激光器
因科镍合金625
失真(音乐)
材料科学
热的
计算机科学
融合
光栅图形
螺旋(铁路)
光学
机械工程
光电子学
人工智能
复合材料
工程类
微观结构
物理
哲学
CMOS芯片
气象学
放大器
语言学
作者
Amit Kumar Ball,Amrita Basak
标识
DOI:10.1016/j.cjmeam.2023.100103
摘要
Metal additive manufacturing, especially laser powder bed fusion (L-PBF), is increasingly being used to fabricate complex parts with fine features. Emerging L-PBF systems have large build volumes and several lasers that operate simultaneously. Hence, they can produce large and complex parts at reduced costs and short build times. However, the thermal distortion remains a critical challenge. Hence, a thorough understanding of the impact of multiple lasers on part distortion in multi-laser PBF (ML-PBF) is imperative. Although experimental investigation is possible, a more conducive approach is to design and create suitable predictive models to understand the impact of multiple lasers consolidating a part into layers. To fulfill this goal, in this study, a commercially available and widely used thermo-mechanical model, Netfabb®, was modified to include the effects of multiple lasers for different scan patterns such as raster, spiral, and Hilbert. Thereafter, a computational model was developed to determine the influence of single vs. multiple lasers on the temperature distribution and thermal distortion of complex scan patterns. The results show that the thermal distortion is minimal for the spiral scan pattern. Additionally, multiple lasers were found to decrease the build time (as expected) while maintaining or reducing the thermal distortion compared with their single-laser counterparts for all scan patterns (except Hilbert scan pattern). Therefore, the newly developed ML-PBF predictive model is capable of providing critical insights into the effects of using multiple lasers, thereby opening new possibilities for the faster production of complex parts. In the future, small-scale computational models will be expanded to include large-scale parts, and probabilistic models will be developed to establish correlations.
科研通智能强力驱动
Strongly Powered by AbleSci AI