|
The surrogate assisted design optimization of a twisted oval tube heat exchanger |
|---|---|
| รหัสดีโอไอ | |
| Creator | Kittinan Wansasueb |
| Title | The surrogate assisted design optimization of a twisted oval tube heat exchanger |
| Contributor | Pitak Promthaisong, Bopit Bubphachot, Teerapat Chompookham, Narin Siriwan |
| Publisher | Khon Kaen University, Thailand |
| Publication Year | 2568 |
| Journal Title | Asia-Pacific Journal of Science and Technology |
| Journal Vol. | 30 |
| Journal No. | 3 |
| Page no. | 5 (16 pages) |
| Keyword | Twisted oval tube, Heat transfer, Surrogate model, Kriging, Surrogate-assisted design optimization |
| URL Website | https://apst.kku.ac.th/ |
| Website title | https://apst.kku.ac.th/the-surrogate-assisted-design-optimization-of-a-twisted-oval-tube-heat-exchanger/ |
| ISSN | 2539-6293 |
| Abstract | A design for a twisted oval tube (TOT) heat exchanger using a surrogate-assisted metaheuristic (MH) optimization technique is proposed in this work. The heat transfer characteristic in the TOTs is presented using the computational fluid dynamic (CFD) method, considering parameters such as pitch ratios (PR), cross-sectional ratios (DR), and Reynolds number (Re) ranging from 0.6 to 1.4, 0.02 to 0.1, and 100 to 2,000, respectively. The fitness functions were considered as multi-objective, including the Nusselt number (Nu) and the Poiseuille number (fRe) [1]. To reduce the time consumed in the design procedure, surrogate-assisted optimization was applied in the optimum design search phase. Well-known surrogate models (SuMo), including the Kriging (KRG), radial basis function interpolation (RBF), and k-nearest neighborhood method (KNN), were investigated and compared when applied with metaheuristic algorithms. The results show that the most acceptable prediction model is the RBF-Inverse Multiquadric kernel with the multi-objective meta?heuristic with iterative parameter distribution estimation (MMIPDE), with average errors of 20.9731 and 15.6011 for Nu and fRe, respectively. |
