An official website of the United States government
During machining, kinetic energy is imparted to a workpiece to remove material. The integrity of the machined surface, which depends on the energy transfer, affects the quality and performance of the product, therefore needs to be quantified. Prior studies have indicated the potential of using machining power, or the power consumption at the tool-chip interface, as a process signature for predicting machined surface integrity. However, direct measurement of machining power is constrained by the availability of special equipment and the associated cost. To address this gap, this paper presents a machine learning-based method for machining power prediction through multi-sensor fusion and sequence-to-sequence translation from acoustic and vibration signals, which represent portions of the in-situ kinetic energy dissipation, to the machining power signal as a process signature. Specifically, a neural network architecture is developed to separately translate the acoustic and vibration signals to corresponding machining power signals. The two predicted power signals are subsequently fused to arrive at a unified power signal prediction. To check for spurious decision logic, the sensor fusion model is interpreted using integrated gradients to reveal temporal regions of the input data which have the most influence on the machining power prediction accuracy of the fusion model. Systematic cutting experiments performed on a lathe using 1018 steel have shown that the developed sensor fusion method for process signature prediction can successfully map machine acoustics to power consumption with 5.6% error, tool vibration to power consumption with 8.2% error, and acoustics and vibration, jointly, to power with 2.5% error. Model parameter interpretation reveals that the vibration signal is more influential on the machining power prediction result than the acoustic signal, but that overall model accuracy is diminished if only the vibration signal is used.
more »
« less
Surface roughness prediction through GAN-synthesized power signal as a process signature
Predicting machined surface roughness is critical for estimating a part’s performance characteristics such as susceptibility to fatigue and corrosion. Prior studies have indicated that power consumed at the tool-chip interface may represent an indicator for the surface integrity of the machining process. However, no quantita-tive association has been reported between the machining power and surface roughness due to a lack of data to develop predictive models. This paper presents a data synthesis method to address this gap. Specifically, a conditional generative adversarial network (CGAN) is developed to synthesize power signals associated with varying process parameter combinations. The quality of the synthesized signals is evaluated against experimentally measured power signals by examining the consistency in: 1) the spatial pattern of the signals induced by the cutting process as shown in the frequency domain, and 2) the temporal pattern as shown in the clustering of the synthesized and measured signals corresponding to the same parameter combination. The synthesized signals are then used to augment the measured signals and develop a convolutional neural network (CNN) for predicting the machined surface roughness. Experiments performed using H13 tool steel have shown that data augmentation by CGAN has effectively reduced the error of the surface roughness prediction from 58 %, when no synthetic data is used for CNN training, to 9.1 % when 250 synthetic samples are used. The results demonstrate the effectiveness of CGAN as a data augmentation method and CNN for mapping machining power to surface roughness.
more »
« less
- Award ID(s):
- 2040288
- PAR ID:
- 10435376
- Date Published:
- Journal Name:
- Journal of manufacturing systems
- ISSN:
- 0278-6125
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract Functionally graded surfaces — surfaces with properties that are engineered to have spatial variations — have numerous applications such as micropumps, auto-mixers, and flow control for lab-on-chip devices. Manufacturing of functionally graded surfaces is an increasingly important topic of research. This study investigates the feasibility of creating a functionally graded surface during channeling of borosilicate glass by the electrochemical discharge machining (ECDM) process. The ability to create surface roughness gradients in microchannels during the machining process was demonstrated by modifying the input voltage, tool feed rate, and tool rotation speed. Microchannels with graded surface roughness having Ra values ranging from 0.35 to 4.07 μm were successfully machined on borosilicate glass by ECDM. Surface profiles were obtained via a stylus profilometer, and roughness values were calculated after detrending and applying a Gaussian filter. To demonstrate the process versatility, micro channels with increasing and decreasing Ra values were machined, one increasing from 1.43 μm to 4.07 μm, another decreasing from 3.29 μm to 1.10 μm. To demonstrate the process repeatability, a micro channel with similar surface roughness on both ends and a lower Ra value in the center was created. In this channel, the Ra value at the start is 0.35 μm, reducing to 0.24 μm, then rising again to 0.38 μm in the final section.more » « less
-
Abstract Machining-induced residual stresses (MIRS) are a main driver for distortion of thin-walled monolithic aluminum workpieces. Before one can develop compensation techniques to minimize distortion, the effect of machining on the MIRS has to be fully understood. This means that not only an investigation of the effect of different process parameters on the MIRS is important. In addition, the repeatability of the MIRS resulting from the same machining condition has to be considered. In past research, statistical confidence of MIRS of machined samples was not focused on. In this paper, the repeatability of the MIRS for different machining modes, consisting of a variation in feed per tooth and cutting speed, is investigated. Multiple hole-drilling measurements within one sample and on different samples, machined with the same parameter set, were part of the investigations. Besides, the effect of two different clamping strategies on the MIRS was investigated. The results show that an overall repeatability for MIRS is given for stable machining (between 16 and 34% repeatability standard deviation of maximum normal MIRS), whereas instable machining, detected by vibrations in the force signal, has worse repeatability (54%) independent of the used clamping strategy. Further experiments, where a 1-mm-thick wafer was removed at the milled surface, show the connection between MIRS and their distortion. A numerical stress analysis reveals that the measured stress data is consistent with machining-induced distortion across and within different machining modes. It was found that more and/or deeper MIRS cause more distortion.more » « less
-
Abstract In recent years, semiconductors, electronics, optics, and various other industries have seen a significant surge in the use of sapphire materials, driven by their exceptional mechanical and chemical properties. The machining of sapphire surfaces plays a crucial role in all these applications. However, due to sapphires’ exceptionally high hardness (Mohs hardness of 9, Vickers hardness of 2300) and brittleness, machining them often presents challenges such as microcracking and chipping of the workpiece, as well as significant tool wear, making sapphires difficult to cut. To enhance the machining efficiency and machined surface integrity, ultrasonic vibration-assisted (UV-A) machining of sapphire has already been studied, showing improved performance with lower cutting force, better surface finish, and extended tool life. Scribing tests using a single-diamond tool not only are an effective method to understand the material removal mechanism and deformation characteristics during such UV-A machining processes but also can be used as a potential process for separating IC chips from wafers. This paper presents a comprehensive study of the UV-A scribing process, aiming to develop an understanding of sapphire’s material removal mechanism under varying ultrasonic power levels and cutting tool geometries. In this experimental investigation, the effect of five different levels of ultrasonic power and three different cutting tool tip angles at various feeding depths on the scribe-induced features of the sapphire surface has been presented with a quantitative and qualitative comparison. The findings indicate that at feeding depths less than 6 μm, UV-A scribing with 40–80% ultrasonic power can reduce cutting force up to 50% and thus improve scribe quality. However, between feeding depths of 6 to 10 μm, this advantage of using ultrasonic vibration gradually diminishes. Additionally, UV-A scribing with a smaller tool tip angle (60°) was found to lower cutting force by 65% and improve scribe quality, effectively inhibiting residual stress formation and microcrack propagation. Furthermore, UV-A scribing also facilitated higher critical feeding depths at around 10 μm, compared to 6 μm in conventional scribing.more » « less
-
Accurate quantification of machined surface roughness is crucial to the characterization of part performance measures, including aerodynamics and biocompatibility. Given this cruciality, there exists a need for predictive metrology models which can predict the surface roughness before a part leaves a machine to reduce metrologyinduced bottlenecks and improve production planning efficiency under emergent production paradigms, e.g., Industry 4.0. Current predictive metrology approaches in machining generally train machine learning models on all available input features at once. However, this approach yields a high number of free parameters during all states of training, possibly leading to suboptimal prediction results because of the complexity of simultaneously optimizing all parameters at once. In addition, previous machine learning-enabled surface roughness prediction studies have used limited test dataset sizes, which reduces the reliability and robustness of the reported results. To address these limitations, this study proposes a two-stage model training approach based on domainincremental learning, wherein a second stage of training is performed using an expanded input domain. The proposed method is evaluated on a 3,000-element experimentally collected testing dataset of machined H13 tool steel surfaces, where it achieves 16.3 % roughness prediction error compared to the 29.5 % error of the conventional single-stage training approach, indicating the suitability of the two-stage training method for reducing surface roughness prediction error.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1016/j.jmsy.2023.05.016
