Center for Mathematical Science and Advanced Technology (MAT)

Seminar Schedule

[FEAT-MAT Joint Seminar]

Date:

2021/09/06(Monday) 15:00 - 16:30

Language:

English

Speaker:

Hirobe Sayako

Title:

Numerical analysis of dynamic fracture in a residual stress field

Abstract:

Residual stress field is a self-equilibrated state of stress induced by the inhomogeneous distribution of the inelastic deformations, temperature and moisture gradients or local volume changes due to phase transition. The high level of tensile residual stress often leads to dynamic fracture resulting in the instantaneous and catastrophic destruction of the materials without any persistent external loading. This dynamic fracture in a residual stress field brings substantial theoretical complexity because we have to solve the mutual interaction among crack propagation, change in the residual stress field, and generation of the elastic wave. In this seminar, I will present a mathematical model and a numerical analysis method for dynamic fracture in a residual stress field. We firstly formulate the Hamiltonian for the dynamic behavior of the solid continuum with residual stress field. Introducing the appropriate treatment of the fracture in a residual stress field, this Hamiltonian formulation provides the rigorous evaluation of (i) release and redistribution of the residual stress due to dynamic propagation of the cracks and (ii) the effect of the elastic wave on crack propagation. For evaluating the validity of the proposed mathematical model and numerical analysis method, we performed experiments and numerical analyses of the self-driven fracture process in chemically tempered glass sheets with different residual stress profiles. Our methodology is applicable regardless of materials and scales, ranging from nano crystal glass to earthquake faults.

Upcoming Seminar schedule
※title and abstract of each seminar will be informed when they are decided.

[MAT Seminar]

Date:

2021/7/21 (Wednesday) 13:00 - 15:00

Language:

English

Speaker:

Gen Li

[MAT Seminar]

Date:

2021/7/7 (Wednesday) 13:00 - 15:00

Language:

English

Speaker:

Minoshima Takashi

[MAT Seminar]

Date:

2021/6/23 (Wednesday) 13:00 - 15:00

Language:

English

Speaker:

Maruoka Hirokazu (CeBN, MRU)

[MAT Seminar]

Date:

2021/6/16 (Wednesday) 13:00 - 15:00

Language:

English

Speaker:

Arthur Bauville

[MAT Seminar]

Date:

2021/06/09(Wednesday) 13:00 - 15:00

Language:

English

Speaker:

Ettore Barbieri

[MAT Seminar]

Date:

2021/04/21(Wednesday) 13:00 - 15:00

Language:

English

Speaker:

Takahashi Tomoko (ASTER Program, X-star)

Title:

Laser spectroscopic techniques for in-situ chemical measurements of solids in the deep-sea environment

Abstract:

Laser spectroscopy is a non-contact, label-free chemical analytical technique, and has a large potential to perform chemical measurements of solid materials such as rocks and particles in-situ in the deep-sea environment. These are typically analysed by sampling-based methods, and in-situ measurement techniques could significantly increase the spatial and temporal resolutions of data acquisition.
In this presentation, the presenter will introduce three laser spectroscopic techniques for in-situ applications, and discuss data interpretation methods:
(i) Particle measurements using holography and laser Raman spectroscopy
By integrating digital holography and laser Raman spectroscopy, morphological and molecular information can be obtained for a floating particle in a large volume of water. Preliminary results of classification of particle types using multivariate analysis and machine learning will be shown.
(ii) Continuous monitoring of microplastics using coherent anti-Stokes Raman scattering (CARS)
Fast chemical imaging of flowing microplastic and bio-organic particles with the size of 20-40μm can be performed by simultaneous detection of CARS and fluorescence. Methods for reconstruction of flowing particle images will be discussed.
(iii) In-situ elemental analysis of hydrothermal deposits using laser induced breakdown spectroscopy
Development and deployments of the deep-sea LIBS anlyser will be introduced. In-situ grid measurement surveys were performed in a hydrothermally active region at the depth more than 1000m, and heavy metals were successfully detected. Quantitative results using partial least squares regression and artificial neural networks will be shown.

[MAT Seminar]

Date:

2021/03/31(Wednesday) 13:00 - 15:00

Language:

English

Speaker:

Nozaki Tatsuo(MRU Submarine Resources Research Center), Kitada Kazuya(X-star, ASTER Program)

Title:

Long-term monitoring of pressure and temperature of hydrothermal fluid using deep-sea artificial hydrothermal vent and Kuroko-ore cultivation apparatus

Abstract:

The IODP Exp. 331 in 2010 drilled the deep-sea active hydrothermal filed at the Iheya-North Knoll, middle Okinawa Trough (Takai et al., 2011, 2012). Due to its drilling operation, four “artificial hydrothermal vents (AHVs)” that were drill holes with hydrothermal fluid discharge were formed. Unnaturally rapid growth of chimney on AHVs inspired us the possibility of the Kuroko-ore (黒鉱; hydrothermal sulfide) cultivation likes as aquafarming (Kawagucci et al., 2013; Nozaki et al., 2016). However, long-term direct measurements targeting the hydrothermal vent and flux data of metal elements which is essential to evaluate the potential of the Kuroko-ore cultivation project are still limited. Thus, during the two drilling cruises performed by using D/V Chikyu in 2016 (cruise CK16-01 and CK16-05), totally three Kuroko-ore cultivation apparatuses were installed at the Original Site of the Iheya-North Knoll (Hole C9024A) and Noho Site of the southern flank of the Iheya Minor Ridge (Holes C9017A and B), middle Okinawa Trough. The Kuroko-ore cultivation apparatus is equipped with two P/T sensors, flowmeter, load cell and their loggers to monitor the pressure, temperature and flow rate of hydrothermal fluid together with the weight of hydrothermal precipitate.
In this seminar, Dr. Nozaki will introduce this project briefly as well as explanations about research purpose, cruise operation, specification of the Kuroko-ore cultivation apparatus and general P/T data of three Kuroko-ore cultivation apparatuses and one flowmeter. Then, Dr. Kitada will do the presentation about detailed analysis about the P/T sensor data equipped with the Kuroko-ore cultivation apparatus at Hole C9017B.

[MAT Seminar]

Date:

2021/03/24(Wednesday) 13:00 - 15:00

Language:

English

Speaker:

Nomura Shun(MAT)

Title:

My research life in MAT

Abstract:

In the end of this March, I will leave MAT, JAMSTEC and work in a new affiliation. I introduce the new affiliation from April and explain my motivation in it. Then I will briefly review my research topics, which I fulfilled or completed in JAMSTEC such as theoretical explanation of soil media, experimental investigation of turbidity currents, and exposing test of concrete specimens in the deep sea. After that I will present my research plan which I want to tackle from April. As the latest topics, I will explain the progress the transparent ground technique induced by the refractive index matching and its possibility to extend the field of geoengineering.

[MAT Seminar]

Date:

2021/02/24(Wednesday) 13:00 - 15:00

Language:

English

Speaker:

Maruoka Hirokazu (CeBN)

Title:

The concept of intermediate asymptotics and a dynamical impact of solid sphere onto the PDMS militextured surface

Abstract:

In the field of mechanics of continua and soft matter physics, we frequently encounter the mixture of physical properties and the scale-local behavior which changes its scaling relation depending on the spatial and temporal scale. These characters are not only of great interest and attraction, but also the difficulties on the mechanics of continua and soft matter physics. In order to deal with these difficulties, I believe and expect that the concept of intermediate asymptotics, which had been elaborated by G.I. Barenblatt with Ya. B. Zeldovich, would be potentially effective tool. Dimensional analysis extracts the degree of mixture of forces as dimensionless numbers (e.g. Reynolds number are the fraction between inertial force and viscous force), which may be expected to be useful to express the mixing physical properties in soft matter. Scale-local behaviors can be understood as the asymptotic behaviors of global functions, which corresponds to the Barenblatt's formalization of intermediate asymptotics.

In this seminar of first part, I will show the characters of phenomena of soft matter physics briefly and the concept of dimensional analysis and intermediate asymptotics to show the potentiality of their application.

In second part, I will show the application of this idea to the dynamical impact of solid sphere onto the PDMS militextured surface. The crossover of the scaling law between the maximum deformation and the impact velocity was observed in the experiment. I have applied the dimensional analysis to find that the problem belongs to the self-similarity of the second kind. The crossover was explained as the result of the competition between two scaling laws which appears from the expansion of the global dimensionless function, of which each scaling laws can be considered as the intermediate asymptotics.

[MAT Seminar]

Date:

2021/02/17(Wednesday) 13:00 - 15:00

Language:

English

Speaker:

Hidaka Mitsuko (Information Engineering Program)

Title:

Application of deep-learning at JAMSTEC

Abstract:

Deep-learning is known as the state-of-the-art underlying technology of AI (Artificial Intelligence). The presenter, who has a background in marine biology, is now involved in some different deep-learning projects, such as super-resolution of bathymetric maps and beached marine litter detection. In this seminar, the presenter will introduce these two topics and the results gained through deep-learning. Any comments and opinions are most welcome.

(1) Super-resolution of bathymetric maps
JAMSTEC launched an interdisciplinary project “mathematical seafloor geomorphology” in 2019. In the project, a technique that can produce high-resolution bathymetric maps from low-resolution maps has been studied. Five deep learning architectures are applied to the same dataset, and the results are compared.
(2) Beached marine litter detection
Beached marine litter is harmful to both finance and the environment. Establishing technologies that can quantify beached marine litter will contribute to mitigation strategies and scientific understanding of the litter. We applied the pixel-wise classification method “semantic segmentation” for beached marine litter estimation. The preliminary results from the study will be introduced.

[MAT Seminar]

Date:

2021/01/20(Wednesday) 13:00 - 15:00

Language:

English

Speaker1:

Ohkouchi Naohiko (Biogeochemistry Research Center)

Title:

What is biogeochemistry?

Speaker2:

Isobe Noriyuki (Biogeochemistry Research Center)

Title:

Mechanical stress of hydrogel by squeezing or absorbing water

Abstract:

Hydrogel is a porous material filled with water. It contains much amount of water (more than 90%) while retaining high integrity. Thanks to these advantages, the hydrogel has been a fascinating material in many applications ranging from biomedicines to foods. For engineering application, the improvement of mechanical property is a major challenge. For this, strengthening the gel matrix (network structure) has been intensively studied, while the effect of water on the mechanical property has rather been underestimated. Here, I show that water can govern the mechanical property of hydrogel based on cellulose (upon squeezing water) and chitin (upon absorbing water).

i) Squeezing water

I reported that a hydrogel made of cellulose can show poroelastic behavior: compressive elastic modulus changes upon the compression speed (Isobe et al. 2018, JTICE). To further understand the poroelastic behavior of cellulose hydrogel, a series of compression tests were performed. Compressive Young's modulus increased up to 4-times larger depending on the compression speed. In addition, the compressive Poisson's ratio, which was calculated from the amount of oozing water, was found to be 0.07, indicating that the horizontal expansion upon vertical compression of the cellulose hydrogel was small. Therefore, it is concluded that the poroelastic response of cellulose hydrogel was given by large amount of water passing through the small pores inside the regenerated cellulose hydrogel.

ii) Absorbing water

Other fascinating property of hydrogel is a volume change induced by external stimuli such as pH, temperature, etc. I found that chitin gel can swell up from the completely dried state by being immersed in acid solution (Isobe et al. 2020, ACS Appl. Polym. Mat.). Swelling materials can serve as a prestressing agent such as expansive cement. To understand the swelling pressure of the chitin hydrogel, here I monitored the time-course change in swelling pressure. Dried chitin hydrogel swelled by the acetic acid solution, and the swelling pressure reached up to 1 MPa in . Moreover, the swelling pressure was kept constant for at least 2 days.

[MAT Seminar]

Date:

2021/01/13(Wednesday) 13:00 - 15:00

Language:

English

Speaker:

Patrick Martineau and Ratnam Jayanthi (APL)

Title:

Malaria prediction efforts at the application laboratory

Abstract:

In this presentation, we are going to report on our progress, ongoing challenges, and future work for the prediction of Malaria in South Africa.

In the first part, we will discuss the development of an early warning malaria prediction framework utilizing machine learning techniques to forecast malaria from observed climate variability. To produce forecasts of malaria incidence with lead times ranging from 3 months to about a year, a set of classifiers are trained on 23 years of historical global climate and South African malaria incidence data to recognize the past influence of climate on the spread of the disease. Through hindcast experiments, we find that the prediction accuracy is adequate during the austral summer when malaria frequency is largest. Among the identified predictors, sea surface temperature variability associated with the El Niño–Southern Oscillation (ENSO) and the Indian Ocean Subtropical Dipole (IOSD) provide an important source of predictability for the classifiers. Challenges associated with the presence of outliers and trends for the identification of useful predictors will be discussed.

In the second part, we are going to discuss the results of a malaria model VECTRI (VECtor borne disease community model of ICTP, TRIeste, Italy), which was used in the project to simulate the number of malaria cases over the north-eastern provinces of South Africa. VECTRI is a dynamical model mimicking the life cycle of malaria and is mainly driven by the climatic parameters, precipitation, and surface air temperature. The model was used to forecast the malaria cases at seasonal time scales using the WRF downscaled SINTEX-F forecasts and also to evaluate the changes in the number of malaria cases over South Africa under the RCP 8.5 climate change scenario by forcing the VECTRI model with the climate change projections of the CORDEX Africa regional model. The results indicate the VECTRI model to be suitable for simulating and predicting the malaria cases over South Africa. However, the model has a bias in simulating a higher number of cases over certain regions of South Africa which needs to be addressed in the future.

©JAMSTEC, www.jamstec.go.jp