What type of study is this?

September 10, 2025Open Access

Study of the influence of the computational grid density on the heat transfer coefficient in numerical modeling of the transverse flow around a tube bundle of the corridor type

Key Points

High grid detail improves accuracy, with modeled heat transfer coefficients closely aligning with experimental data.
The average heat transfer coefficient varied among six distinct grid configurations, highlighting the influence of grid density.
A systematic comparison of numerical results with the Baer method confirmed the superiority of detailed models for accurate simulation.
Results suggest detailed computational grids are essential for effective modeling of coupled heat exchange in heat exchangers.

Abstract

Justification. The article considers numerical modeling of the transverse flow of a five-row corridor-type tube bundle with an air flow during coupled heat exchange. The studied model is typical for air-cooled heat exchangers such as air condensers, oil coolers or cooling towers. The outer surface of the pipes had a temperature higher than the temperature of the air flow. The air velocity corresponded to the Reynolds number in the narrow section of the tube bundle, typical for heat exchangers of the above type. The results of numerical modeling are determined by the quality of the computational grid. At the same time, the larger the calculation cells and the smaller their size, the more computing and time resources are required to solve the problem. However, the results obtained are closer to a field experiment. Goal. Determination of the level of detail of the computational domain, at which it is possible to obtain correct results of numerical simulation of coupled heat transfer. Results. The influence of the quantitative characteristics of six design grid variants on the value of the average heat transfer coefficient is estimated. The variants differed in the degree of detail: the number and size of cells (from large to small). In each case, the values of the average heat transfer coefficient are obtained, determined by the degree of error of the numerical experiment. The data obtained were compared with the data obtained when solving the criterion equations of convective heat transfer using three methods. Conclusion. The error of finding the heat transfer coefficient by the considered methods is determined. It is established that the results of numerical modeling with a high degree of detail are close to the results obtained in the calculation using the Baer method. The results obtained can be useful in numerical modeling of flows in heat exchangers with transverse flow around tube bundles.

Read Full Paperexternally

Perguntar à IA

Bookmark

View Full Paper