Quality assurance (QA) of clinched joints is predominantly performed by destructive testing. Recently, non-destructive evaluation (NDE) methods received increasing attention as a potential alternative. However, the inherently indirect measurement of underlying effects poses a significant challenge to its broader application. To tackle this, two experimental data sets, containing a total of 43 potential process deviations and defects are established using transient dynamic analysis (TDA). On these, several machine learning (ML) models are trained to detect the underlying deviations. The best-in-class model is used to identify a frequency band at which a classification accuracy of 88.58% across all 43 classes is achieved. Further analysis of the most discriminative model features reveals the importance of measuring both excitation as well as specimen response. This lays the foundation for further research towards defect specific in-line measurements of mechanical joints, further improving joint reliability.

Ring rolling is an important incremental forming process for manufacturing seamless rings. However, fast and accurate prediction for process parameters remains challenging, especially in the context of generating large datasets. Therefore, this paper introduces a novel semi-analytical approach, based on the upper bound method, to calculate process parameters incrementally in cold ring rolling. The proposed method is validated against experimental data and finite element (FE) simulations, demonstrating a trade-off between computational speed and accuracy. While slightly less precise than FE simulations, the approach drastically reduces computation time, making it suitable for rapid parameter prediction prior to manufacturing. Incremental calculations further enhance optimization potential, offering a practical solution for improving manufacturing efficiency and sustainability. The development of data-driven approaches, particularly in machine learning (ML), has revealed new possibilities, such as the early identification of form errors. However, training machine learning algorithms requires extensive datasets, which can be obtained in ring rolling from experiments, FE simulations, or (semi-)analytical methods. Therefore, the proposed method enables the generation of large interpolative and extrapolative datasets, which can serve as inputs for ML models.

Sheet metal components with complex geometries are typically recycled by remelting. Direct remanufacturing necessitates the flattening of parts, which requires the implementation of cuts to facilitate unwinding. The exact positioning of these cuts is a complex planning task, because several influencing factors can be considered, such as material usage, ease of flattening, or minimal forming required. This study presents a geometry-based concept addressing this challenge and demonstrates its use for a test geometry. The finite element method is applied to simulate the flattening process of the resulting sections, and the results are evaluated in terms of planarity and induced plastic strain. The findings of the present work indicate a discernible dependency of results on the selection of the flattening directions. In particular, curved areas impact the induced plastic deformation and springback of flattened sections. This is a crucial consideration when planarity is prioritised over material utilisation.

Modeling metal forming processes, and specifically ring rolling as considered in this research, is crucial for process design and the development of new product variants. Although finite element simulations and analytical modeling are common practices to perform process predictions, they can still present limitations such as high computational time, adherence to reality, and modeling accuracy. Trying to address all the above, this research proposes a novel generative machine learning solution for the creation of accurate and physics-compliant synthetic data for the ring rolling process, enabling almost real time and accurate process modeling within latent space. The proposed data generation method is based on multivariate generative adversarial network (GAN) combined with a physics-guided neural network (PGNN) to incorporate constraints relevant to the ring rolling process and integrate them into an auxiliary loss granting physical consistency. The proposed GAN architecture is then annotated as a physics-guided auxiliary GAN (PG-A-GAN). The physical constraints introduced are relevant to an analytical slip-line model for the force, a surrogate model for the torque, volume consistency, and an analytical model for the time–diameter relationship. Considering the duality of the physical modeling approach for the force and torque, where data-driven machine learning is stirred by analytical models, the proposed approach is defined as analytical transfer learning. Findings reveal that employing PGNNs in the GANs learning process improves physical loss in data generation, combined with a slight increase in data distribution similarity with respect to experimental instances. To enable modeling of specific and customized rolling instances, a condition based on the final sought ring geometry was introduced, within an auxiliary classifier generative adversarial network (ACGAN) framework. The proposed architecture allows the generation of multivariate, physically constrained, rolling time series and highlights the feasibility of such a modeling approach within the latent space and might be extended to other manufacturing processes by adapting modeling and retraining the proposed architecture.

This paper examines the impact of a rotationally superimposed punch stroke on the binding mechanisms of clinched joints of aluminum sheets. As part of the development of a method for ensuring the versatility of clinching, an additional rotational movement of the punch was introduced as a control variable to influence friction in the mechanical joining process. The effect of rotational superimposition on the force-displacement curve of the clinching processes was investigated using four test variants with different kinematics. The primary objective was to evaluate the binding mechanisms that maintain the integrity of the clinched joint. To evaluate the force closure of the resulting joint, two testing methods were employed throughout the course of the research, non-destructive resistance measurement using four-wire sensing method and destructive torsion testing. A crucial factor influencing the efficacy of the process is surface cleanliness, as contaminants between joining partners can impede the effectiveness of the clinched joint. Therefore, all specimens were meticulously cleaned prior to experimentation. This method exhibits promising potential in creating clinched joints that align with the demands of flexible manufacturing environments.

Process optimization in ring rolling is crucial for achieving resource efficiency, competitiveness and the enabling of new products and process strategies. To gain an overview for data-centric optimization approaches and its utilized databases, a systematic literature review is presented. Major aspects of the review are to identify the data generation method, the strategy for this, the regarded process and the underlying optimization. Moreover, the literature has been screened for machine learning applications, a novel approach within ring rolling. The focus in the analysis is set on the utilized databases for optimization and machine learning, especially the data generation and the data sources. Based on the found results, future challenges and possible trends are discussed. The presented review gives a comprehensive overview to display current data-centric trends in ring rolling and aims to provide a foundation for data generation methods for optimization and the application of machine learning.

Design for manufacturability is considered part of the general skill set of mechanical engineers. However, the requirements in practice are significantly more complex. An approach that uses learning algorithms to set parameters and requirements for specific manufacturing processes, materials, tools, and machines related to component geometries is presented. This approach is evaluated, and forecasts are created. A user interface is being developed to integrate the approach into the working environment of design engineers. This interface is designed to equip engineers with the pertinent information necessary for informed design decisions regarding component geometry. The efficacy of this assistance system, encompassing both the system itself and its user interface, was empirically evaluated in the present study. The results indicated that the system was perceived as helpful by the study participants, and the usability of the user interface was favorably assessed.

This paper discusses the influence of joint orientation with non-rotationally symmetric geometry, on load distribution and structural behavior. The focus is on understanding how changes in the alignment of individual joints affect the distribution of load, neighboring joints, and the overall performance of the component. Lap shear specimens with multiple joints arranged in a line are analyzed to explore these effects. Simplified models are used to model the joints in finite element simulations, allowing for efficient yet accurate analysis of the load distribution and structural response under varying joint orientations. Variations in joint orientation result in measurable changes in the distribution of forces on adjacent joints, influencing their behavior and that of the overall assembly. Experimental validation confirms the numerical results, providing deeper insights into the interaction between individual joints and their surroundings. This work contributes to the development of systematic approaches for optimizing the design of components with non-rotationally symmetric joints. The study highlights the importance of considering directional properties of joints in designing structural components.

Supervised learning can be employed to predict product quality, enabling the optimization of parameter configurations before and during the manufacturing process. This approach helps to reduce material and resource waste. The focus of this research is the cold ring rolling process. In this context, process parameters serve as features for the machine learning (ML) algorithm, while form errors are used as labels. However, a sufficiently large database is necessary. 3D FE simulation presents a more cost-effective and efficient method for generating substantial amounts compared to experiments. Both from experiments and simulations, point cloud files of final rings can be obtained, which are essential for the comparative calculation of form errors. A novel method is developed to calculate form errors based on point cloud files. The calculated form errors enable the ML-based quality prediction as labels for supervised ML. The accuracy of the algorithms used for this calculation is subsequently verified, ensuring the reliability of the ML predictions.

Defects such as cracks, overlaps and impressions are prevalent in the manufacturing of press-hardened automotive body components. The prevailing industrial practices rely on manual visual inspections, which are both costly and less effective, thereby posing a risk of undetected defects. To address these challenges, the potential of smart vision sensors for automated component inspection is being investigated. A dedicated test rig was constructed for the purpose of studying the key influencing factors on the output similarity values of the image processing system. These factors include the temperature of the component subsequent to the processing stage and the exposure to the light conditions during the inspection. The performance of the system was evaluated using confusion matrices in order to assess precision and repeatability. For the deformation and crack defect types discrepancies were not observed between the actual and predicted classifications. For the purpose of a practical acceptance of the test system, a left-tailed hypothesis test is carried out for the overlap defect type. The results of the study demonstrate the potential of inspection systems to improve accurate defect detection, thereby paving the way for their implementation in production environments.

This paper focuses on the failure behavior of clinched specimens with various stiffnesses under shear tensile loading. The primary objective is to assess the influence of the specimen stiffness with an arrangement of clinched joints. The specimen stiffness depends on several variables. In addition to the material selection, the specific choice of geometry and the design of the clinched joints must also be taken into account. A number of experiments was conducted to investigate the failure behavior of specimens with an arrangement of three clinched joints under shear tensile loading. These configurations were subjected to shear tensile tests, with force displacement curves recorded for each specimen to provide a detailed characterization of their structural response. The stiffness is modified by altering the specimen width, which has marginal impact on the maximum force. The experimental findings indicate that reducing the specimen stiffness results in a shift in the type of stress, with the failure behavior becoming increasingly influenced by bending stress. These results offer important insights for the design of clinched joint assemblies, indicating that it is feasible to achieve the desired properties by changing the specimen stiffness.

Non-rotationally symmetrical joints can have different properties that can be controlled by the joint orientation. This hypothesis is tested using a Reuleaux triangle joint geometry. A tool design is carried out, followed by a numerical sensitivity analysis of the tool geometry. Initial tools were manufactured for experimental investigations and then adapted based on the findings of the sensitivity analysis. The joints are characterized by micrographs, 3D scans, shear tensile tests, head tensile tests and three-point bending tests and compared with a round geometry. The analysis confirms the hypothesis. Thus, joints with adaptable properties can be produced with one tool set. (c) 2025 The Author(s). Published by Elsevier Ltd on behalf of CIRP. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)

The present study introduces a new experimental setup for the accurate measurement of interfacial heat transfer coefficients during hot stamping with focus on an appropriate temperature measurement. For the evaluation of measurements with varying quenching start temperatures and pressures for the blank material 22MnB5 coated with 150 g & sdot;m2 of Al-Si, Beck's regularization method has proven to be most reliable, whereas the energy conservation method is associated with difficulties in a correct description of the release of phase-transition enthalpies in the blank. For quenching experiments performed at an insert temperature of 700 degrees C and pressures between (1.0 and 18.7) MPa, the time-averaged and maximum values of the measured interfacial heat transfer coefficients ranged from (3.0 to 6.0) kW & sdot;m-2 & sdot;K-1 and (8.1 to 9.5) kW & sdot;m-2 & sdot;K-1, respectively. The obtained results indicate that the setup is applicable for future studies aiming at systematic investigations of factors influencing interfacial heat transfer during hot forming.

In modern lightweight design, mechanical joining methods such as clinching are increasingly used due to their efficiency and suitability for joining dissimilar materials. However, variations in process parameters and material properties can lead to significant deviations in the resulting joint geometry. This study investigates the ability of global polynomial regression models to predict such deviations in clinch joint properties, based on finite element (FE) simulation data and evaluates them through variation simulations and experimental testing. A comprehensive dataset was generated using a validated simulation model to train polynomial regression models. These models were then applied to six distinct clinching process configurations. The metamodels show excellent agreement with the variation simulations, achieving coefficients of prognosis (CoP) above 0.95. Experimental validation using z-scores and Empirical Coverage Probability (ECP) indicates high predictive accuracy for bottom thickness (BT), partially accurate results for neck thickness (NE), and a systematic underestimation of interlock (IL). The predicted 95.5 % confidence intervals are overly conservative for bottom thickness, while for neck thickness and interlock, the intervals are often misaligned with the actual measurements, reflecting biased predictions. The results underline both the potential and the limitations of polynomial regression models for predicting variations in clinch joint properties. While the approach shows promise for designing reliable clinch joints, the study highlights challenges in transferring simulation-trained metamodels to experimental conditions due to uncertainties in the metamodels, the numerical simulations and the experiments.

This study focuses on the phenomenological change in material strength caused by a specific heat treatment and the subsequent analysis of the influence on the clinching process and the resulting joint properties. For this purpose, three series of tests were performed. In the first series of tests, the influence of heat treatment up to 340 °C on the mechanical properties of an age-hardenable AlMgSi alloy was investigated. Holding time and temperature were varied and the material strength was evaluated by tensile and hardness tests. Two strength-increasing and two strength-reducing heat treatment parameters were identified. In the second series of tests, selected heat treatment parameters were applied to a larger number of specimens and the joint strength was investigated by shear and head tensile tests. In the shear tensile test, mainly the properties of the punch-side material have an influence on the resulting joint strength. A change in strength of the die-side material can be neglected. In contrast, the properties of both sheets are important in the head tensile test. The strength of the joint will only increase if the strength of both sheets is increased. In general, a strength increasing heat treatment resulted in higher joint strength. In the third series of tests, the factor of punch displacement was considered, which was demonstrated to directly influence the formation of the clinched joint geometry.

Subject of this study is the usage of macro-structured tools for deep drawing of aluminum alloys at cryogenic temperatures in order to increase the drawing depth. The advantage of the macro-structure is the reduced contact area, which primarily leads to a reduction in frictional forces in the flange area and also minimizes the heat transfer from the tool to the cooled sheet metal. In addition to the geometry of the tools, the heat transfer is also influenced by the thermal properties of the tool materials. Therefore, the tool must be designed to reduce heat flux and thermal diffusivity. In this regard, substitute materials for the tools, specifically polymer, and steel-shielded polymer, were investigated. In this context, the heat transfer coefficients between the tool material and the aluminum blank are determined experimentally. The influence of thermal conditions on the cryogenic deep drawing process are investigated numerically and experimentally. The result of the study is an improved process understanding to ensure the blank temperature in a cryogenic range and increase the resistance against bottom cracks.

Clinching is a mechanical joining technology, in which a mainly form-fit joint is created by means of local cold forming. To characterize the load-bearing behavior of such joints, they are typically analyzed destructively, for example by tensile-shear tests in combination with metallographic sections. However, both the initiation and progress of failure can only be described to a limited extent by this method. Furthermore, these tests allow only limited conclusions about clinch points under in-service loading. More purposefully, clinch points can be analyzed nondestructively by combining in-situ computed tomography (CT) and transient dynamic analysis (TDA). The TDA continuously measures the dynamic behavior of the specimen and indicates failure events like crack initiation, which then can be evaluated thoroughly by stopping the test and performing a CT scan. To qualify the TDA for this task, it is necessary to link the observed damage behavior with specific dynamic characteristics. In this work, the complementation of in-situ CT and TDA is investigated by testing a clinched single-lap tensile-shear specimen made of aluminum. The testing procedure is stepwise: at certain displacement levels, the specimen is investigated by in-situ CT and TDA. While the in-situ CT provides the location, extent, and development of the failure phenomena, the TDA uses this information to evaluate the dynamic signal and detect relevant frequency ranges, which indicate damage events. The results demonstrate, that failure initiation and progression can be analyzed efficiently by combining both measuring systems. The TDA reliably detects relevant signal changes in the monitored frequency band. By means of in-situ computed tomography, the corresponding failure phenomena can be described in detail, enhancing the understanding of the load-bearing and deformation behavior of clinch points. The concatenation of characteristic signal changes and observed failure phenomena can henceforth be transferred to analyze complex structures during operation nondestructively by TDA.

Reaching robustness of forming processes is one of the main challenges in manufacturing technologies. Volatile material properties [1] and fluctuations in process parameters often result in insufficient part quality and rising production costs, especially in production chains [2]. As a common solution for this problem, numerical simulations are used to consider scattering of material parameters and define a useable process window with a specified safety margin. A new approach is the use of macro-structured tools. Result shown here indicate, in addition to the general enlargement of the useable process window [3], a high tolerance toward volatile material properties and variations of the blank. The produced parts show consistent quality with a reduction in residual stresses [4]. To examine the robustness in particular, virtual studies for steel alloy DX54 with varying alignment of the blank were performed. The resulting properties, such as the blank draw-in, thinning and springback, were examined in predefined positions on the blank to compare the experiments. The enlarged process window will be analysed and a strategy for a deeper understanding of the forming process will be generated, which will sustain the design of robust tools and processes.