[
home]
[
Personal Program]
[
Help]
tag
10:20
20 mins
Artificial Neural Network-Based Design of Air Duct Outlets for Household Refrigerators
Eunseop Yeom, Heewook Jung, Yongbum Cho, Hoyoon Kim
Session: Control of heat transfer I
Session starts: Wednesday 05 November, 10:20
Presentation starts: 10:20
Room: Lecture room A
Eunseop Yeom ()
Heewook Jung ()
Yongbum Cho ()
Hoyoon Kim ()
Abstract:
The performance of a household refrigerator is highly influenced by the internal temperature distribution and the
efficiency of airflow. In particular, uniform temperature distribution within the refrigerator plays a critical role in
maintaining energy efficiency and preserving food freshness, making its optimization essential to meet consumer
demands. Computational Fluid Dynamics (CFD) has long been a vital tool for visualizing thermal and flow fields
and optimizing refrigerator designs.
Oh, M. J. conducted a three-dimensional steady-state numerical simulation to analyze the characteristics of cold
airflow inside a refrigerator, accurately modeling internal flow and heat transfer involving fan and evaporator
operation. (1) Similarly, Yoo, J. H. carried out a steady-state simulation aimed at improving airflow uniformity by
redesigning the air duct structure. (2) In addition, Wang, L. performed CFD simulations under transient conditions
to evaluate three-dimensional airflow and temperature variation inside a freezer cabinet, thereby optimizing
temperature uniformity and cooling performance while analyzing internal flow patterns and thermal behavior. (3)
Recently, advances in artificial neural networks have opened new opportunities for refrigerator design
optimization. AI models can learn from large-scale simulation data to identify optimal design parameters that
maximize temperature uniformity and cooling efficiency. This study integrates conventional CFD techniques with
AI-based design optimization methods to analyze internal temperature characteristics and propose optimal design
solutions for improved refrigerator performance.