A typical case of die failure owing to low-cycle fatigue fracture during hot forging of a complicated auto-part is presented. The reason for the fracture is revealed through precise rigid-thermoviscoplastic finite element analysis of the three-stage hot forging process and die structural analysis. The fracture occurred not at the very corner, where stress concentration may occur, but near the corner, a little apart from the very corner. The experiment shows that several cracks occur in parallel, indicating the die fracture is owing to the cyclic loading. The FE prediction shows that the high principal stress band coincides with the crack and that the maximum principal axis is perpendicular to the crack. The cycle of maximum principal stress is traced to investigate the low-cycle fatigue fracture, showing that the amplitude of stress cycle and mean stress at the maximum principal stress point in the fracture region back up the low-cycle fatigue fracture.

In manufacturing a tripod housing using multi-stage cold and warm forging process, a process modification example is presented for approaching net-shape forging. The process modification in this study is mainly performed on target shape improvement at warm upsetting, and the target layouts of the intermediate billets of the following warm backward extrusion and the cold sizing are also changed minutely. Using mechanical and thermal properties of C48P medium carbon steel bar, elasto-plastic finite element simulation of the warm upsetting is re-started based on the prior results from the ejection operation in the second forward extrusion, and then a series of numerical simulations related to the warm backward extrusion and the cold sizing processes are followed. The numerically predicted results in view of the deformed feature before and after the process modification are compared with the tripod housing prototypes produced experimentally. As a result, the forged configuration of the tripod housing by the process modification is revealed to be improved.

This study aims to develop both a mathematical and a finite element method (FEM) model for predicting work-roll deformation in a four-high hot rolling mill, specifically targeting improved crown control accuracy. To do this, the theoretical model extends the conventional slit-beam deflection model to comprehensively incorporate the complex interactions of roll deflection, backup-roll support, roll-to-roll/strip flattening, initial crown geometry, and external bender forces. To enhance computational efficiency for on-line implementation in industrial manufacturing systems, the nonlinear equations are effectively linearized using a first-order Taylor expansion. The developed models show excellent agreement with the FEM simulation results across various input loading profiles, initial crown conditions, and bender force levels. The proposed linear approximation model offers a reliable and computationally efficient analytical tool highly suitable for implementation in industrial applications for on-line crown control, sensitivity analysis, and optimal roll-gap design.

Sheet metal forming is widely applied in manufacturing components such as automotive panels and battery cans, enabling high-rigidity products with minimal welding. The rolling process used to manufacture these sheets induces grain orientation, leading to material anisotropy that varies with material type and thickness. Typically, the Lankford value (R-value) is determined via Digital Image Correlation (DIC) by measuring longitudinal(ε_l) and transverse(ε_w) strains during uniaxial tensile tests, utilizing data from the uniform elongation region (10% ~ 15%). However, high-speed tensile tests, essential for simulating actual forming processes, require limited parallel lengths due to equipment and dynamic equilibrium constraints. This limitation reduces the uniform elongation zone, complicating the acquisition of reliable R-values. To address this, this study designed a specimen geometry to secure a sufficient uniform elongation region under highspeed conditions for accurate anisotropy measurement. Using this design, we compared anisotropic characteristics across strain rates from quasi-static to high-speed regimes (1 /s ~ 100 /s). This research successfully obtained material properties relevant to actual forming speeds, contributing to enhanced precision in future forming simulations.

Crack formation after Flow drilling screw (FDS) fastening remains a critical issue in composite structures, particularly when heterogeneous laminate configurations are applied. In this study, the crack initiation mechanism in composite laminates subjected to FDS fastening was systematically investigated through mechanical testing, cross-sectional observation, and X-ray computed tomography (CT) analysis. Composite specimens consisting of chopped strand and woven layers were prepared with different laminate configurations to clarify the influence of fiber distribution around the fastening area. The mechanical test results revealed that chopped strand-based laminates exhibited relatively vulnerable behavior under localized stress conditions, showing a reduction in maximum fastening load of approximately 10-20% compared to laminates with woven layers at the outer region. Cross-sectional observations confirmed localized deformation concentrated at the outer region of the fastening area after FDS installation. CT analysis further demonstrated that cracks preferentially initiated and propagated along the outer region dominated by chopped strands. In contrast, specimens with woven layers positioned at the outer region showed sound fastening behavior without crack formation, even under identical FDS fastening conditions. These comparative results indicate that crack formation after FDS fastening is governed not by the fastening process itself, but by the laminate configuration and fiber distribution around the joint. Specifically, the placement of chopped strands at the outer region promotes stress concentration and crack initiation, whereas woven layers effectively suppress crack formation by enabling more continuous load transfer. The findings of this study provide practical design guidelines for optimizing laminate configurations in FDS-fastened composite structures to enhance joint reliability.

In this study, a combined experimental and numerical approach was employed to investigate the dynamic tensile properties of STS430 thin sheet specimens under high strain rate conditions. Although the Split Hopkinson Tensile Bar (SHTB) test is widely used to characterize dynamic tensile behavior, obtaining reliable data directly for thin specimens with a thickness of less than 1 mm is challenging due to the low Signal-to-Noise Ratio (SNR) caused by the significant impedance mismatch between the specimen and the pressure bars. To overcome these limitations, a constitutive model was constructed by integrating the Swift strain hardening model and the Cowper-Symonds strain rate factor model based on quasi-static and dynamic tensile test results. Furthermore, Finite Element Analysis (FEA) using LS-DYNA was performed to simulate the specimen clamping force and impact conditions, followed by a comparative analysis with the experimental data. The results showed that the reflected waves and flow stress curves derived from the numerical analysis effectively reproduced the trends of the experimental data, demonstrating high consistency, particularly in the work-hardening region. This study confirms that the integrated experimental-numerical analysis is effective for characterizing the strain rate dependency of thin sheet specimens and is useful for deriving reliable parameters for constitutive models.

The use of high-strength steel sheets in automotive body and chassis components continues to grow to reduce vehicle weight while maintaining crashworthiness. As sheet strength increases, variability in material properties and forming conditions tends to increase as well, which can raise process sensitivity and compromise production stability. This study evaluates the robustness of two-stage split drawing process designs for an automotive crossmember against variations in material properties and process conditions. Yield strength, ultimate tensile strength, and sheet thickness were selected as material-related noise factors, whereas pad force and the friction coefficient were considered process-related noise factors. Finite element simulations were conducted to quantify key forming defects, including thinning, wrinkling, and springback, under the prescribed noise-factor variations. The resulting distributions of defect responses were assessed using the interquartile range (IQR), total range, and skewness. Relative to the initial design, the optimized design showed markedly reduced dispersion in defect responses, indicating a significant improvement in forming robustness.