The University of Tokyo · Engineering
Professor Kazuki Shibanuma's research lab specializes in multiscale materials modeling with a focus on predicting the mechanical behavior of metallic materials, particularly steels, under complex loading conditions. The lab develops advanced computational models that bridge microstructural features—such as grain boundaries, phase distribution, and heterogeneous microstructures—with macroscopic fracture and fatigue performance. Key research directions include ductile and cleavage fracture mechanisms, fatigue life prediction in welded joints, and the effects of microstructure evolution on material toughness and reliability. The lab uniquely combines experimental validation with innovative modeling strategies to address real-world engineering challenges in structural materials.
Figures are computed from collected data and may differ slightly.
Although various effective models for simulating ductile fracture in metallic materials have been established, the methods to identify the required parameters have scarcely been clarified. The difficulty for parameters identification of ductile fracture models is the actual bottleneck for accurate numerical predictions in ductile fracture simulation. In this paper, a strategy to efficiently identify the post-necking strain behaviour and the ductility diagram parameters is proposed by a hybrid ex
It is known that a cracking of brittle phase such as cementite works as a trigger of cleavage fracture initiation. This study shows a microscopic observation of cracked cementite and its quantitation of the cracking nucleation in ferrite-cementite steels. Seven steels with various sizes of microstructures are produced by laboratory scale vacuum melting and rolling. The cementite particle thickness was measured by a SEM observation and an image analysis. Tensile tests using circumferential notche
• The fatigue performance of welded joints is predicted by multiscale model simulations. • Microstructure distribution, strength distribution in HAZ and 3D weld toe shapes are considered simultaneously in the multiscale model. • The modelling strategy for multiple crack initiation, growth, and coalescence behaviour is proposed. • The proposed strategy for predicting the fatigue performance of welded joints is validated by experimental results. This study predicts the fatigue performance of welde
Multiscale modelling strategy for predicting fatigue lives and limits of steels is proposed based on a generalised method for evaluating grain boundaries’ (GBs) effects on fatigue crack growth. The proposed strategy is a modification of our previous works that extends the applicability from only low-grade ferrite steels to high-grade bainite steels. A microstructure model was developed considering distances between GBs and misorientations between adjacent grains. The strategy was validated by co
Open papers in the app to read, cite, and organize with AI.
