influence of TIG arc characteristics on the weld morphology and structure of AISI 444 ferritic stainless steel using pulsed current. Their findings revealed that when subjected to low-frequency pulsing at 50 Hz, the weld microstructure exhibited varying amounts of equiaxed crystals across different positions. However, increasing the pulse frequency to 100 Hz led to a higher concentration of equiaxed crystal grains at the weld center. This phenomenon is attributed to the combined effects of arc pressure and electromagnetic forces within the molten pool, creating a force gradient during pulsing. As a result, these interactions promote the formation of smaller equiaxed crystal grains in the molten pool.

In the context of welding Ti-64 alloy, enhancing plasticity is critical for improving the overall performance of the welded joint. Ti-64 alloy is often used in applications where high strength-to-weight ratios and corrosion resistance are essential, such as in aerospace and biomedical industries. However, these welds can be brittle and susceptible to cracking due to their high reactivity and the residual stresses introduced during welding. By improving plasticity through post-weld heat treatment (PWHT), the welded joint can absorb more deformation before fracturing, thereby increasing the durability and integrity of the structure under service conditions. This plasticity improvement, accompanied by some reduction in hardness and strength, ensures that the welds are more resilient to mechanical stresses and better suited for critical applications [33]. The mechanical properties of Ti-alloy weldments can be influenced by modifying the microstructure through PWHT [34]. The PWHT leads to the transformation of martensite into β-phase, the precipitation of α-phase within the retained β, and the coarsening of the existing α-phase. Tsai and Wang [35] have reported a 5% decrease in the yield strength (YS) and UTS of the Ti-64 EB welds that were subjected to PWHT at 705 °C compared to the as-welded condition. This softening of the weldment has been attributed to the decomposition of α′ martensite into equilibrium α and β phases within the FZ. Similarly, Kabir et al. [36] observed that PWHT of Ti-6Al-4V welds fabricated by LBW reduced UTS, YS, and hardness compared to the as-welded samples. Although the strength and ductility values change during PWHT, it is noteworthy that, after PWHT, weldments produced with or without pulsing show identical strength and ductility values.

Existing literature highlights the significant influence of pulsing frequency on GTAW mechanical properties of Ti-alloys, but research has primarily focused on low frequencies (0–10 Hz) and ultra-high frequencies (kHz). The intermediate frequency range of 100–500 Hz remains relatively unexplored. Advances in current control technology now enable precise manipulation of current set-points within this range, offering new possibilities for improving weld properties. Micro-pulsing, a technique with millisecond-scale pulse durations, can achieve this frequency range. This study investigates the effects of micro-pulsing frequency on the microstructure and mechanical properties of Ti-6Al-4V weldments prepared using GTAW. A comparison is made between the weldments in un-pulsed and micro-pulsed conditions.

2. Experimental Details

In the present study, a 3 mm thick mill annealed Ti-6Al-4V sheet was used. The chemical composition of the BM is provided in

Table 1. Chemical composition of Ti-6Al-4V (wt. %).

Element	Al	V	Cr	Fe	H	C	Ni	O	Ti
Wt. %	6.26	4.04	0.17	0.05	0.001	0.013	0.008	0.16	Balance

. Prior to welding, the sheet was cut into coupons with dimensions of 110 mm × 100 mm × 3 mm using electric discharge machining (EDM). The samples were rubbed mechanically with a wire brush and cleaned with acetone to remove dirt particles prior to welding. Autogenous bead-on-plate welds were produced using the GTAW process in both un-pulsed and micro-pulsed modes (125 Hz and 250 Hz) conditions using direct current (DC). The parameters employed for GTAW are mentioned in

Table 2. Welding parameters for TIG welding of Ti-6Al-4V.

Parameters	Values
Electrode Dia (mm)	2.4
Peak current (A)	130
Background current (A)	42
Micro-pulsing frequency (Hz)	125, 250
Travel speed (cm/min) (both un-pulsed and micro-pulsed)	15
Argon gas flow rate (L/min) (both un-pulsed and micro-pulsed)	20
Average un-pulsed current (A)	100
Voltage (V) (both un-pulsed and micro-pulsed)	10.5
Arc length (mm) (both un-pulsed and micro-pulsed)	2

. To minimize contamination from atmospheric gasses like nitrogen, oxygen, and hydrogen, backing and trailing gasses were supplied, alongside the regular shielding gas that flowed from the torch. Upon completion of the welding, some coupons underwent PWHT in a vacuum furnace (VAC AERO International, Canada). The PWHT was conducted at 700 °C for 3 h in a vacuum environment, followed by furnace cooling to room temperature. The specimens were subsequently prepared for metallographic analysis and mechanical testing to compare with the as-welded joints.

Standard metallographic procedures were employed for both the as-welded and PWHT samples. The process involved rough polishing with SiC emery papers ranging from 600 to 2000 grit, followed by fine polishing using colloidal diamond suspensions on a cloth. The welds were then etched with Kroll’s reagent (a mixture of 1–3 mL HF, 2–6 mL HNO₃, and 100 mL distilled water) for 15 s. Microstructural features were analyzed with an optical metallurgical microscope (QUASMO QMM 500), whereas macrostructural features were evaluated using a stereo microscope (Leica Microsystems). A scanning electron microscope (TESCAN VEGA 3LMV) equipped with energy-dispersive spectroscopy capabilities was employed to investigate weld morphology, elemental chemical composition, and fractography.

Vickers micro-hardness measurements were performed across the weld regions using a Shimadzu HMV-G20S (Japan) microhardness tester. Measurements were taken at 0.5 mm intervals with a 200 g load applied for 15 s, using a diamond pyramid indenter, for both the PWHT and as-welded conditions. In accordance with ASTM E8M standards, three sub-sized specimens were extracted from each weld coupon for tensile testing. Each specimen had a gauge length of 32 mm and a width of 6 mm, with the weld positioned at the middle of the gauge length. The tensile tests were performed at room temperature using a Zwick Roell universal testing machine (Zwick Roell, Germany), with a cross-head speed of 0.5 mm/min and a load capacity of 100 kN. Finally, the fracture surfaces of the weld samples were examined with a scanning electron microscope (SEM).

3. Results and Discussion

3.1. Base Metal Microstructure

Figure 1 illustrates the optical microstructure of the Ti-6Al-4V BM, revealing a distinct heterogeneous mixture of primary α grains embedded within a matrix of transformed β phases. Figure 2 presents the BM’s scanning electron micrograph (SEM), which displays equiaxed α grains surrounded by intergranular β phases. The line intercept method was used to estimate the average grain size, which was found to be approximately 4 ± 1 μm.

3.2. Macrostructure

Figure 3a–c depict the cross-section macrostructures of Ti-6Al-4V welds produced under un-pulsed, 125 Hz micro-pulsed, and 250 Hz micro-pulsed conditions, respectively. In all cases, the BM, HAZ, and FZ are distinguishable. Macrostructural analysis of the transverse cross-sections revealed noticeable differences in weld bead shape and full penetration across all weldments. The widths of the FZ and HAZ are calculated for different processing conditions using image analysis software and are presented in

Table 3. The average grain size and width of different zones in Ti-6Al-4V weldments under various conditions.

Condition	Width of FZ + HAZ (mm)	Width of FZ (mm)	Width of HAZ (mm)	Average Grain Size in FZ (µm)	Average Grain Size in HAZ (µm)	Heat Input (J/mm)
Un-pulsed	14.64	11.22	3.42	374 ± 5	105 ± 5	70
Micro-pulsed-125 Hz	11.77	10.14	0.81	303 ± 6	85 ± 6	60.2
Micro-pulsed-250 Hz	10.5	9.57	0.75	272 ± 5	64 ± 6	55

. A careful observation of weld macrostructure revealed that the widths of FZ and HAZ are lower in the micro-pulsed condition compared to the unpulsed condition. This is attributed to decreased weld heat input as the welds are subjected to micro-pulsing, which results in rapid cooling rates. The calculated heat input values of the welds produced under un-pulsed, 125 Hz micro-pulsed, and 250 Hz micro-pulsed conditions are 70 J/mm, 60.2 J/mm, and 55 J/mm, respectively. The heat input values of the welds are also displayed in

Table 4. Tensile properties of GTA welded Ti-6Al-4V in as-welded and PWHT conditions.

Condition	Yield Strength (YS), MPa	Ultimate Tensile Strength (UTS), MPa	%Elongation (%El)
As-welded un-pulsed	980 ± 4	1014 ± 5	14 ± 0.5
As-welded 125 Hz	994 ± 6	1048 ± 4	16 ± 0.6
As-welded 250 Hz	1020 ± 5	1089 ± 5	18 ± 0.5
PWHT un-pulsed	944 ± 6	967 ± 6	22 ± 0.4
PWHT 125 Hz	957 ± 5	989 ± 6	23 ± 0.5
PWHT 250 Hz	972 ± 4	998 ± 4	23 ± 0.5

for a clear comparison of the welding conditions and their effect on the FZ and HAZ. In addition, welds made at higher micro-pulsed conditions (250 Hz) appear to have relatively narrow beads compared with those made at lower micro-pulsed conditions (125 Hz). The micro-pulsed welding with high frequency produces a more stable arc, enabling better control and potentially resulting in arc constriction [37]. This constriction may have contributed to adequate depth of penetration and narrower weld beads. Conversely, the less stable arc of the un-pulsed welding process led to wider weld beads.

In all the welds, the fusion zone (FZ) displays large columnar grains. These grains grow epitaxially in a direction perpendicular to the fusion boundary, driven by the higher thermal gradients near the boundary. The un-melted base material (BM) grains serve as a substrate for the epitaxial growth to initiate. All the welds exhibited coarse equiaxed grains in the heat-affected zone (HAZ) and finer equiaxed grains in the base material (BM). This microstructural variation is due to the different thermal cycles undergone during the welding process. Similar observations were made by Mishra et al. [38].

3.3. Microstrcutre

Figure 4a–c shows the FZ microstructure of the weldments in un-pulsed, micro-pulsed conditions at 125 Hz and 250 Hz, respectively. The fusion zone (FZ) consists of acicular α′ martensite within the prior β grains in all conditions. The welds produced using micro-pulsing exhibited narrower prior β grain widths compared to un-pulsed welds. This is attributed to the increased cooling rates and shortened exposure time above the β-transus experienced by the welds subjected to micro-pulsing. Additionally, all the welds displayed an acicular martensitic structure throughout the FZ. The FZ grain size measurements for welds prepared under un-pulsed, 125 Hz micro-pulsed, and 250 Hz micro-pulsed conditions showed average grain diameters of approximately 374, 303, and 272 µm, respectively, with grain aspect ratios ranging from two to six. These measurements were taken using the line intercept method with an image analyzer. Table 3 summarizes these findings. It seems the coarse columnar grain structure in the un-pulsed condition tends to be gradually transitioning to an equiaxed morphology in the micro-pulsed condition at both frequencies. This is attributed to the higher frequencies in micro-pulsing that led to increased arc forces, resulting in more nucleating sites and producing finer grains in FZ. Notably, the grain size of welds produced at a pulsing frequency of 125 Hz is larger compared to those created at 250 Hz. This may be due to the lower number of nucleating sites at a lower frequency compared to a higher frequency. These findings align with those reported by Mehdi et al. [8].

Figure 5 presents an optical micrograph of a Ti-64 micro-pulsed (250 Hz) weld cross-section, comparing the BM, FZ, and HAZ microstructures. The grain growth in the HAZ near the fusion line is due to the high temperatures approaching the melting point at the solid/liquid (S/L) interface. This is likely due to the region’s proximity to the heat source, resulting in higher peak temperatures. During the heating process, slower cooling allows a full transformation from α to β phase, whereas faster cooling results in the formation of acicular α′ martensite. The calculated average grain size in the HAZ of un-pulsed weldments is 144 µm, while micro-pulsed weldments at 125 Hz and 250 Hz exhibited slight grain refinement to 125 µm and 115 µm, respectively. This may be due to lower heat sinks in micro-pulsed welds compared to un-pulsed welds. These results align with the observations made by Kumar et al. [37]. Also, the width of the HAZ is lower in micro-pulsing compared to the un-pulsed condition due to the more controlled and localized heat input provided by the pulsed current. In micro-pulsing, the alternating periods of high and low current create intermittent heating, which reduces the overall thermal exposure of the surrounding material. This leads to less thermal diffusion and a narrower HAZ. Additionally, the repeated cooling intervals between pulses allow for better dissipation of heat, preventing excessive thermal accumulation in the HAZ. In contrast, the continuous heat input in un-pulsed welding results in a wider HAZ due to prolonged exposure to higher temperatures [39].

Figure 6a–c show the SEM FZ microstructure of the weldments in un-pulsed, micro-pulsed conditions at 125 Hz and 250 Hz, respectively. All the microstructures have dominance of acicular morphologies, which is possibly martensitic α′. Since, during welding, the temperature goes above beta transfuse-temperature, followed by rapid cooling tends to transform the beta phase to α՛ martensitic phase. For the Ti-64 alloy, the Ms (martensitic start temperature) is above room temperature; hence, this alloy, after the quenching or rapid cooling, can lead to the formation of the martensitic phase at room temperature. Therefore, FZ microstructure in both un-pulsed and micro-pulsed conditions reveals acicular martensitic structure, as shown in Figure 6a–c. However, the prior beta grain boundaries are not visible because of the dominance of the acicular microstructure. The microstructural transformations in the FZ occur due to micro-pulsing during the welding process. These results align with the findings of Sattar et al. [40] in their study of GTA welding of the Ti-6Al-4V alloy.

3.4. Mechanism of Grain Refinement

Grain refinement during solidification occurs when the following two conditions are met: (1) adequate undercooling and (2) the availability of suitable nucleating particles (nucleating agents) ahead of the solid–liquid (S/L) interface. This can be achieved by various methods, including heterogeneous nucleation via inoculation, physical disturbances such as electromagnetic stirring, and the use of pulsed welding currents. However, current pulsing is a key welding parameter affecting the weld cooling rates and weld heat input. Notably, current pulsing plays a significant role in welding, affecting the cooling rate and heat input. It disrupts fluid motion, amplifying the existing convective forces in the weld pool. Periodic fluctuations in arc current lead to corresponding variations in the arc forces acting on the weld pool or weld puddle [15]. In the current study, it was evident that micro-pulsing (125 Hz and 250 Hz frequency) in the arc current induces stronger vibrations in the weld pool compared to un-pulsed conditions. These vibrations may cause the re-melting and fragmentation of growing dendrites (Figure 7), while the weld pool turbulence transports the dendritic fragments to the S/L interface. This movement induces undercooling ahead of the S/L interface, allowing more nucleation sites to persist, disrupting columnar growth, and encouraging finer grain formation in the FZ [9,24].

The growth and fragmentation of dendrites occur predominantly in the molten pool rather than the mushy zone because the molten pool provides the necessary conditions for dendrite formation, such as a fully liquid environment with super-cooling and adequate thermal gradients. In the molten pool, the temperature is lower than the liquidus temperature, allowing nucleation and growth of dendrites to begin as the liquid metal cools. In contrast, the mushy zone consists of both solid and liquid phases, where dendritic growth is restricted due to limited space and lower mobility of atoms, making the molten pool more favorable for dendrite growth and fragmentation. The fluid flow in the molten pool also enhances dendrite fragmentation by breaking off growing dendrite arms, which can be transported to other areas before solidification completes [41].

Comparing the macrostructures and microstructures of welds produced under un-pulsed and micro-pulsed conditions, it is evident that the grains within the FZ are significantly finer in pulsed welding. Moreover, increasing the pulse frequency leads to even finer grain sizes. The average grain size in micro-pulsed welds with high frequencies of 125 Hz and 250 Hz was found to be 303 µm and 266 µm, respectively, compared to 384 µm in un-pulsed welds (Table 3). These results confirm that pulsed welding leads to grain refinement, especially at higher frequencies. The observed grain refinement in pulsed welding conditions can be attributed to arc constriction and increased convective forces within the weld pool. Arc constriction, more prevalent in pulsed welding, can concentrate heat energy and promote grain refinement. Additionally, the increased convective forces within the weld pool can contribute to grain refinement. Moreover, the grain size decreased as the frequency increased from 125 Hz to 250 Hz. This is likely due to the higher arc force at the elevated frequency, resulting in more vital convective forces and enhanced grain refinement.

3.5. Post Weld Heat Treatment

Figure 8a,b show the FZ microstructure of welds prepared in un-pulsed and micro-pulsed conditions at 250 Hz pulse frequency, after being subjected to PWHT. It is observed that there is variation in prior β grain size after PWHT. After the post-weld aging material tries to attain thermodynamic equilibrium by the formation of β from martensitic α′. Vanadium will be absorbed from the surrounding α′ by β as it nucleates. This results in the conversion of martensitic α′ to α. This leads to coarsening the α plates and increasing the α platelet width. The analysis is consistent with the study made by Babu et al. [9] during GTA welding of Ti-64 alloy. Additionally, PWHT reduces dislocation density, which further contributes to changes in the weld’s microstructure [18]. Despite these transformations, the microstructure retains a needle-like morphology reminiscent of the as-welded condition. The acicular (needle-like) appearance of the α and α′ phases makes it challenging to differentiate between the two, as their morphologies are quite similar.

4. Mechanical Properties

4.1. Hardness

The hardness profiles for all the welds in PWHT and as-welded samples are shown in Figure 9. The microhardness is measured from the center of the weld towards the BM on either side of the weld. The hardness profiles of the GTA welds, in both the as-welded and PWHT conditions, exhibit a progressive increase in hardness from the base metal (BM) through the HAZ to the FZ. The average FZ hardness is 350–380 HV, while the BM has a hardness of 305 ± 4 HV. The increased hardness in the FZ results from the formation of α′ in the weld. Further, the welds prepared in the micro-pulsed condition at 125 Hz frequency have exhibited a higher FZ hardness than the un-pulsed weldment, attributing to a refinement in the prior β-grain size. A further increase in FZ hardness is observed with an increase in frequency from 125 Hz to 250 Hz, owing to the further grain refinement due to an increase in convective forces and arc forces. The grain refinement leads to stronger material due to the enhanced grain boundary strengthening effect. The fluctuating heat input during micro-pulsing also promotes the formation of harder phases, such as α′ martensite, further increasing FZ hardness. On the other hand, the elevated hardness in the HAZ relative to the BM is caused by the formation of α′ martensite in the as-welded state, regardless of the welding speed.

A comparable hardness trend is seen in welds that undergo PWHT across all welding conditions. The hardness of the as-welded weldments is higher than that of those in the PWHT condition. The reduced hardness in the PWHT samples results from the breakdown of α′ martensite into a diffusional α and β lamellar structure. Pulsed welds may retain slightly higher hardness than unpulsed welds due to their finer initial microstructure. The grain size of all weldments shows no change after PWHT compared to the as-welded condition. The reduction in weld metal hardness following PWHT aligns with the findings of Reda et al. [42]. Additionally, pulsing reduces peak thermal stresses, allowing PWHT to enhance further the weld’s mechanical properties with improved stress relief and uniformity [18].

4.2. Tensile Properties and Fractography

Figure 10 shows the tensile test results for the pulsed and un-pulsed weldments in the PWHT and as-welded conditions. The percentage of tensile elongation and UTS for different processing conditions are compared in Table 4. Among the weldments in the as-welded samples, the welds prepared with the micro-pulsed condition at 250 Hz exhibited the highest ductility and strength (YS: 1020 ± 5 MPa, UTS: 1089 ± 5 MPa, %El: 18 ± 0.5), while the un-pulsed weldment has shown lower ductility and strength (YS: 980 ± 4 MPa, UTS: 1014 ± 5 MPa, %El: 14 ± 0.5). The increase in strength is due to the narrowing of prior β grains and the acicular structure in the FZ. The grain refinement mainly because of micro-pulsing during welding enhances both the welded material’s strength and ductility and promotes a more uniform microstructure. The rapid thermal cycling caused by pulsing breaks down larger grains into finer ones, which increases strength through grain boundary strengthening. Additionally, pulsing helps achieve a more controlled phase transformation, resulting in a balance between strength and ductility, especially in the fusion and heat-affected zones. These results align with those documented by Li et al. [43].

The weldments subjected to PWHT have shown an increase in ductility (%Elongation) and reduction in strength compared to the respective weldments in the as-welded condition (Figure 10). The reduced strength in the PWHT samples results from the transformation of α′ martensite into diffusional α and β phases. Among the weldments in the PWHT, the welds prepared without pulsing exhibited lower ductility and strength (YS: 944 ± 6 MPa, UTS: 967 ± 6 MPa, %El: 22 ± 0.4) compared to that of the welds made of micro-pulsing at 125 Hz (YS: 957 ± 5 MPa, UTS: 989 ± 6 MPa, %El: 23 ± 0.5) and 250 Hz (YS: 972 ± 4 MPa, UTS: 998 ± 4 MPa, %El: 23 ± 0.5) in the PWHT condition. This decrease in strength is attributed to the diffusional phase of α + β decomposed from α′ and coarse prior β grains in FZ of un-pulsed as compared to micro-pulsed condition. Additionally, the uniform grain structure from pulsing ensures that strength is distributed more evenly throughout the weld, further contributing to improved mechanical properties after heat treatment. The findings align well with the observations made by Thomas et al. [44] regarding Ti-6Al-4V welds created through EBW.

All the as-welded and PWHT samples exhibited BM failure, indicating stronger FZ, attributed to acicular martensitic α′ formation in the FZ. For instance, Figure 11a,b depict the tensile fracture surfaces of Ti-64 GTA welds made with micro-pulsed at 125 Hz frequency in as-welded and PWHT conditions, respectively. The fracture morphology for both samples reveals equiaxed fine dimples through the microstructure, indicating a ductile type of fracture, suggesting a BM failure. A few tear edges are also observed in the fractured surfaces. However, the ductility of both specimens is lower than the BM, attributed to the coarse prior β-grain size and large acicular intergranular microstructure in FZ. These results are consistent with the research carried out by Wang et al. [6].

5. Conclusions

Ti-64 GTA welds were successfully produced under un-pulsed, 125 Hz micro-pulsed, and 250 Hz micro-pulsed conditions. The microstructural and mechanical properties of both the as-welded and PWHT weldments were evaluated, and the key findings are summarized below:

• Micro-pulsing significantly influences the weld bead geometry of GTA-welded Ti-64 alloy. As the micro-pulsing frequency increases, the FZ and HAZ’s average width decreases, attributed to the reduced heat input at higher frequencies;
• The micro-pulsing during the GTA welding has resulted in mild grain refinement of prior β grains in the HAZ and FZ. This is due to dendrite fragmentation, which induces undercooling ahead of the S/L interface, allowing more nucleation sites to survive, disrupting columnar growth, and promoting finer grains in the FZ;
• The microstructure of the as-welded FZ in Ti-64 welds, regardless of the welding conditions, features an acicular martensitic α′ phase that forms due to the elevated thermal cycles experienced during welding. However, after PWHT, the microstructure shows coarsening of the α phase due to the transformation of martensitic α′ into α and β phases through a diffusional process;
• In the as-welded condition, the hardness, ductility, and UTS have shown a significant improvement in the micro-pulsed condition (125 Hz and 250 Hz) compared to the un-pulsed condition due to grain refinement in the FZ;
• Welds subjected to PWHT exhibited reduced strength but improved ductility in all samples due to the transformation of martensitic α′ into α and β phases;
• During the tensile test, all the samples failed in BM, indicating the higher strength of FZ than BM. The dimples in the fracture surface indicate ductile failure.