Laser Additive Manufacturing (LAM) involves complex laser-matter interactions characterized by extreme non-equilibrium thermodynamic processes. During this process, localized high-energy heat input and rapid cooling establish a highly non-uniform temperature field within the material. The interplay between severe temperature gradients and the material's thermal expansion and contraction behavior inevitably leads to the accumulation of significant residual stresses within the fabricated component. The presence of these internal stresses not only induces component warping and compromises geometric precision but also acts as a driving force for micro-crack propagation. This severely degrades fatigue life, fracture toughness, and corrosion resistance, thereby posing potential risks to structural safety during service. Consequently, the effective mitigation of residual stress has become a critical bottleneck restricting the large-scale application of LAM technology in high-end manufacturing sectors such as aerospace and biomedicine. This paper first elucidates the fundamental nature, classification system, and thermo-mechanical coupling origins of residual stress. It then systematically reviews multi-dimensional characterization techniques and the effects of key process parameters. Finally, integrating recent research advancements, it provides an in-depth analysis of the mechanisms, merits, limitations, and future trends of existing control strategies.

A physics data fusion driven strategy for estimating the multiaxial fatigue life of titanium alloys is proposed. Using TC4 titanium alloy as the research subject, multiaxial fatigue experiments are performed and a physical framework is created by adopting von-Mises equivalent strain as the damage indicator. A Backpropagation (BP) neural network is then developed in which phase difference, normal strain amplitude, shear strain amplitude, normal stress amplitude, shear stress amplitude, and von-Mises equivalent strain serve as inputs, while the life predicted by the physical framework serves as the output. The connection weights between the hidden layer and the output layer of the neural network are subsequently refined through measured fatigue life data to form a parameter fusion model. By using the measured fatigue life as the reference, the predictive performance of the physical framework, the neural network, and the parameter fusion scheme is evaluated. The results reveal that the parameter fusion architecture achieves a substantial improvement in predictive accuracy compared with the single physical framework or the BP neural network, with every predicted life value remaining inside the twofold scatter zone of the experimental results. This approach integrates the foundational understanding of physical behavior with the adaptive learning abilities of data centered models, provides a promising route for reliable multiaxial fatigue life estimation of titanium alloys under limited sample conditions, and contributes theoretical value for fatigue resistant design of titanium alloy structural elements.

Accurate localization of simulated acoustic emission (AE) sources representing weld cracks is critical for ensuring the structural integrity and operational safety of welded components. Traditional methods such as the Least Squares Method (LSM) often yield significant positioning errors, while the Simplex Algorithm (SA) is highly sensitive to initial values and converges slowly. This paper proposes a hybrid LSM-SA algorithm that leverages the analytical initial estimate provided by LSM and the local refinement capability of SA. The method employs an autoregressive (AR) model for AE signals from simulated crack sources, using the Akaike Information Criterion (AIC) to pick arrival times at each sensor, from which time differences across sensors are computed. An initial source location estimate from LSM serves as the starting point for SA, enabling refined optimization and accurate crack source coordinates. Lead-break experiments conducted on an aluminum alloy plate at a single simulated crack location demonstrate that the proposed LSM-SA hybrid algorithm achieves a 42.2% reduction in localization error (from 11.22 mm to 6.48 mm) compared to the standalone LSM, while exhibiting a 55.6% decrease in the number of required optimization iterations (from 9 to 4) relative to the SA alone. By improving both accuracy and computational efficiency, this method can support online weld crack monitoring via acoustic emission.