As a critical element in rotating machines, remaining useful life (RUL) prediction of rolling bearings plays an essential role in realizing predictive and preventative machine maintenance in modern manufacturing. The physics of defect (e.g. spall) initiation and propagation describes bearing’s service life as generally divided into three stages: normal operation, defect initiation, and accelerated performance degradation. The transition among the stages are embedded in the variations of monitored data, e.g., vibration. This paper presents a multi-mode particle filter (MMPF) that is aimed to: 1) automatically detect the transition among the three life stages; and 2) accurately characterize bearing performance degradation by integrating physical models with stochastic modeling method. In MMPF, a set of linear and non-linear modes (also called degradation functions) are first defined according to the physical/empirical knowledge as well as statistical analysis of the measured data (e.g. vibration). These modes are subsequently refined during the particle filtering (PF)-based bearing performance tracking process. Each mode corresponds to an individual performance scenario. A finite-state Markov chain switches among these modes, reflecting the transition between the service life stages. Case studies performed on two run-to-failure experiments indicate that the developed technique is effective in tracking the evolution of bearing performance degradation and predicting the remaining useful life of rolling bearings.

An oxyfuel cutting torch was biased with a positive DC voltage relative to steel and copper work pieces, and ion currents were seen to flow due to ions generated chemically in the flame and at the work surface. A repeatable 150% rise in these currents occurs over steel at a surface temperature around 1050°C. Equally strong intermittent spikes appear over clean and corroded steel samples alike. They are are easily removed by a median value filter, and are believed to be due to small reacting particulates. Tests in which salt was deliberately deposited on the steel surface produces a signal similar in shape, but exaggerated in magnitude to those currents supposed to be from other surface impurities.

This paper describes a novel method to analyze the effect of the fiber laser induced plasma on welding of dissimilar Al and Cu. Based on spectroscopic studies, specific wavelengths were found to be suitable for process monitoring during laser welding. For aluminum, the double peak at 394.4 and 396.1 nm was studied, for cu the 578 nm peak was studied. The characteristic peaks were observed by multiple photodiodes equipped with narrow optical filters with varying FWHM. Analyzing the plasma emissions of the process enabled to detect material changes in the plasma plume. This novel method has the advantage of higher sampling rate combined with reduced costs compared to a spectrometer and is thus able to capture the process dynamics.

One of the major challenges in producing highly accurate graphene-based nanoelectromechanical (NEMS) resonators is the poor fabrication repeatability of graphene-based NEMS devices due to small variations in the residual stress and initial tension of the graphene film. This has meant that graphene-based nanoelectromechanical resonators tend to have large variations in natural frequency and quality factor from device to device. This poor repeatability makes it impossible to use these resonators to make accurate, high-precision force and displacement sensors or electromechanical filters. However, by actively controlling the tension on the graphene resonator it is possible both to increase repeatability between devices and to increase the force/mass sensitivity of the nanoelectromechanical resonators produced. Such tension control makes it possible to produce electrometrical filters that can be precisely tuned over a frequency range of up to several orders-of-magnitude. In order to controllably strain the graphene resonator, a microelectromechanical system (MEMS) is be used to apply tension to the graphene. The MEMS device consists of a graphene resonator connected between a set of gold electrodes. Each gold electrode is located on a different MEMS stage. Each stage is connected to a set of flexural bearings which are used to guide the motion of the stage. The displacement stage is actuated using a thermal actuator that allows a uniform and constant tension to be applied to the graphene resonator. The displacement of the actuator and the tension applied to the graphene are measured using a pair of differential capacitive actuators. The resonator is actuated electrostatically using the electrical back gate, and the resonant frequency is measured from the change in conductance of the graphene as it approaches resonance. Using this setup, it is possible to tune the natural frequency of the graphene resonator with high precision and accuracy. In addition to designing devices that can compensate for manufacturing errors in nanomanufactured devices, this paper will present several methods that can greatly expand the scope and rate at which nanomaterials-based devices can be fabricated. For example, this paper will present a transfer-free, wafer-scale manufacturing process that can be used to produce suspended graphene-based devices such as the graphene-based NEMS resonators. This new method involves the growth of graphene directly on the device wafer and release of the graphene-based device through etching of the copper catalyst layer. This method replaces traditional graphene fabrication methods, such as mechanical exfoliation, electron beam lithography, or transfer from copper foils, which are slow and require a transfer step that is the source of much of inconsistency in suspended graphene-based devices. Therefore, these new transfer-free, wafer-scale fabrication methods offer the potential to increase the throughput, yield, and repeatability of manufacturing processes for graphene resonators while reducing manufacturing costs and complexity.

Recent developments in science and engineering have advanced the atomic manufacturing of nanoscale structures, allowing for improved high-performance technologies. The challenge now is to expand nanomanufacturing capabilities in order to fabricate a broader range of structures with higher complexity, greater precision and accuracy, and with substantially improved performance. Also, reducing the cost by avoiding the use of expensive equipment/processes is highly desirable. While current nanomanufacturing capabilities limit the commercialization of such nanoscale structures, AFM-based nanomachining is a promising tool to address issues and is considered a potential manufacturing tool for operations including machining, patterning, and assembling with in situ metrology and visualization. Most existing techniques used in the fabrication of nanofluidic channels involved the use of electron-beam lithography systems, which are very expensive processes. In addition, one of the issues associated with electron-beam lithography is when the base material is changed; the energy of ejecting electrons needs to be adjusted. This results in time spending on developing the formula for different material. In this work, atomic force microscopy (AFM) is employed in the fabrication of a nanofluidic device for medical applications. Nanofluidic channels with various depths and widths are fabricated using AFM indentation and scratching techniques. Nanoscale channel is mainly used in the study of molecular behavior at single molecule level. The interaction of molecules with the nanochannels at persistence length has led to a new way of detecting, analyzing and separating these biomolecules. The resulting device from the current work can be used for the DNA stretching application and the separation of elite group of lysosome and other viruses. The nanochannels are integrated between the microchannels and act as a filter to separate biomolecules. Sharply developed vertical microchannels are produced from a deep reaction ion etching followed by deposition of different materials such as gold and polymer on the top surface in order to study alternative ways of manufacturing the nanofluidic device. PDMS bonding is performed to close the top surface of the device. An experimental setup is used to test and validate the device by flowing fluid through the channels. A detailed cost evaluation is conducted to assess the economical merits of the proposed process. It is shown that there is 47.7% manufacturing-time savings and 60.6% manufacturing-cost savings, when using the proposed process compared to more traditional processes.

The lack of plug-and-play programmability in conventional toolpath planning approach in subtractive manufacturing, i.e., machining leads to significantly higher manufacturing cost for CNC based prototyping. In computer aided manufacturing (CAM) packages, typical B-rep or NURBS based representations of the CAD interfaces challenge core computations of tool trajectories generation process, such as, surface offsetting to be completely automated. In this work, the problem of efficient generation of free-form surface offsets is addressed with a novel volumetric representation. It presents an image filter based offsetting algorithm, which leverages the parallel computing engines on modern graphics processor unit (GPU). The scalable voxel data structure and the proposed hardware-accelerated volumetric offsetting together advance the computation and memory efficiencies well beyond the capability of past studies. Additionally, in order to further accelerate the offset computation the problem of offsetting with a large distance is decomposed into successive offsetting using smaller distances. The accuracy of the offset algorithms is thoroughly analyzed. The developed GPU implementation of the offsetting algorithm is robust in computation, easy to comprehend, and achieves a 50-fold speedup on single graphics card (NVIDIA GTX780Ti) relative to prior best-performing dual socket quad-core CPU implementation.

Due to the need of high speed and efficient biodosimetric assays for triage and therapy in the event of radiological or nuclear attack, a robotically-based automated biodosimetry tool (RABiT) has been developed over the past few years. Adapting the micronucleus assay from filter plates to V-shaped plates presented challenges in the liquid handling, namely cell splashing out of the V-shaped well plate during the cell harvesting, poor cell distribution on the bottom of the image plate during the dispensing, and cell loss from the image plate during the aspiration in the liquid handling process. Experimental and numerical investigations were carried out to better understand the phenomena and mitigate the problems. Surface tension and contact angle among the fluids and the plate wall were accounted for in the discrete and multiphase numerical models. Experimental conditions were optimized based on the numerical results showing the relationship between nozzle speed and amount of splashed liquid, and the relationship between aspiration speed and number of escaped cells. Using these optimized parameters, numbers of micronuclei in binucleated cells showed the same dose dependence in the RABiT-prepared samples as those in the manually prepared ones. Micronucleus assay protocol was fully realized on RABiT.

In this paper, monitoring and prediction of cutting tool wear condition based on dynamic data driven approaches were investigated. Sensor signals obtained from the machining processes were processed through wavelet denoising to filter the noise un-related to cutting, features in time and frequency domains were extracted using classical signal processing approaches, and then were selected with Pearson correlation coefficient. The most related features were sent to the feature fusion approaches including neural network (NN), adaptive neural fuzzy inference system (ANFIS), or support vector regression (SVR) to estimate the tool wear. Statistics performance evaluation based on correlation coefficient (R 2 ), average absolute error (AAE), and Se/Sy, as well as cross validation, selected the most proper feature fusion approach. Further, prediction models based on Bayesian model average were applied to predict the future tool wear. A case study based on the end mill experiment with signals of 3-axis cutting forces, 3-axis vibrations and acoustic emission, illustrated the proposed approach. It showed that ANFIS has the best estimation accuracy with the R 2 of 0.99, AAE of 0.42, Se/Sy of 0.12, and cross validation error of 13.36. In the prediction stage, the prediction model has high prediction accuracy with all the experiment results covered by 95% confidence interval of prediction.

The ease of entering a car is one of the important ergonomic factors that car manufacturers consider during the process of car design. This has motivated many researchers to investigate factors that affect discomfort during ingress. The patterns of motion during ingress may be related to discomfort, but the analysis of motion is challenging. In this paper, a modeling framework is proposed to use the motions of body landmarks to predict subjectively reported discomfort during ingress. Foot trajectories are used to identify a set of trials with a consistent right-leg-first strategy. The trajectories from 20 landmarks on the limbs and torso are parameterized using B-spline basis functions. Two group selection methods, group nonnegative garrote (GNNG) and stepwise group selection (SGS), are used to filter and identify the trajectories that are important for prediction. Finally, a classification and prediction model is built using support vector machine (SVM). The performance of the proposed framework is then evaluated against simpler, more common prediction models.

Force transducers such as the piezoelectric Kistler dynamometer and the wireless strain gage Smart Tool are used to monitor the cutting force during CNC machining. Due to sensor dynamics, there are differences between the actual cutting force and the measured force, characterized by phase delay and additional vibrations in the cutting profile. In this work, actual cutting forces are estimated from the measured sensor data using a Kalman filter. Since cutting forces are composed of harmonics of the tooth passing frequency and runout frequency, both a sensor dynamic model and harmonic cutting force model are included in the Kalman filter model in order to improve performance of the filter. The harmonic computations of the measured cutting force are used to self-tune the Kalman filter. Comparisons between the measured and estimated force show the ability of the Kalman filter to reduce sensor vibration and noise with no phase delay.

An experienced technician can usually identify the cutting condition by hearing the sound generated during the cutting and the sound signal can be expected to detect the features closely related to the tool condition. However, the background noise always contaminates the signal obtained by microphone system during cutting and reduces the chance of applying the sound based micro tool condition monitoring system in industry. In order to reduce the noise effect and improve the system performance, a microphone array integrated with Wiener filter was designed and implemented in this study to enhance the noise reduction capability for monitoring system. The experimental results show that the microphone array integrated with Wiener filter provides a better solution than single microphone integrated with Wiener filter or the microphone array without the post filter design in reducing the broadband background noise.

In the present study, the shape of an object is first obtained from photographic images using the photometric stereo method, and then the target object is created using material that gives an artificial impression of wood. We found through experimentation that the output gradient on the surface of the target object is limited to 67°. Moreover, there is a fault that a lustrous object cannot be reproduced by a general photometric stereo method. Then, to overcome the fault, we propose using a Gaussian filter. The results showed that the maximum error margin of the shape of the reproduced object could be adjusted from 8.3 to 3.3%. At the end, the feature that the normal vector in respect is obtained by using the photometric stereo method is made the best use of, and we propose a new method to obtain cutting location (CL) data for ball-end milling using the normal vector of the object geometries generated by the photometric stereo method. As a result, the tool path was quickly and easily derived, without the use of CAD data.

This paper uses simulation to investigate measurement errors resulting from the camera point spread function (PSF) when measuring the temperature of segmented chip formation using infrared (IR) thermography. The PSF of the IR camera effectively filters the results which can cause significant errors due to the large temperature gradients and abrupt transitions between features and their corresponding emissivity values. The different emissivity values of the tool, workpiece, chip body, and shear band affect the apparent difference in the emitted energy of these different features. This decreases the measured temperature in the regions of most interest: along the tool-chip interface and the periodic shear zone. The method in this study creates an appropriate emissivity map from post-process measurements and applies it to results from the temperature distribution of the cutting zone predicted by commercial finite element analysis (FEA) software. Comparisons between the simulation results and experiment results show that the emissivity values obtained form the post process chip analysis lead to good agreement. The resulting radiant intensity distribution becomes the input for an IR camera simulation module developed by the authors and presented in earlier work [1]. The earlier work used the true temperature distribution predicted by the FEA as the simulation module input, and did not incorporate the IR camera’s PSF. Implementation of the actual IR camera’s PSF allows the simulation module to more accurately represent the measurements of the IR camera and ultimately allow the comparison of the simulation results to the measurement results. Simulation results show that the PSF accounts for 45% of the 42 °C radiance temperature error at the tool-chip contact along the rake face. The PSF accounts for approximately 15% of the 46 °C radiance temperature measurement error at a point in the center of the catastrophic shear band. These errors consider the effects of motion blur (integration time) and magnification (size-of-source), as described in the earlier work [1].

In this paper the dynamical behavior of a part and an actuated fixture system is studied. The part is modeled using the finite element method. Subsequently, a reduced model is established using the Craig-Bampton reduction method in order to create a small-sized model, accurately describing the dynamic behavior of the part. The clampers and locators of the fixture are modeled as springs. The fixture frame is considered to be much stiffer than the locators such that it provides zero displacement boundary conditions to the locators. An electromechanical actuator is utilized to provide the adaptive clamping forces. After investigating proportional control, integral control, the three term classical PID controller and a lag filter for dynamic compensation, the analysis shows that position feedback can be used effectively in combination with integral control action to minimize unnecessary displacement of the workpiece and increase the bandwidth of the actuated fixture system.

Electrospun polymer/nanoparticle composite-fiber strips are introduced as permeable, flexible strain sensors. The composite strips are referred as nanocomposite in this study. In the experiments with uniaxial elongation, the nanocomposite strips showed a reproducible increase in electrical resistance indicating the strain level. The resistance increase was interpreted with a novel percolation model that accommodated changing strips geometry. A theoretical framework for nanocomposites as possible detectors of local clogging in large-scale filters is also proposed.