Advancing aircraft engine RUL predictions: an interpretable integrated approach of feature engineering and aggregated feature importance.

In this study, we present a comprehensive approach for predicting the remaining useful life (RUL) of aircraft engines, incorporating advanced feature engineering, dimensionality reduction, feature selection techniques, and machine learning models. The process begins with a rolling time series window, followed by the extraction of a multitude of statistical features, and the application of principal component analysis for dimensionality reduction. We utilize a variety of feature selection methods, such as Genetic Algorithm, Recursive Feature Elimination, Least Absolute Shrinkage and Selection Operator Regression, and Feature Importances from a Random Forest model. As a significant contribution, we introduce the novel aggregated feature importances with cross-validation (AFICv) technique, which ranks features based on their mean importance. We establish a selection criterion that retains features with a cumulative mean sum equal to 70%, thereby reducing the complexity of machine learning models and enhancing their generalizability. Four machine learning regression models-Natural and Extreme Gradient Boosting, Random Forest, and Multi-Layer Perceptron-were employed to evaluate the effectiveness of the selected features. The performance of our proposed method is assessed by the evaluation metrics Root Mean Square Error (RMSE) and R2 Score, and also considered within-interval percentages and relative accuracy metrics. Importantly, a novel PCA interpretability was introduced to provide real-world context and enhance the utility of our findings for domain experts. Our results indicate that the proposed AFICv technique efficiently achieves competitive performance across the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) sub-datasets using a significantly smaller subset of features, thus contributing to a more effective and interpretable RUL prediction methodology for aircraft engines.

Concurrent shape and build orientation optimization for FDM additive manufacturing using the principal stress lines (PSL).

Additive Manufacturing (AM) with the consisting constantly evolving technologies is a particularly popular research area. Based on the shape forming freedom, size, shape, and topology optimization techniques can be validated by AM produced parts. However, in every manufacturing process, AM also has some adverse inherent properties. One and maybe the most significant optimization problem is the mechanical anisotropy caused by the layered structure. In this paper, a simultaneous build orientation and shape optimization method is presented. Both of the approaches are intended to increase the mechanical performance of the produced parts. Shape optimization was accomplished by varying the cross-section of the beam geometries, based on the angle between a PSL section and the characteristic load direction. To test the efficiency and validate the method 2D structures (with relatively small 3rd dimension) and their tensile properties were tested. Based on the results, we can prove that the PSL method works and help to increase the mechanical performance by 19.2% with only 7.8% size increment.

Simplified local infill size optimization for FDM printed PLA parts.

The great advantage of additive manufacturing is the fact that hollowed parts with a given infill can be created. However, the standardized commercial slicer software offers a uniform infill pattern creation solution. In engineering practice, the manufactured parts are functional, therefore the appropriate load bearing capacity is mostly mandatory. In this paper a simplified local infill size optimization method has been presented. Based on a Finite Element Analysis the local density of the pattern can be adjusted, according to the emerged local stresses. The results show that independently of the pattern type, if the scaling was applied, the mechanical resistance was improved to the same extent. In case of the worst-performing uniform pattern, 84% improvement in mechanical resistance was achieved with the optimization. In addition, an FDM printing problem has been highlighted, which must be eliminated if the proposed method is used.

3D Scanner-Based Identification of Welding Defects-Clustering the Results of Point Cloud Alignment.

This paper describes a framework for detecting welding errors using 3D scanner data. The proposed approach employs density-based clustering to compare point clouds and identify deviations. The discovered clusters are then classified according to standard welding fault classes. Six welding deviations defined in the ISO 5817:2014 standard were evaluated. All defects were represented through CAD models, and the method was able to detect five of these deviations. The results demonstrate that the errors can be effectively identified and grouped according to the location of the different points in the error clusters. However, the method cannot separate crack-related defects as a distinct cluster.

Part orientation optimization for Wire and Arc Additive Manufacturing process for convex and non-convex shapes.

Building orientation optimization for Additive Manufacturing (AM) process is a crucial step because it has a vital effect on the accuracy and performance of the created part. Wire and Arc Additive Manufacturing's (WAAM) working space is less limited, and the production time is significantly shorter than the other metal 3D printers. However, one of the adverse effects of WAAM is the defect at the start and endpoints of the welding beads. In this paper, an algorithm has been invented to define the optimal printing position, reducing the number of these defects by rotating the 3D object in a loop around the X and Y axes by a small constant degree and then selecting the degree of rotation that has the fewest uninterrupted surfaces and the largest area of the first layer. The welding process will be interrupted as little as possible by the torch if there are the fewest possible uninterrupted surfaces. As a result, there will be fewer defects in the production and finishing of the welding beads. In order to have a sufficient connection surface with the build tray, which will aid in holding the workpiece in place, the largest first layer should also be sought. Therefore, it has been found that a properly defined orientation relative to the build tray can reduce the number of uninterrupted surfaces within the layers, which will improve the expected dimensional accuracy of the parts. The efficiency of the process is highly affected by the shape of the part, but in most cases, the print errors can be drastically minimized.

Layer Adhesion Test of Additively Manufactured Pins: A Shear Test.

Additive Manufacturing (AM) became a popular engineering solution not only for Rapid Prototyping (RP) as a part of product development but as an effective solution for producing complex geometries as fully functional components. Even the modern engineering tools, such as the different simulation software, have a shape optimization solution especially for parts created by AM. To extend the application of these methods in this work, the failure properties of the 3D-printed parts have been investigated via shear test measurements. The layer adhesion can be calculated based on the results, which can be used later for further numerical modeling. In conclusion, it can be stated that the layer formation and the structure of the infill have a great influence on the mechanical properties. The layers formed following the conventional zig-zag infill style show a random failure, and the layers created via extruded concentric circles show more predictable load resistance.