Advanced Laser Weld Brazing of Zinc-Coated Automotive Steels: Process Optimization for AI-Enhanced Inspection, Tailored Intermetallic Formation, and the Evolution of Mechanical Behavior

Abstract

Laser weld brazing (LWB) is a key joining process in automotive manufacturing, offering minimal substrate melting, reduced Zinc (Zn) coating burn-off, and superior joint appearance. However, challenges such as sub-surface defect formation and intermetallic compound (IMC) embrittlement hinder its optimization and broader adoption. Due to LWB's relative novelty for Zn-coated steels, limited research has explored brazing defects, segregation behavior, and IMC formation, as well as their direct impact on mechanical performance. The potential for maximizing braze joint mechanical performance is limited. Even with the elimination of interfacial premature failures caused by the brittle IMC layer, the extent of mechanical performance improvement in defect-free joints remains confined to the strength of the brazing filler material. Moreover, the role of IMCs and any second phases forming within the braze matrix in either degrading or improving mechanical performance within the literature is inconclusive. This work systematically investigates LWB across Zn-coated steel grades using Si-bronze filler wire, advancing the understanding of laser-wire-substrate interactions that dictate microstructure and mechanical behavior, to maximize mechanical performance of LWB joints. Additionally, it establishes a novel LWB optimization approach that enables modification of microstructure of LWB joints as well as improvement of an in-line non-destructive testing (NDT) method assisted by artificial intelligence (AI) for real-time defect inspection and precise braze geometry measurement. Comparative analysis with gas metal arc brazing (GMAB) confirms that LWB significantly enhances wettability of molten filler by refining the surface morphology of substrates and alters elemental segregation. This study introduces a novel wire-adjusted heat input (HI) approach, a systematic optimization strategy that directly correlates developed HI-related parameters with defect formation, overcoming inconsistencies observed in conventional nominal heat input methods. This approach enables accurate distinction between defect-free and defective seams, including lack of adhesion (LoA), pores, base metal melting (BMM), and non-conforming geometry. Process window validation through three-dimensional (3D) X-ray micro-computed tomography (μCT) using semi-automated pore segmentation and statistical analysis of pore evolution, supports the robustness of the optimization strategy. The braze joints made within the process window successfully enabled acquisition of reliable UT signal feedback for defect detection. The two-dimensional (2D) slices of μCT data, enriched AI training data base, effectively contributed into AI predictions of the seam geometry from the UT data. Another major breakthrough of this study is developing a novel IMC category, termed surrounded interface-IMCs (SI-IMCs), formed under controlled elevated HI, distinct from conventional interface-IMCs (I-IMCs). A critical HIRelative threshold of 32 J/mm is identified for defect-free brazing in roof joints of GI-IF steels, with an additional 12.44 J/mm required to promote SI-IMC formation, occupying up to 38.2 ± 16.9% of the braze cross-sectional area. These SI-IMCs, consisting of a shell-like Fe-Si layer and a (Fe-rich)-Cu eutectic phase, enhance mechanical properties. Increasing SI-IMC area fraction from 1.2 ± 2.4% to 38.2 ± 16.9% results in a 14% increase in tensile peak load, a 350% increase in displacement, and improvement of joint toughness by 525%. The complex geometry of roof joint coupons, combined with the low strength and high ductility of galvanized coated interstitial free (GI-IF) steel substrates, makes it difficult to fully attribute the improved mechanical performance to the presence of SI-IMCs. To fully elucidate the contribution of SI-IMCs to joint mechanical performance, this study conducts the first comprehensive investigation into a wide range of SI-IMC structures formed at different HIRelative levels during LWB in a lap joint configuration for GI-coated Gen 3 Q&P980 steel. It provides entirely novel insights into characterization of SI-IMC structures, the Cu-rich braze matrix, and how they impact mechanical performance. SI-IMCs formed at lower ∆HIRelative (2.3–7.5 J/mm) improve mechanical performance, increasing peak load by 91.7%, displacement by 319.8%, and toughness by 1200%, whereas those formed at higher ∆HIRelative (13–17.5 J/mm) degrade properties. Systematic elemental and texture analyses identify five SI-IMC types (α, β, γ, δ), with fine α-, β-, and γ-SI-IMCs delaying failure, while coarse α- and δ-SI-IMCs accelerate it. A strong {100} texture alignment between SI-IMCs and the Cu-rich braze matrix enhances strain accommodation and crack resistance. The newly introduced GND.Ratio (Cu/Fe) quantifies dislocation density, revealing that when GND.Ratio (Cu/Fe) > 0.67, SI-IMC-induced twinning activates, locally rotating the Cu-rich matrix orientation to <112> and forming twins near crack paths. This transition shifts fracture behavior from brittle to ductile, enhancing mechanical performance. These findings redefine the role of IMCs and second phases that are formed in the braze interior, demonstrating their potential to optimize strain distribution, strain-hardening behavior, and crack propagation. Tailoring SI-IMC structures strategically can further enhance the mechanical performance and reliability of multi-material automotive structures.