Elast-plasto deformation of metal, Ductile fracture analysis（Matsuno）
Establishment a Foundation for AI Assimilation Analysis of Simulation/Observation Images
Structural metal materials frequently undergo plastic deformation during their processing to structual componets and subsequent performance. Steel materials for automotive applications are processed through plastic deformation during manufacturing and absorb energy through plastic deformation during automobile crush.Material failure must be prevented in these circumstances. While clearly predictable fractures can be readily addressed, fractures that occur occasionally at frequencies of approximately once per several hundred instances present significant challenges. These cannot be detected through preliminary material testing, yet their frequency is not negligible. In numerous cases, the causes remain unidentified, leaving limited options for remediation.
Our laboratory attempts to reproduce stress and strain distributions immediately preceding material failure through assimilation techniques integrating measurement and observation via AI technology. This enables identification of causative factors in materials that experience incidental material failure. The methodology involves conducting deep learning as an inverse deformation simulation, then utilizing this deep learning model to inversely deform the microstructure near the ductile fracture zone, thereby visualizing stress and strain distributions. This technology facilitates determination of fracture initiation sites within the microstructure and identification of factors generating abnormally elevated stress or strain conditions conducive to fracture. Recent accomplishments include successful rendering of stress-strain distributions on deformed regions of microstructures.
Current research focuses on reconstructing the microstructure at fracture sites and depicting stress-strain distributions and fracture resistance variation distributions on the reconstructed microstructure.
Analysis Examples
Identification of Post-necking flow stress curves and Ductile Fracture Loci in Ultra-High Tensile Steel Sheets
Post-necking flow stress curves are essential for predicting press forming failure and automobile crash-worthiness performance.Our laboratory has developed a technique for identifying flow stress curves in large plastic strain regions up to the point immediately preceding ductile fracture by combining shape measurement and FE simulation. Anticipating application to ultra-high tensile steel sheets for automotive use, we conduct evaluations using small round bar tensile specimens with 1 mm diameter extracted from thin steel sheets.
While measuring flow stress curves in large plastic strain regions, we have successfully measured flow stresses in notched specimens that simulate a variable stress states. Through these measurements, we have newly identified a "weakening" phenomenon in 1.5 GPa-class ultra-high strength steel sheets, wherein work-hardening decreases depending on stress state. Additionally, we are identifying ductile fracture loci, which are uncommon in thin steel sheets. For 980 MPa-class and 1.5 GPa-class ultra-high strength steel sheets, the only publicly available ductile fracture loci are data from our laboratory.
Similar approaches are being implemented for thinner high-strength aluminum and other lightweight metal materials. Furthermore, while we currently identify deformation resistance by assuming functions, we are also attempting non-parametric identification using the aforementioned AI assimilation.
Analysis Examples
Digital Twin of Shear Cut Surface and Visualization of Residual Stress Distribution
Thin sheet materials delivered from material manufacturers are in roll form similar to toilet paper. Therefore, they must be cut into specified work-piece shapes before press-forming. Hole punching or blanking using punches is termed shear cut, and most steel cutting utilize this method.Although our laboratory focuses on analyzing ductile fracture phenomena, we apply these insights to shearing simulation. Shearing achieves cutting by controlling ductile fracture using punches.
Existing simulations primarily "predict" sheared surface characteristics from anticipated tool conditions; however, our laboratory addresses the visualization of residual stress from post-sheared surface morphology. Tensile residual stress contributes to fatigue fracture and delayed fracture (a phenomenon where fracture occurs over time without applied load) originating from sheared surfaces. Currently, X-ray methods are commonly employed for residual stress measurement, but spatial resolution sufficient to identify fracture causative factors remains unachieved.
As simulation enables stress output through arbitrarily fine modifications, our laboratory reproduces sheared surface morphologies nearly identical to actual specimens and visualizes residual stress distributions on actual sheared surfaces. This process also derives ductile fracture loci, constituting a dual-purpose technology applicable to conventional prediction methods.
This analysis also incorporates attempts at non-parametric shape assimilation using artificial intelligence.
Analysis Examples
Digital Twin of JSC980Y Sheared Cross-section and Inversely Identified Ductile Fracture Locus（Matsuno et al., 2025, Matec Web Conf.)
放射光X線イメージング，水素脆性に関する研究（清水）
これまでの主な共同研究先：物質・材料研究機構，日本原子力研究開発機構，東北大学，富山大学，九州大学，京都大学
高強度アルミニウム合金の水素脆性
Al-Zn-Mg合金（7000系合金）は代表的な高強度アルミニウム合金であり，航空機・新幹線・自動車など，強度と軽さの両方が要求されるところに実用されています。世界的な燃費規制や環境負荷の低減といったニーズに応えるためには，材料の高強度化が必要になります。ただし，一般に構造用金属材料は，高強度になるほど水素によって脆化しやすくなります。これは，単純に静的な強度を上昇させても，水素脆化を克服しない限りは真の高強度化を達成できないことを意味します。我々の研究グループでは，放射光施設SPring-8を使って合金の水素脆化挙動を「その場観察」します。そして，どのような力学状態で・どのような水素濃度で・いつ・どのタイミングで・どこから破壊が始まったのか，三次元可視化します。このように，SPring-8で三次元的に可視化し，様々な材料挙動を明らかにすることを得意にしています。
研究事例
For participants of our labo.
If participants aim to enroll master students in October, the process would be as follows:1) August: Travel to Japan on a tourist visa for a short-term stay, take the entrance exam, and then return to your home country.
2) September: Upon passing the exam, obtain the necessary documents issued by the university and apply for a student visa. The staffs will assist the participants in applying for their residence status in Japan. The issuance of documents typically takes about three weeks, which should allow sufficient time for enrollment.
3) October: Enroll in the graduate program.
This is the most expedient route, contingent upon passing the entrance exam. The exam is typically taken in Japan, and most applicants arrive early to complete the payment procedure from within the country. The staffs can assist with this process during the short stay.
Please note that financial assistance for these processes cannot be provided.
The university recommends that students enroll as research students for about six months to acquire Japanese language skills and basic academic knowledge. For more information, please refer to https://www.ciatu.tottori-u.ac.jp/en/study-tottori-enroll. In this case, the process would be:
1) Apply as a research student. The university will deliberate on the participant application.
2) Obtain a student visa and residence status, following the same procedure as the October enrollment.
3) Arrive in Japan and prepare for the graduate school entrance exam in December.
4) Enroll as a regular student from April.
Please note that the financial assistance for these processes cannot be provided.
For those who wish to enroll as doctoral students, several financial support options are available, e.g.,JSPS Invitational Fellowships. All of these require application and successful selection. Please contact the laboratory staff for details regarding enrollment.