Real-time monitoring of raster temperature distribution and width anomalies in fused filament fabrication process  

作　　者：Feng Li Zhong-Hua Yu Hao Li Zhen-Sheng Yang Qing-Shun Kong Jie Tang 

机构地区：[1]The State Key Laboratory of Fluid Power Transmission and Control,College of Mechanical Engineering,Zhejiang University,Hangzhou,310027,People’s Republic of China

出　　处：《Advances in Manufacturing》2022年第4期571-582,共12页先进制造进展（英文版）

基　　金：Our work is funded by the National Natural Science Foundation of China(Grant Nos.51675481 and 51905335);the Shanghai Science and Technology Committee Research Project(Grant No.19040501500).

摘　　要：The aim of this study is to monitor the raster temperature distribution and width anomalies in a fused filament fabrication(FFF)process by an infrared(IR)array sensor.To achieve this goal,two experiments were conducted on a desktop FFF machine.For the first experiment,three normal samples with different raster widths were fabricated,and thermal images of the newly deposited rasters were collected during the process.To process the low-resolution images,a segmentation-based image processing method was proposed.The temperature distributions along the horizontal direction of the raster section and along the raster length were obtained.The temperature features that could indicate the raster widths were extracted and then fed to recognition models for training and testing.The classification performance of the models were evaluated based on the F-score.The models with high F1-scores could be used to recognise width anomalies online.For the second experiment,an abnormal sample with raster width anomalies was fabricated.The temperature features were extracted from the collected experimental data.The obtained features were then fed to the built and evaluated models to recognise the width anomalies online.The effectiveness of the monitoring method was validated by comparing the recognition results with the actual optical images.The support vector machine(SVM)and k-nearest neighbour(KNN)were adopted to build the recognition models.The F1-score and online recognition results of the models were compared.The comparison study shows that SVM is more suitable for our situation than KNN.A method for monitoring the temperature distribution and width anomalies of the FFF raster is provided in this paper.To the best of the authors’knowledge,this is the first study to explore the actual temperature distribution along the horizontal direction of the raster section,and the first study to monitor the width anomalies of the raster in the FFF process.

关 键 词：Fused filament fabrication(FFF) Process monitoring INFRARED Temperature distribution Width anomalies 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]