九州大学伊都キャンパスにて不定期に現象数理セミナー（NPＡセミナー）を開催しています。応用や現象に関連した解析の話題を広く取り上げる予定です。多くの方々のご参加をお待ちしております。
セミナー幹事：手老篤史、福本康秀（九大MI研究所）
連絡先：手老 篤史（tero(at)imi.kyushu-u.ac.jp）
現象数理セミナー予定
通常 木曜不定期15:30-17:00・
（日時・場所は変更になることがあります）
ホームページ担当者変更により、アドレスが変わりました。
以前のホームページアドレスは下のリンクより移動してください。
https://sites.google.com/site/npaseminar/
*************************************************************
第113回現象数理セミナー
Date: 16:30--17:30, Feb 17 (Tue), 2026
Venue: IMI Conference room (W1-D-414)
Speaker: Fermin Franco Medrano (Autonomous University of Baja California)
Title:
Mathematical Modeling of Atomizing Liquid Jets:
From Conservation Laws to Modern Extensions with Machine Learning
Abstract:
Atomizing liquid jets are central to numerous industrial processes,
including fuel injection in automotive and aerospace engines, spray
coating, and medical atomizers. This presentation reviews a family of
one-dimensional two-phase fluid models developed in collaboration with
Prof. Yasuhide Fukumoto, which describe the macroscopic dynamics of
full-cone turbulent round jets in a simple yet comprehensive
manner. The models rely on the assumptions of statistical stationarity
and locally homogeneous flow, approximating the jet as a mixture of
liquid and entrained gas phases in dynamic equilibrium. By imposing
conservation of initial mass flux, momentum flux, and/or total power
flux (with optional partial loss factors), the governing equations
admit implicit analytical or numerical solutions for composite density
and velocity as functions of axial distance from the nozzle. From
these, key quantities such as dynamic pressure and gas entrainment
rate are derived. In particular, the local entrainment rate
coefficient is obtained analytically in the far-field limit, depending
solely on the cone apex angle, and predictions compare favorably with
experimental data for both single-phase turbulent air jets and
high-speed atomizing liquid jets. The talk will highlight foundational
aspects from our earlier works, including stability analyses of
evaporating jets, spray impact modeling, and conservation-based
dynamical frameworks. Looking ahead, modern data-driven techniques,
such as surrogate models based on machine learning, offer exciting
potential to enhance parameter exploration, uncertainty
quantification, and integration with high-fidelity simulations or
experimental datasets. These extensions remain under active
consideration and could bridge classical analytical models with
emerging computational tools for more robust industrial
applications. This work reflects ongoing efforts to advance
mathematical fluid mechanics in the spirit of mathematics-for-industry
collaborations.
****************************************************************
セミナーHP:
https://sites.google.com/site/npaseminar2/
セミナー幹事：手老篤史、福本康秀（九大MI研究所）
連絡先：手老 篤史（tero(at)imi.kyushu-u.ac.jp）
_______________________________________________
Npa mailing list