Page 35 - MSAM-4-3

Materials Science in Additive Manufacturing Numerical simulation of plasma WAAM for Ti-6Al-4V

observations. The final values give an arc efficiency of 0.4, is represented as a flat surface, as no wire feedstock is
an emissivity of 0.7, a convective heat transfer coefficient deposited. In addition, arc pressure, which exerts a force on
of 8 W/(m²⋅K), and a contact heat transfer coefficient the melt pool, influences the weld pool shape by directing
of 10 W/(m²⋅K). These parameters are assumed to be the flow of liquid metal. Thus, melt pool dynamics play a
temperature-independent throughout the simulation. crucial role in the melting process, influencing the weld
pool geometry, heat flow, temperature, and temperature
3.2.1. Double-ellipsoid heat source validation gradients. In Figure 5C, the simulation overestimates the
32
Accurate modeling of the heat source is critical for realistic penetration depth of the weld pool. This effect is attributed
simulations of welding processes. The calibration process to the larger-than-expected pre-heating fusion zone, which
ensures that the simulated weld pool dimensions match delivers excess energy to the baseplate, causing deeper
the experimental measurements. During calibration, key penetration than experimentally observed.
parameters such as weld penetration and weld width were The ratio of melt width to melt depth φ = 2b/d is an
systematically fine-tuned to achieve the best agreement with indicator of the evolution of the weld pool geometry, with
experimental results. A visual comparison of experimental higher values corresponding to wider and shallower pools.
and simulated weld profiles is shown in Figure 5.
Experimental melt pool aspect ratios from Table 4
The fusion zone boundary in the simulation is defined show that pre-heating results have significantly larger
by the solidification temperature of Ti6Al4V, which is ratios than single beads. Unlike single beads, pre-heating
1,550°C. As shown in Figure 5A, the experimental fusion does not involve feedstock material deposition, and melt
zone of the single bead is closely reproduced, although the pool dynamics are primarily governed by forces based on
simulated penetration depth is slightly underestimated.
However, for the pre-heating pass (Figure 5B), the surface tension gradients in the melt. This result in lateral
metal transport and the formation of a wide, shallow melt
simulation overestimated the weld pool size compared to pool (Figure 5B). In contrast, the melt pool ratio of the
the experimental measurements. One possible explanation
is the omission of melt pool flow effects in the simulation single bead with preceding pre-heating is slightly lower
that affect heat dissipation and melt zone morphology, than that of the single bead alone, indicating greater weld
penetration. In this case, the numerical simulation will
leading to numerical predictions that are different from
those experimentally measured. Visual assessment of overestimate the weld penetration, which will affect the
the micrograph in Figure 5B reveals that the liquid calculated aspect ratio. Table 4 confirms that increasing
heat input leads to deeper melt zones, which reduces the
metal is drawn downward in the peripheral areas while
being forced upward toward the center. In contrast, the melt pool aspect ratio.
numerical model for pre-heating does not account for Overall, the simulated weld pool geometries and
any weld pool geometry; instead, the pre-heating pass corresponding melt pool ratios are in fair agreement with

A B

C

Figure 5. Comparison of calculated and experimental micrographs. (A) Single bead. (B) Pre-heating. (C) Single bead with preceding pre-heating. Scale
bar: 2,000 µm

Volume 4 Issue 3 (2025) 8 doi: 10.36922/MSAM025140021