Microstructure Evolution in Double-Sided AC TIG Welding of Cast AZ91D Magnesium Alloy

Literature Overview

The study by Guo Qiang, You Guoqiang, and Wu Jingting from Chongqing University and the National Magnesium Alloy Materials Engineering Technology Research Center addresses a critical but often overlooked challenge in lightweight structural engineering: the weldability of cast AZ91D magnesium alloy using double-sided AC TIG (GTAW) welding. Published in the journal Hot Working Technology in 2011, this work investigates 3 mm thick cast-state AZ91D plates welded with a root-side current of 80, 90, and 100 A while maintaining a consistent face-side current of 120 A. The research is supported by Chongqing Municipal Science and Technology Project (CSTC2009AB4010) and Guangdong Provincial Science and Technology Project (2008A090300004), underscoring the industrial relevance of magnesium alloy joining in China's growing lightweight vehicle and aerospace sectors.

Core Technical Points

AZ91D is a cast magnesium alloy containing approximately 9 wt% aluminum and 1 wt% zinc, with the balance magnesium. The equilibrium microstructure consists of an α-Mg matrix with interdendritic β-Mg₁₇Al₁₂ phases. During TIG welding, the high thermal input inevitably causes localized melting and rapid solidification, fundamentally altering the phase morphology and distribution. The authors employed AC TIG welding to leverage the cathodic cleaning action on the backside, which is essential for magnesium alloys due to the rapid formation of a refractory MgO oxide layer.

The key finding is that increasing the backside welding current from 80 A to 100 A progressively reduces the network-like (reticular) β-Mg₁₇Al₁₂ phase while increasing the proportion of particulate (spheroidal) β-Mg₁₇Al₁₂ phase. This is a significant observation because network-type intermetallics act as crack initiation sites and severely degrade the fracture toughness and fatigue life of the weld joint.

Welding Parameters and Observations

Microstructural Interpretation

From a metallurgical perspective, the reduction in network β-Mg₁₇Al₁₂ at higher backside currents can be attributed to enhanced thermal input on the backside, which promotes more complete dissolution of the interdendritic phase during solidification. The subsequent cooling rate from the backside is governed by the thermal mass of the backing plate or the air gap, and higher currents create a wider, more fully melted weld zone where the β-phase has more opportunity to re-nucleate as isolated particles rather than forming continuous interdendritic networks. This phenomenon is analogous to what is observed in aluminum alloys where higher heat input promotes grain refinement through increased nucleation sites, although the mechanisms differ due to the different phase equilibria in Mg-Al systems.

However, the authors also note that the weld surface profile is poor, with visible defects on the surface. This is a well-known challenge in TIG welding of magnesium alloys, where the combination of high thermal conductivity, low melting point, and rapid oxide reformation creates a narrow process window for achieving sound weld bead geometry.

Engineering Practice Implications

For engineers working with magnesium alloy components, this study provides several actionable insights:

• Backside current optimization is critical. The choice of backside current should not be arbitrary but should be systematically varied to optimize the β-phase morphology. A current of 100 A on the backside (approximately 83% of the face current) appears to provide the most favorable balance of phase distribution.
• Surface quality remains a concern. The reported poor surface profile suggests that additional measures such as backing gas shielding, tighter current control, or even pulsed TIG welding may be necessary for production-quality welds.
• Thermal management is paramount. Magnesium alloys are sensitive to excessive heat input, which can lead to grain coarsening in the heat-affected zone (HAZ) and potential cracking. The AC component of the TIG process must be carefully balanced to provide adequate cleaning without excessive thermal damage.

Defect Analysis and Countermeasures

Study Insights and Reflections

Having worked extensively with metallic joining processes over decades, I find this study particularly valuable for its systematic approach to backside current variation. In practice, many engineers focus exclusively on the face-side parameters and treat the backside as a passive process. This work demonstrates that the backside current is not merely a means to achieve penetration but is a powerful metallurgical control lever that directly influences phase morphology and, by extension, mechanical performance.

One area where I would urge further investigation is the quantitative correlation between β-phase morphology and mechanical properties. The study focuses primarily on microstructural characterization but does not report tensile strength, fatigue life, or fracture toughness data. For engineering qualification purposes, such data are essential. Additionally, the lack of specified welding speed and gas flow rate values limits the reproducibility of the results. In my experience, welding speed is often the single most influential parameter in determining weld bead geometry and heat input, and its omission is a notable gap.

The findings also raise an important question about the thermal history of the backside weld. In double-sided welding without a backing plate, the backside weld cools primarily by radiation and convection, which is significantly slower than air-cooled solidification. This slower cooling rate may actually contribute to the reduction in network β-phase by allowing more time for diffusion-controlled precipitation. Future work should consider the use of water-cooled backing plates to study the effect of rapid solidification on phase morphology, which could potentially produce even finer particulate β-phase distributions.

For magnesium alloy applications in automotive lightweighting and aerospace structures, where fatigue performance is often the governing design criterion, the elimination of network β-phase is of paramount importance. Network-type intermetallics act as stress concentrators and crack initiation sites, and their reduction can lead to substantial improvements in fatigue life. This study provides a foundation for developing welding procedures that produce fatigue-resistant magnesium alloy welds, which is a critical need in the industry.