Technology
Tech Breakthrough
University of Wisconsin Studies Acoustic Signal to Predict PBF-LB Mechanical Properties
2026-05-21
A research team from the University of Wisconsin has used machine learning to establish relationships between acoustic signals in the laser powder bed fusion (PBF-LB) process and the mechanical properties of parts. The study printed 100 tensile specimens of CoCrFeMnNi high-entropy alloy under 13 process conditions, training a model to predict mechanical properties from acoustic features. By combining process parameters with acoustic information, the prediction accuracy for yield strength improved by 18%, and for elongation at break by 10%. The research found that acoustic signals in the 12 to 16 kHz frequency range correlate with material ductility, where lower power spectral density corresponds to higher ductility. This method could replace destructive testing by directly assessing part performance through acoustic signatures.
📰 News Source
This article was translated from a Chinese source.
Source URL not available
Show Chinese original
威斯康星大学研究声学信号预测PBF-LB力学性能
美国威斯康星大学的研究团队利用机器学习建立激光粉末床熔融工艺中声学信号与零件机械性能之间的关系。该研究在13种工艺条件下打印100个CoCrFeMnNi高熵合金拉伸试样，通过声学特征训练模型预测力学性能。结合工艺与声学信息后，屈服强度预测准确率提升18%，断裂延伸率提升10%。研究发现12至16kHz频段的声学信号与材料延展性相关，较低功率谱密度对应更高延展性。该方法可替代破坏性检测，直接通过声学特征评估零件性能。