4.7 Article

Multi-purpose prediction of the various edge cut twisted tape insert characteristics: multilayer perceptron network modeling

Journal

JOURNAL OF THERMAL ANALYSIS AND CALORIMETRY
Volume 145, Issue 4, Pages 2005-2020

Publisher

SPRINGER
DOI: 10.1007/s10973-021-10904-1

Keywords

Cut twisted tape; Artificial neural network; Heat exchanger performance; Heat transfer augmentation; Friction factor

Funding

  1. Brain Pool Program through the National Research Foundation of Korea (NRF) - Ministry of Science and ICT [NRF-2020H1D3A2A01104062]
  2. National Research Foundation of Korea (NRF) - Korean Government (MSIT) [2020R1A5A8018822]

Ask authors/readers for more resources

The current study used artificial neural network modeling to analyze the Nusselt number and friction factor of a heat exchanger with twisted tapes, showing that the model performed exceptionally well in terms of time and cost, and accurately predicted heat transfer enhancement based on the specification parameters of the cuts.
In the current study, the artificial neural network modeling of the Nusselt number and friction factor of a heat exchanger, equipped with the twisted tapes to enhance the heat transfer capacity, has been done. In the modeling procedure, alongside the Reynolds number, four geometrical characteristics of the twisted cut tapes, including the depth of cut, the area of cut, the perimeter of cut, and the pitch length of the twisted tapes, have been considered. The required data were gathered from both experiments and the CFD method. The results clarified that compared to the experimental procedures and numerical methods, the artificial neural network model has an outstanding performance in considerably less time and cost. Despite a variety of cuts, the present work introduced generalized and unique tools to design and control these systems. The work interestingly reveals that the model can forecast the friction factor and the heat transfer augmentation only by the specification parameter of cuts rather than the shape of cuts. Further sensitivity studies also proved that, if necessary, some variables could be removed from the input vectors, and the models predicting ability remains acceptable.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.7
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available