Diode Area Melting of Ti6Al4V: Effects of Multi-Laser Processing on Microstructure and Residual Stress

Technical Summary: Diode Area Melting of Ti6Al4V – Effects of Multi-Laser Processing on Microstructure and Residual Stress

Problem Statement
Laser Powder Bed Fusion (LPBF) is a dominant additive manufacturing technique for metal components, yet it is constrained by rapid thermal cycles that generate steep temperature gradients and high residual stresses. These conditions often lead to warping, cracking, and dimensional instability in components such as Ti6Al4V. While LPBF produces fine microstructures beneficial for strength, the associated thermal history limits the control over solidification behavior. Diode Area Melting (DAM) has emerged as a novel alternative, utilizing arrays of low-power diode lasers (808 nm) rather than a single high-power fiber laser reflected by a galvo system. Although previous studies have demonstrated DAM's feasibility for various metals and its potential to reduce cooling rates, there remains a limited understanding of how multi-laser interactions specifically influence thermal penetration, solidification behavior, and residual stress development in Ti6Al4V.

Methodology
This study utilized a custom DAM system developed at the University of Sheffield, featuring a gantry-based optical head capable of integrating up to 100 low-power diode lasers. For this investigation, single-layer Ti6Al4V samples were fabricated using 3, 4, and 5 aligned diode lasers (808 nm wavelength, 5 W per laser). The experimental design systematically varied the number of activated lasers and scanning speeds (100, 150, 200, and 250 mm/min) while maintaining a constant hatch spacing of 200 µm and a powder bed thickness of 1 mm.

Characterization methods included:

• Surface Analysis: 3D profilometry to measure surface roughness ($R_a$).
• Microstructural Analysis: Optical microscopy (OM) and Scanning Electron Microscopy (SEM) for melt pool geometry; Electron Backscattered Diffraction (EBSD) for grain structure, phase identification, and $\alpha$-lath spacing; and X-Ray Diffraction (XRD) for phase constitution.
• Residual Stress Analysis: XRD using the $\sin^2\psi$ method to measure near-surface residual stresses in both the scanning ($\sigma_{xx}$) and perpendicular ($\sigma_{yy}$) directions.
• Simulation: Computational Fluid Dynamics (CFD) using FLOW-3D™ software to estimate cooling rates under the specific processing conditions.

Key Results

1. Surface Morphology: The DAM process produced significantly smoother surfaces compared to typical as-built LPBF Ti6Al4V (which often exhibits $R_a$ of 10–15 µm). The 5-laser configuration yielded the smoothest surfaces ($R_a$ = 1.2–1.3 µm), attributed to a higher overlap ratio (approx. 45.9%) between adjacent melt tracks, which promoted partial remelting and suppressed balling. In contrast, the 3-laser configuration resulted in rougher surfaces ($R_a$ = 5.4–6.6 µm) due to lower overlap and thermal discontinuities.
2. Melt Pool and Thermal Penetration: Increasing the number of lasers increased both melt pool depth and Heat Affected Zone (HAZ) depth. The 5-laser configuration at 100 mm/min achieved a melt pool depth of ~108 µm and an HAZ depth of ~475 µm. Higher scanning speeds reduced thermal penetration, leading to shallower pools and increased instability.
3. Microstructure and Phase: The DAM process resulted in a coarser $\alpha$-lath structure (0.8–1.3 µm) compared to LPBF (typically <0.5 µm). This coarsening is linked to the significantly lower cooling rates observed in DAM. Crucially, detectable $\beta$-phase retention was observed in 4- and 5-laser specimens, particularly at lower scanning speeds. This contrasts with LPBF, where rapid cooling typically suppresses $\beta$-phase retention in favor of martensitic $\alpha'$. The presence of $\beta$-phase was attributed to slower cooling rates allowing for diffusion-controlled transformations.
4. Cooling Rates: Simulated cooling rates ranged from 2,211 K/s to 8,400 K/s, which is substantially lower than the $10^4$–$10^7$ K/s typical of LPBF. Cooling rates decreased with an increasing number of lasers (due to heat accumulation) and increased with scanning speed.
5. Residual Stress: Residual tensile stresses in DAM samples were markedly lower than those reported for LPBF. The maximum tensile stress measured was 138.2 MPa (in the 3-laser, 250 mm/min condition), compared to LPBF values that can exceed 688 MPa. Stress levels decreased as the number of lasers increased (5-laser samples showed the lowest stress) and increased with scanning speed. The stress anisotropy followed LPBF trends, with $\sigma_{xx}$ (scanning direction) consistently higher than $\sigma_{yy}$.

Significance and Claims
The authors claim that the DAM process offers improved control over solidification behavior and stress evolution in Ti6Al4V components. By employing multiple low-power diode lasers, DAM achieves significantly lower cooling rates, which facilitates the formation of coarser $\alpha$-laths and the retention of $\beta$-phases, deviating from the martensitic structures typical of LPBF.

The study posits that these lower cooling rates are the primary mechanism for the substantial reduction in residual tensile stress. The slower thermal cycles allow for stress relaxation via plastic deformation at elevated temperatures, thereby mitigating the high residual stresses that often cause warping and cracking in LPBF parts. Furthermore, the improved surface finish and reduced stress levels suggest that DAM has the potential to extend the service lifetime of components by suppressing fatigue crack initiation.

The authors maintain a modest scope, noting that the current findings are limited to single-layer specimens where stress accumulation from inter-layer thermal cycling is minimized. They conclude that while the process shows promise for reliable fabrication, future work must extend to multi-layer specimens to fully understand cumulative thermal effects in bulk component production.