Naoki Iinuma, Boshi Chen, Tappei Kawasato, Y. Kakinuma
4K and 8K technologies are attracting attention in optical industries. The most important mechanical element to enhance the imaging performance is the aspherical lens requiring higher surface quality and higher form accuracy. Currently, the production process of optical lenses consists of brittle-mode grinding and pro-longed polishing process, which play a role of shaping the form and producing the fine surface, respectively. However, this process is not considered to be suitable for manufacturing such higher-quality lenses for 4K and 8K imaging devices because a required form accuracy could not be ensured, and the polishing time gets longer. To enhance the form accuracy and production efficiency, application of ductile-mode grinding is expected to reduce polishing amount. However, the shape error generated by the ductile mode grinding is not clear. Therefore, the purpose of this research is to analyze the relation between the shape error and the grinding force estimated from motor-current in the grinding machine. The motor-current acquisition system in all translational axes and the work spindle is constructed and implemented into a 4-axis ultra-precision aspherical machine. The grinding force in each axis is derived by subtracting the motor current during non-grinding previously obtained in air-grinding test from the current during grinding. Firstly, the behavior of the motor current in each axis is investigated from the viewpoint of repeatability and position dependency. While the periodic fluctuation of the motor current affected by the influence of permanent magnet in the linear motor is confirmed, it shows high repeatability at each position. This result indicated that grinding force is easily calculated from the motor current with less uncertainty. Then, influence of grinding condition in the range of ductile mode grinding on the shape error is analyzed by monitoring the motor current. Toward the outside of the workpiece, the shape error gradually increases with the increase of motor current, which means larger grinding force at the outer side causes the deformation of the resin grinding wheel.
{"title":"Shape Error Analysis in Ultra-Precision Grinding of Optical Glass by Using Motor-Current-Based Grinding Force Monitoring","authors":"Naoki Iinuma, Boshi Chen, Tappei Kawasato, Y. Kakinuma","doi":"10.1115/msec2022-85472","DOIUrl":"https://doi.org/10.1115/msec2022-85472","url":null,"abstract":"\u0000 4K and 8K technologies are attracting attention in optical industries. The most important mechanical element to enhance the imaging performance is the aspherical lens requiring higher surface quality and higher form accuracy. Currently, the production process of optical lenses consists of brittle-mode grinding and pro-longed polishing process, which play a role of shaping the form and producing the fine surface, respectively. However, this process is not considered to be suitable for manufacturing such higher-quality lenses for 4K and 8K imaging devices because a required form accuracy could not be ensured, and the polishing time gets longer. To enhance the form accuracy and production efficiency, application of ductile-mode grinding is expected to reduce polishing amount. However, the shape error generated by the ductile mode grinding is not clear. Therefore, the purpose of this research is to analyze the relation between the shape error and the grinding force estimated from motor-current in the grinding machine. The motor-current acquisition system in all translational axes and the work spindle is constructed and implemented into a 4-axis ultra-precision aspherical machine. The grinding force in each axis is derived by subtracting the motor current during non-grinding previously obtained in air-grinding test from the current during grinding. Firstly, the behavior of the motor current in each axis is investigated from the viewpoint of repeatability and position dependency. While the periodic fluctuation of the motor current affected by the influence of permanent magnet in the linear motor is confirmed, it shows high repeatability at each position. This result indicated that grinding force is easily calculated from the motor current with less uncertainty. Then, influence of grinding condition in the range of ductile mode grinding on the shape error is analyzed by monitoring the motor current. Toward the outside of the workpiece, the shape error gradually increases with the increase of motor current, which means larger grinding force at the outer side causes the deformation of the resin grinding wheel.","PeriodicalId":23676,"journal":{"name":"Volume 2: Manufacturing Processes; Manufacturing Systems; Nano/Micro/Meso Manufacturing; Quality and Reliability","volume":"224 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89169427","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The onset of Industry 4.0 brings a greater demand for Human-Robot Collaboration (HRC) in manufacturing. This has led to a critical need for bridging the sensing and AI with the mechanical-n-physical necessities to successfully augment the robot’s awareness and intelligence. In a HRC work cell, options for sensors to detect human joint locations vary greatly in complexity, usability, and cost. In this paper, the use of depth cameras is explored, since they are a relatively low-cost option that does not require users to wear extra sensing hardware. Herein, the Google Media Pipe (BlazePose) and OpenPose skeleton tracking software packages are used to estimate the pixel coordinates of each human joint in images from depth cameras. The depth at each pixel is then used with the joint pixel coordinates to generate the 3D joint locations of the skeleton. In comparing these skeleton trackers, this paper also presents a novel method of combining the skeleton that the trackers generate from each camera’s data utilizing a quaternion/link-length representation of the skeleton. Results show that the overall mean and standard deviation in position error between the fused skeleton and target locations was lower compared to the skeletons resulting directly from each camera’s data.
工业4.0的到来为制造业带来了对人机协作(HRC)的更大需求。这导致迫切需要将传感和人工智能与机械和物理必需品连接起来，以成功增强机器人的意识和智能。在HRC工作单元中，用于检测人体关节位置的传感器的选择在复杂性、可用性和成本方面差异很大。在本文中，深度相机的使用进行了探索，因为它们是一种相对低成本的选择，不需要用户佩戴额外的传感硬件。本文使用Google Media Pipe (BlazePose)和OpenPose骨骼跟踪软件包来估计深度相机图像中每个人体关节的像素坐标。然后将每个像素处的深度与关节像素坐标一起使用，以生成骨骼的3D关节位置。在比较这些骨骼跟踪器时，本文还提出了一种新的方法，利用骨骼的四元数/链接长度表示，将跟踪器从每个相机的数据生成的骨骼结合起来。结果表明，融合骨架与目标位置之间的总体位置误差均值和标准差比直接由每个相机数据产生的骨架要低。
{"title":"Comparison of Human Skeleton Trackers Paired With a Novel Skeleton Fusion Algorithm","authors":"Jared T. Flowers, G. Wiens","doi":"10.1115/msec2022-85269","DOIUrl":"https://doi.org/10.1115/msec2022-85269","url":null,"abstract":"\u0000 The onset of Industry 4.0 brings a greater demand for Human-Robot Collaboration (HRC) in manufacturing. This has led to a critical need for bridging the sensing and AI with the mechanical-n-physical necessities to successfully augment the robot’s awareness and intelligence. In a HRC work cell, options for sensors to detect human joint locations vary greatly in complexity, usability, and cost. In this paper, the use of depth cameras is explored, since they are a relatively low-cost option that does not require users to wear extra sensing hardware. Herein, the Google Media Pipe (BlazePose) and OpenPose skeleton tracking software packages are used to estimate the pixel coordinates of each human joint in images from depth cameras. The depth at each pixel is then used with the joint pixel coordinates to generate the 3D joint locations of the skeleton. In comparing these skeleton trackers, this paper also presents a novel method of combining the skeleton that the trackers generate from each camera’s data utilizing a quaternion/link-length representation of the skeleton. Results show that the overall mean and standard deviation in position error between the fused skeleton and target locations was lower compared to the skeletons resulting directly from each camera’s data.","PeriodicalId":23676,"journal":{"name":"Volume 2: Manufacturing Processes; Manufacturing Systems; Nano/Micro/Meso Manufacturing; Quality and Reliability","volume":"7 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83606615","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Shoichi Tamura, Takeru Ishikuri, Katsufumi Inazawa, T. Matsumura
Hybrid manufacturing involving additive and subtractive process in the same platform has been attracted to produce complex shaped parts with excellent surface finish and accuracy in surgical, aerospace, tool and die industries. The microstructures of metals fabricated by laser powder bed fusion are different from the wrought material. The machining of the material hardened in situ locally heat treated due to laser scanning is conducted without lubrication in the hybrid manufacturing. The paper discusses the milling process of AISI 420 martensitic stainless steel manufactured in additive process with comparing to a quenched wrought material and a quenched-tempered material. The cutting tests were conducted in slotting with a helical square end mill. The chip and the burr formations are discussed with the measured cutting forces. Then, the cutting forces are analyzed in a chip flow model, in which the three-dimensional chip flow is interpreted as a piling up of orthogonal cuttings in the planes containing cutting and chip flow velocities. The shear angle and the shear stress on the shear plane during chip formation are associated with the cutting force.
{"title":"Cutting Process in Slot Milling of AISI420 Stainless Steel Manufactured in Additive Process","authors":"Shoichi Tamura, Takeru Ishikuri, Katsufumi Inazawa, T. Matsumura","doi":"10.1115/msec2022-85416","DOIUrl":"https://doi.org/10.1115/msec2022-85416","url":null,"abstract":"\u0000 Hybrid manufacturing involving additive and subtractive process in the same platform has been attracted to produce complex shaped parts with excellent surface finish and accuracy in surgical, aerospace, tool and die industries. The microstructures of metals fabricated by laser powder bed fusion are different from the wrought material. The machining of the material hardened in situ locally heat treated due to laser scanning is conducted without lubrication in the hybrid manufacturing. The paper discusses the milling process of AISI 420 martensitic stainless steel manufactured in additive process with comparing to a quenched wrought material and a quenched-tempered material. The cutting tests were conducted in slotting with a helical square end mill. The chip and the burr formations are discussed with the measured cutting forces. Then, the cutting forces are analyzed in a chip flow model, in which the three-dimensional chip flow is interpreted as a piling up of orthogonal cuttings in the planes containing cutting and chip flow velocities. The shear angle and the shear stress on the shear plane during chip formation are associated with the cutting force.","PeriodicalId":23676,"journal":{"name":"Volume 2: Manufacturing Processes; Manufacturing Systems; Nano/Micro/Meso Manufacturing; Quality and Reliability","volume":"45 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83674045","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Additive manufacturing (AM) has shown great potentials in fabricating titanium aluminides (TiAl-based alloys) toward high-temperature components in aerospace and automotive applications. However, due to the complex thermal conditions during AM, the as-printed components typically contain heterogeneous microstructure, leading to nonuniform mechanical properties. A thorough understanding of microstructure evolution during AM is necessary to fabricate high-performance TiAl-based components. In this work, the mechanism for the formation of heterogeneous microstructure during selective laser melting (SLM), particularly the spatial variations in sub-grain cellular structure, was revealed by a computational framework. Specifically, a binary Ti-45Al (at.%) alloy was used for the SLM experimental observation and model development to investigate the process-microstructure relationship. The computational framework integrates a finite element thermal model and a phase-field microstructural model. A particular focus was put on the local sub-grain cellular structure evolution within the melt pool. The microstructural sensitivity to spatial variations and individual processing parameters were investigated to better understand the non-equilibrium solidification during SLM. Good agreements in the sub-grain size were achieved between experimental measurements and modeling predictions. This work presents valuable insights and guidance toward the process optimization and alloy design for fabricating high-performance TiAl-based alloys.
{"title":"Investigating the Heterogeneity in Microstructure Evolution During Selective Laser Melting of Titanium Aluminides: An Integrated Experimental and Modeling Study","authors":"Xing Zhang, L. Mushongera, Y. Liao","doi":"10.1115/msec2022-85722","DOIUrl":"https://doi.org/10.1115/msec2022-85722","url":null,"abstract":"\u0000 Additive manufacturing (AM) has shown great potentials in fabricating titanium aluminides (TiAl-based alloys) toward high-temperature components in aerospace and automotive applications. However, due to the complex thermal conditions during AM, the as-printed components typically contain heterogeneous microstructure, leading to nonuniform mechanical properties. A thorough understanding of microstructure evolution during AM is necessary to fabricate high-performance TiAl-based components. In this work, the mechanism for the formation of heterogeneous microstructure during selective laser melting (SLM), particularly the spatial variations in sub-grain cellular structure, was revealed by a computational framework. Specifically, a binary Ti-45Al (at.%) alloy was used for the SLM experimental observation and model development to investigate the process-microstructure relationship. The computational framework integrates a finite element thermal model and a phase-field microstructural model. A particular focus was put on the local sub-grain cellular structure evolution within the melt pool. The microstructural sensitivity to spatial variations and individual processing parameters were investigated to better understand the non-equilibrium solidification during SLM. Good agreements in the sub-grain size were achieved between experimental measurements and modeling predictions. This work presents valuable insights and guidance toward the process optimization and alloy design for fabricating high-performance TiAl-based alloys.","PeriodicalId":23676,"journal":{"name":"Volume 2: Manufacturing Processes; Manufacturing Systems; Nano/Micro/Meso Manufacturing; Quality and Reliability","volume":"8 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89108027","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Wrought superalloy IN718 and powder metallurgy (P/M) FGH96 were joined by linear friction welding (LFW). The variation of microstructure and mechanical properties at different welding parameters has been investigated. Macrostructural examination of the double flash morphology indicated a conservative shortening length of 2.57 mm that was recommended to extrude out the original surface contaminants into the flash. Weld zone of the joint was featured with weld interface zone (WIZ) and thermo-mechanically affected zone (TMAZ), where deformed morphology tended to be more narrow with increasing applied pressure. The increasing oscillatory frequency or decreasing applied pressure promoted the refinement of dynamically recrystallised γ matrix grain. The analysis of electron backscatter diffraction (EBSD) mapping of the weldments showed that dynamic recrystallisation (DRX) occurred in the weld zone of dissimilar nickel-based superalloy. Continuous dynamic recrystallisation (CDRX) became the predominant behaviour, accompanied by inconspicuous discontinuous dynamic recrystallisation (DDRX). The scanning electron microscope (SEM) shows that dissolution of the strengthening phase occurred from WIZ to TMAZ, strongly influencing hardness distribution across the interface. Sound joints with a higher interface strength than the base metals of IN718 were obtained.
{"title":"Mechanical Property and Microstructure of IN718/FGH96 Dissimilar Superalloy Linear Friction Weldment","authors":"Mingxiang Wang, Peihao Geng, Hong Ma, G. Qin","doi":"10.1115/msec2022-85288","DOIUrl":"https://doi.org/10.1115/msec2022-85288","url":null,"abstract":"Wrought superalloy IN718 and powder metallurgy (P/M) FGH96 were joined by linear friction welding (LFW). The variation of microstructure and mechanical properties at different welding parameters has been investigated. Macrostructural examination of the double flash morphology indicated a conservative shortening length of 2.57 mm that was recommended to extrude out the original surface contaminants into the flash. Weld zone of the joint was featured with weld interface zone (WIZ) and thermo-mechanically affected zone (TMAZ), where deformed morphology tended to be more narrow with increasing applied pressure. The increasing oscillatory frequency or decreasing applied pressure promoted the refinement of dynamically recrystallised γ matrix grain. The analysis of electron backscatter diffraction (EBSD) mapping of the weldments showed that dynamic recrystallisation (DRX) occurred in the weld zone of dissimilar nickel-based superalloy. Continuous dynamic recrystallisation (CDRX) became the predominant behaviour, accompanied by inconspicuous discontinuous dynamic recrystallisation (DDRX). The scanning electron microscope (SEM) shows that dissolution of the strengthening phase occurred from WIZ to TMAZ, strongly influencing hardness distribution across the interface. Sound joints with a higher interface strength than the base metals of IN718 were obtained.","PeriodicalId":23676,"journal":{"name":"Volume 2: Manufacturing Processes; Manufacturing Systems; Nano/Micro/Meso Manufacturing; Quality and Reliability","volume":"7 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84042203","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Human-robot load-sharing is a potential application for human-robot collaborative systems in production environments. However, knowledge of the appropriate data-driven models for this application type is limited due to a lack of physical real-world data and validation metrics. This paper describes and demonstrates a load-sharing testbed for evaluating data-driven models in a human-robot load-sharing application. Specifically, the testbed consists of a single operator and single robot relocating a payload to a desired destination. In this work, the operator initially communicates to the robot using audio feedback to initiate and alter robotic motion commands. During the payload relocation, human, payload, and robot state data are recorded. The measurements are then used to train three data-driven models (neural network, naïve Bayes, and random forest). The data-driven models are then used to transmit movement commands to the robot during human-robot load-sharing without the use of audio feedback, thus improving robustness and eliminating audio signal processing time. Evaluation of the three data-driven models shows that the random forest model was demonstrated to be the most accurate model followed by naïve Bayes and then the neural network. Hence, the results of this study provide novel insight into the types of data-driven models that can be used in load-sharing applications in addition to development of a real-world testbed.
{"title":"Evaluation of Data-Driven Models in Human-Robot Load-Sharing","authors":"Vinh Nguyen, J. Marvel","doi":"10.1115/msec2022-83907","DOIUrl":"https://doi.org/10.1115/msec2022-83907","url":null,"abstract":"\u0000 Human-robot load-sharing is a potential application for human-robot collaborative systems in production environments. However, knowledge of the appropriate data-driven models for this application type is limited due to a lack of physical real-world data and validation metrics. This paper describes and demonstrates a load-sharing testbed for evaluating data-driven models in a human-robot load-sharing application. Specifically, the testbed consists of a single operator and single robot relocating a payload to a desired destination. In this work, the operator initially communicates to the robot using audio feedback to initiate and alter robotic motion commands. During the payload relocation, human, payload, and robot state data are recorded. The measurements are then used to train three data-driven models (neural network, naïve Bayes, and random forest). The data-driven models are then used to transmit movement commands to the robot during human-robot load-sharing without the use of audio feedback, thus improving robustness and eliminating audio signal processing time. Evaluation of the three data-driven models shows that the random forest model was demonstrated to be the most accurate model followed by naïve Bayes and then the neural network. Hence, the results of this study provide novel insight into the types of data-driven models that can be used in load-sharing applications in addition to development of a real-world testbed.","PeriodicalId":23676,"journal":{"name":"Volume 2: Manufacturing Processes; Manufacturing Systems; Nano/Micro/Meso Manufacturing; Quality and Reliability","volume":"62 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80489888","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The past decade has seen a significantly increased use of high-power ultrafast lasers in micromachining applications. With the continual increase of the laser power for ultrafast lasers, an increase in the ablation rate has been brought about. However, it also created some negative effects, such as the heat-affected zone (HAZ) and thermal damages, which hardly occur at lower power. This issue was reported in the literature but has not been systematically addressed by previous research. This paper presents a systematic study on using the burst mode ablation to limit the HAZ while maintaining a high ablation efficiency using a high-power industrial picosecond laser with burst fluence larger than 10 J/cm2. An extended three-dimensional two-temperature model (3D-TTM) was employed to study the mechanism of the HAZ development and to predict the ablation efficiency with experimental validation. The essentiality of including the lattice heat conduction to predict accurate HAZ was discussed. The effect of the number of pulses per burst and pulse to pulse separation time was investigated. The optimal number of pulses per burst was obtained by using the 3D-TTM for copper and stainless steel. The 3D-TTM suggested that by using the optimal number of pulses per burst, a maximum reduction of 77% and 61% in HAZ could be achieved for copper and stainless steel respectively. And the corresponding ablation efficiency will be increased by 24% and 163% for copper and stainless steel at the same time. This study showed that burst mode laser machining at high fluence is an effective way of increasing efficiency while limiting the HAZ.
{"title":"Analysis of the Heat-Affected Zone and Ablation Efficiency in Terms of Burst Mode Parameters During High Power Picosecond Laser Micromachining of Metals","authors":"Sijie Zhang, Y. Shin","doi":"10.1115/msec2022-86171","DOIUrl":"https://doi.org/10.1115/msec2022-86171","url":null,"abstract":"\u0000 The past decade has seen a significantly increased use of high-power ultrafast lasers in micromachining applications. With the continual increase of the laser power for ultrafast lasers, an increase in the ablation rate has been brought about. However, it also created some negative effects, such as the heat-affected zone (HAZ) and thermal damages, which hardly occur at lower power. This issue was reported in the literature but has not been systematically addressed by previous research. This paper presents a systematic study on using the burst mode ablation to limit the HAZ while maintaining a high ablation efficiency using a high-power industrial picosecond laser with burst fluence larger than 10 J/cm2. An extended three-dimensional two-temperature model (3D-TTM) was employed to study the mechanism of the HAZ development and to predict the ablation efficiency with experimental validation. The essentiality of including the lattice heat conduction to predict accurate HAZ was discussed. The effect of the number of pulses per burst and pulse to pulse separation time was investigated. The optimal number of pulses per burst was obtained by using the 3D-TTM for copper and stainless steel. The 3D-TTM suggested that by using the optimal number of pulses per burst, a maximum reduction of 77% and 61% in HAZ could be achieved for copper and stainless steel respectively. And the corresponding ablation efficiency will be increased by 24% and 163% for copper and stainless steel at the same time. This study showed that burst mode laser machining at high fluence is an effective way of increasing efficiency while limiting the HAZ.","PeriodicalId":23676,"journal":{"name":"Volume 2: Manufacturing Processes; Manufacturing Systems; Nano/Micro/Meso Manufacturing; Quality and Reliability","volume":"38 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83577850","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
S. Chandra Mouli, S. Sedaghat, Muhammed Ramazan Oduncu, Ajanta Saha, R. Rahimi, Muhammad A. Alam, Alexander Wei, A. Shakouri, Bruno Ribeiro
Roll-to-roll printing has significantly shortened the time from design to production of sensors and IoT devices, while being cost-effective for mass production. But due to less manufacturing tolerance controls available, properties such as sensor thickness, composition, roughness, etc., cannot be precisely controlled. Since these properties likely affect the sensor behavior, roll-to-roll printed sensors require validation testing before they can be deployed in the field. In this work, we improve the testing of Nitrate sensors that need to be calibrated in a solution of known Nitrate concentration for around 1–2 days. To accelerate this process, we observe the initial behavior of the sensors for a few hours, and use a physics-informed machine learning method to predict their measurements 24 hours in the future, thus saving valuable time and testing resources. Due to the variability in roll-to-roll printing, this prediction task requires models that are robust to changes in properties of the new test sensors. We show that existing methods fail at this task and describe a physics-informed machine learning method that improves the prediction robustness to different testing conditions (≈ 1.7× lower in real-world data and ≈ 5× lower in synthetic data when compared with the current state-of-the-art physics-informed machine learning method).
{"title":"Physics-Informed Machine Learning for Accelerated Testing of Roll-to-Roll Printed Sensors","authors":"S. Chandra Mouli, S. Sedaghat, Muhammed Ramazan Oduncu, Ajanta Saha, R. Rahimi, Muhammad A. Alam, Alexander Wei, A. Shakouri, Bruno Ribeiro","doi":"10.1115/msec2022-85392","DOIUrl":"https://doi.org/10.1115/msec2022-85392","url":null,"abstract":"Roll-to-roll printing has significantly shortened the time from design to production of sensors and IoT devices, while being cost-effective for mass production. But due to less manufacturing tolerance controls available, properties such as sensor thickness, composition, roughness, etc., cannot be precisely controlled. Since these properties likely affect the sensor behavior, roll-to-roll printed sensors require validation testing before they can be deployed in the field. In this work, we improve the testing of Nitrate sensors that need to be calibrated in a solution of known Nitrate concentration for around 1–2 days. To accelerate this process, we observe the initial behavior of the sensors for a few hours, and use a physics-informed machine learning method to predict their measurements 24 hours in the future, thus saving valuable time and testing resources. Due to the variability in roll-to-roll printing, this prediction task requires models that are robust to changes in properties of the new test sensors. We show that existing methods fail at this task and describe a physics-informed machine learning method that improves the prediction robustness to different testing conditions (≈ 1.7× lower in real-world data and ≈ 5× lower in synthetic data when compared with the current state-of-the-art physics-informed machine learning method).","PeriodicalId":23676,"journal":{"name":"Volume 2: Manufacturing Processes; Manufacturing Systems; Nano/Micro/Meso Manufacturing; Quality and Reliability","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89625130","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper analyzed the human-robot collaborative disassembly line balancing problem, which is significantly different from the traditional disassembly line balancing problem. In a human-robot collaborative disassembly line, multiple people and robots perform disassembly tasks at each workstation. Due to the uncertainties such as product quality and human capabilities, the human-robot collaborative disassembly line balancing problem is a dynamic optimization problem. We take into account the uncertainty of product quality and personnel capabilities. In addition, dynamic optimization problems require fast and accurate tracking of Pareto’s optimal solution set in a changing environment, and transfer learning has been proven appropriate. Therefore, an individual-based transfer learning-assisted evolutionary dynamic optimization algorithm has been developed to handle the human-robot collaborative disassembly line balancing problem. The algorithm uses an individual-based transfer learning technique to reuse experience, which accelerates the generation of the initial population and improves the convergence speed of solutions. Finally, based on a set of problem examples generated in this paper, the proposed algorithm is compared and analyzed with several competitors in terms of the mean inverted generational distance and the mean hyper-volume, verifying the effectiveness of the proposed algorithm on the dynamic human-robot collaborative disassembly line balancing. The results show that the proposed algorithm has good performance in large scale problems.
{"title":"Individual-Based Transfer Learning for Dynamic Human-Robot Collaborative Disassembly Line Balancing","authors":"Yilin Fang, Xiao Zhang","doi":"10.1115/msec2022-85362","DOIUrl":"https://doi.org/10.1115/msec2022-85362","url":null,"abstract":"\u0000 This paper analyzed the human-robot collaborative disassembly line balancing problem, which is significantly different from the traditional disassembly line balancing problem. In a human-robot collaborative disassembly line, multiple people and robots perform disassembly tasks at each workstation. Due to the uncertainties such as product quality and human capabilities, the human-robot collaborative disassembly line balancing problem is a dynamic optimization problem. We take into account the uncertainty of product quality and personnel capabilities. In addition, dynamic optimization problems require fast and accurate tracking of Pareto’s optimal solution set in a changing environment, and transfer learning has been proven appropriate. Therefore, an individual-based transfer learning-assisted evolutionary dynamic optimization algorithm has been developed to handle the human-robot collaborative disassembly line balancing problem. The algorithm uses an individual-based transfer learning technique to reuse experience, which accelerates the generation of the initial population and improves the convergence speed of solutions. Finally, based on a set of problem examples generated in this paper, the proposed algorithm is compared and analyzed with several competitors in terms of the mean inverted generational distance and the mean hyper-volume, verifying the effectiveness of the proposed algorithm on the dynamic human-robot collaborative disassembly line balancing. The results show that the proposed algorithm has good performance in large scale problems.","PeriodicalId":23676,"journal":{"name":"Volume 2: Manufacturing Processes; Manufacturing Systems; Nano/Micro/Meso Manufacturing; Quality and Reliability","volume":"71 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72970438","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Machine tool dynamics has an important role in machine health monitoring, chatter prevention, machining process control, and improvement of manufacturing accuracy. For industrial applications, system dynamics vary with the machining process. To achieve this information during operation, the operational modal analysis (OMA) method, which can observe dynamic performance with output only, has rapidly developed and evolved into transmissibility function-based operational modal analysis (TOMA), with the ratio of outputs used for the estimation of system dynamics. However, this analysis is limited by excitation variance and nature of coherence deformation for the entire system when machining position or posture change. To precisely estimate machine dynamics in process, this study proposes the use of feature fitting from a mode shape database after collection by TOMA. By segmenting the continuous machining process, the variation in system dynamics during operation can be separated into domains under the requirement of machining accuracy. A numerical experiment was performed with a no-damping finite element model of one machining center to verify the feasibility and accuracy of the proposed method. Subsequently, the experimental performance of the mode shape with several accelerometers was evaluated, and the differences with finite element results were discussed with further consideration of application in practice.
{"title":"Establishment of Mode Shape Database for Machine Tool and its Application in Estimating System Dynamics","authors":"Jiahui Liu, Toru Kizaki, Shogo Yamaura, N. Sugita","doi":"10.1115/msec2022-83881","DOIUrl":"https://doi.org/10.1115/msec2022-83881","url":null,"abstract":"\u0000 Machine tool dynamics has an important role in machine health monitoring, chatter prevention, machining process control, and improvement of manufacturing accuracy. For industrial applications, system dynamics vary with the machining process. To achieve this information during operation, the operational modal analysis (OMA) method, which can observe dynamic performance with output only, has rapidly developed and evolved into transmissibility function-based operational modal analysis (TOMA), with the ratio of outputs used for the estimation of system dynamics. However, this analysis is limited by excitation variance and nature of coherence deformation for the entire system when machining position or posture change. To precisely estimate machine dynamics in process, this study proposes the use of feature fitting from a mode shape database after collection by TOMA. By segmenting the continuous machining process, the variation in system dynamics during operation can be separated into domains under the requirement of machining accuracy. A numerical experiment was performed with a no-damping finite element model of one machining center to verify the feasibility and accuracy of the proposed method. Subsequently, the experimental performance of the mode shape with several accelerometers was evaluated, and the differences with finite element results were discussed with further consideration of application in practice.","PeriodicalId":23676,"journal":{"name":"Volume 2: Manufacturing Processes; Manufacturing Systems; Nano/Micro/Meso Manufacturing; Quality and Reliability","volume":"4 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76094688","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：