Experiments are performed to study surface curvature effects on the impingement cooling flow and the heat transfer processes over a concave and a convex surface. A single air jet issuing from different size slots continuously impinges normally on the concave side or the convexside of a heated semicylindrical surface. An electrical resistance wire is used to generate smoke, which allows us to visualize the impinging flow structure. The local heat transfer Nusselt number along the surfaces is measured. For impingement on a convex surface, three-dimensional counterrotating vortices on the stagnation point are initiated, which result in the enhancement of the heat transfer process. For impingement on a concave surface, the heat transfer Nusselt number increases with increasing surface curvature, which suggests the initiation of Taylor–Go¨rtler vortices along the surface. In the experiment, the Reynolds number ranges from 6000 to 350,000, the slot-to-plate spacing from 2 to 16, and the diameter-to-slot-width ratio D/b from 8 to 45.7. Correlations of both the stagnation point and the average Nusselt number over the curved surface, which account for the surface curvature effect, are presented.
Due to the increasing demand of electrical vehicles (EVs), prognostics of the battery state is of paramount importance. The nonlinearity of the signal (e.g. voltage) results in the complexity of analyzing the degradation of the battery. Aging characteristics extracted from the voltage, current, and temperature when the battery is fully charged/discharged were commonly used by previous researchers to determine the battery state. The drawbacks of the previous prediction algorithms are insufficient or irrelevant features to explicitly model the battery aging and the use of fully charged/discharged datasets, which might result in poor prediction accuracy. Therefore, this study proposes a feature selection technique to adequately select optimum statistical feature subset and the use of partial charge/discharge data to determine the battery remaining useful life (RUL) using Recurrent Neural Network-Long Short-Term Memory (RNN-LSTM). The proposed approach demonstrated exceptional RUL prediction results, with the root mean square error (RMSE) of 0.00286 and mean average error (MAE) of 0.00222 using partial discharge data. The proposed method shows prediction improvement in comparison with the use of full data and state-of-the-art outcomes from previous studies of the same open data from the National Aeronautics and Space Administration (NASA) prognostic battery data sets. INDEX TERMS Recurrent neural network, long short-term memory, remaining useful life, battery management systems, feature selection.
This study describes a prediction model for the impact vibration of raised access floors by a Finite Element Method (FEM) numerical analysis. The residual ratio of the natural frequency of the peak wave is 1.3% and the residual ratio of the maximum vibration acceleration is 0.8%, which indicates the accuracy in application of the numerical model. With the validated numerical model, this study evaluates the influence on the vibration of various construction parameters. It is shown that the panel material, bracing, and panel thickness may reduce the impact vibration.
A novel tool wear predictive model was developed based on the current signals in this study. The system adapts to different part geometry with accurate prediction of the tool wear during the operation. The current sensor was utilized presenting a practical and better choice for tool wear monitoring which is inexpensive and no need to be attached to the working table or spindle. To avoid interruptions during the machining process, the tool wear was only measured at the end of the operation. The Long Short-Term Memory model was used to develop the tool wear prediction system. The tool wear prediction results indicate 23.92% and 36.41% average error for all the testing samples after 1/3 of the operations for profiling and straight turning, respectively. When the tool wear prediction was carried out after 2/3 of the operations, excellent results are observed with 6.15% error for profiling and 9.44% error for straight turning. The prediction results at the end of the operation shows 0.18% and 0.68% error for profiling and straight turning. The performance of the model using the current sensor shows that the model can predict the tool wear with less than 10% error after 2/3 of the turning operation without interfering with the turning process.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.