Multi-material stacks comprising composite and metallic layers are widely used in aerospace components. For the assembly of structural part, high-quality holes are required to ensure the performance of the mechanical fastening (rivet/bolt). On the final assembly lines of aircraft structures, these machining operations are often performed using an electric automated drilling unit (eADU). When drilling hybrid stacks, the difference in machinability of the materials (CFRP, titanium alloy, and aluminum alloy) makes it difficult to avoid delamination, fibre pullout, matrix degradation, burrs, roughness, and size defects. Therefore, each material must be drilled with suitable machining conditions to meet the demanding quality requirements. To this end, automated material detection would allow for the adjustment of appropriate cutting parameters for each material. Due to the numerous stack configurations (material, thickness) and the variability of process parameters (tool geometry, cutting conditions, lubrication, etc.), automated material detection is not an easy task. To address this issue in eADU applications, this paper presents a novel approach to identify during the process the drilled material being manufactured using a random forest (RF) machine learning model and multi-sensor data fusion. Cutting forces, vibration, micro-lubrication conditions (flow rate and pressure), and eADU spindle and feed motor currents are monitored on a dedicated drilling test rig. Numerous tests were performed on Al7175/CFRP stacks with different cutting conditions to validate the proposed methodology. Advanced signal processing and analysis in the time and frequency domains are used for feature extraction to identify Al7175 from the CFRP. The input features of the RF model were selected using the feature importance measure embedded in the RF model learning process. The knowledge process of drilling multiple material stacks is also considered. The results showed that the features extracted from the frequency domain are more effective in identifying the CFRP Al7175 than those extracted from the time domain. This optimal subset was then used to build the RF model. The proposed methodology resulted in a highly accurate classification allowing the implementation of an adaptive machining process on the eADU for hybrid stack drilling.

Failure monitoring of industrial systems is imperative in order to ensure their reliability and competitiveness. This paper presents an innovative hybrid modelling approach applied to DC electric motors, specifically the Kollmorgen AKM42 servomotor. The proposed Cooperative Hybrid Model for Classification (CHMC) combines physics-based and data-driven models to improve fault detection and extrapolation to new usage profiles. The integration of physical knowledge of the healthy behaviour of the motor into a recurrent neural network enhances the accuracy of bearing fault detection by identifying three health states: healthy, progressive fault and stabilised fault. Additionally, Singular Value Decomposition (SVD) is employed for the purposes of feature extraction and dimensionality reduction, thereby enhancing the model’s capacity to generalise with limited training data. The findings of this study demonstrate that a reduction in the input data of 90% preserves the essential information, with an analysis of the first harmonics revealing a narrow frequency range. This elucidates the reason why the first 20 components are sufficient to explain the data variability. The findings reveal that, for usage profiles analogous to the training data, both the CHMC and NHMC models demonstrate comparable performance without reduction. However, the CHMC model exhibits superior performance in detecting true negatives (90% vs. 89%) and differentiating between healthy and failure states. The NHMC model encounters greater difficulty in distinguishing failure states (83.92% vs. 86.56% for progressive failure). When exposed to new usage profiles with increased frequency and amplitude, the CHMC model adapts better, showing superior performance in detecting true positives and handling new data, highlighting its superior extrapolation capabilities. The integration of SVD further reduces input data complexity, and the CHMC model consistently outperforms the NHMC model in these reduced data scenarios, demonstrating the efficacy of combining physical models and dimensionality reduction in enhancing the model’s generalisation, fault detection, and adaptability. This approach has the advantage of reducing the need for retraining, which makes the CHMC model a cost-effective solution for motor fault classification in industrial settings. In conclusion, the CHMC model offers a generalisable method with significant advantages in fault detection, model adaptation, and predictive maintenance performance across varying usage profiles and on unseen operational scenarios.

To improve the assembly process of aeronautical structures by the mean of One-Way Assembly strategies (no deburring of metallic parts before the installation of final fasteners), it is mandatory to monitor the burr size of drilled holes. Indeed, burrs can have a significant effect on the fatigue life of structures. It was shown that cracks are initiated from exit burrs. This paper presents an experimental analysis of the burr geometry through cutting forces and thermal imaging measurements. The effect of the tool wear on burr geometry is also analyzed to identify phenomena occurring during burr formation.

The generalisation of Deep Learning (DL) models under extrapolated usage conditions, particularly in the context of fault diagnosis, remains a significant challenge. Traditional methods encounter difficulties in terms of adaptability, necessitating extensive retraining when confronted with evolving conditions and unanticipated fault patterns. Moreover, existing DL approaches often demonstrate a lack of robustness in dynamic environments due to their limited capacity for generalisation. To address these limitations, this work introduces a novel Curriculum Learning (CL)-based approach, the Residual Singular Value Decomposition Curriculum Learning algorithm (ReSVD-CL), which integrates a Long Short-Term Memory (LSTM) model to capture temporal dependencies in fault patterns and Singular Value Decomposition (SVD) preprocessing to enhance feature representation. A complexity-based criterion segments the dataset into subsets, which are gradually incorporated into training via a pacing function. A comparison of ReSVD-CL with existing methods, such as Baby-Step Curriculum Learning (BS-CL), One-Pass Curriculum Learning (OP-CL) and baseline approaches of fault diagnosis, demonstrates that ReSVD-CL attains superior classification performance, particularly in extrapolated scenarios. The approach facilitates continuous adaptation to new usage profiles without the necessity of full retraining, thereby reducing computational costs and accelerating convergence. The study maintains a strong diagnostic performance with minimal usage profiles, thus demonstrating the potential of combining CL with dimension reduction and temporal modelling to develop more efficient and adaptable fault diagnosis systems.

The orbital drilling process is a very complex machining operation. Due to the helical path of the tool in the material and the sometimes very complex tool geometry, the geometry of the chip is very variable along the cutting edge and during a revolution of the tool. tool. This complexity explains why the cutting forces are very variable during drilling and they are very difficult to model and estimate for different tool geometries. The aim of this study is therefore to develop a cutting force model taking into account the geometry of the tool and the cutting conditions. The final objective is to control the final quality of the machined borehole. First, the geometry of the chip is modeled from the cutting parameters defining the trajectory and from the macro-geometry of the tool. Cutting force models, based on the instantaneous chip thickness and applied to the drilling, are then implemented. An experimental study validates the modeling by cutting force measurements carried out during orbital drilling tests. From this modeling, it is now possible to study the influence of the geometry of the cutting tool on the forces in order to control the loading on the tool and therefore the final quality of the drilling.