Elnaz, Yousefi (2024) Nano enhanced phase change material properties driven by artificial intelligence method. Masters thesis, Universiti Malaysia Pahang Al-Sultan Abdullah (Contributors, Thesis advisor: Farzad, Jaliliantabar).
|
Pdf
ir.Nano enhanced phase change material properties driven by artificial intelligence method.pdf - Accepted Version Download (648kB) | Preview |
Abstract
A phase change material (PCM) is a substance thatcan remarkably release or absorb adequate energy to provide heat or cooling applications. An essential step in using phase change materials is selecting the proper PCM according to its characteristics and prospective application; many items, such as the extent of toxicity, thermal conductivity, chemical stability, and expenses, can be included. Recently, considerable development has occurred in applying the nano-particles (NPs) combined with the PCMs to augment their thermal conductivity, especially in latent heat. This thesis is divided into two main parts, each addressing the performance of enhanced phase change materials (PCMs). The initial section focuses on investigating the thermal characteristics—thermal conductivity, thermal stability, and latent heat—of nanostructure-enhanced phase change materials (NePCM) based on eicosane. Eicosane (CH3(CH2)18CH3), with its melting point at 37°C, serves as the fundamental PCM. Copper (II) oxide (CuO) nano-particles (NPs) are chosen as nanoscale enhancers for thermal conductivity. Multiple eicosane-CuO batches with 0.5, 0.7, and 1 mass fractions are analyzed to assess the influence of these fractions on PCM characteristics. Furthermore, determining latent heat values is executed using differential scanning calorimetry (DSC). Comprehensive thermophysical attributes of NePCM suspensions, including thermal conductivity, have been methodically evaluated across various temperatures and NePCM concentrations. Moreover, the characterization of NePCM involves scrutiny through FTIR and SEM techniques. In the latter part of this thesis, machine learning techniques are employed to predict NePCM characteristics and train on data amassed during the experimental process. MLP (multilayer perceptron) is used as the ANN model, and the model predicted two thermophysical properties of the NePCM (latent heat and thermal conductivity). The input of the models was the temperature and phase of the NePCM and concentrations of NPs. The results of the thermophysical analysis of the samples showed improvement in the desirable characteristics of the PCM without changing the chemical properties of the PCM. Remarkably, the introduction of CuO NPs has enhanced the composite's thermal conductivity and latent heat. The maximum latent and thermal conductivity were observed for the highest concentration of NPs (1%wt), which were 44.03 °C and 0.54 W/mK, respectively. Furthermore, FT-IR analysis has confirmed the absence of chemical interactions between the PCM and CuO NPs. Overall, the findings have underscored the potential of incorporating inorganic materials (CuO NPs) to significantly enhance the thermal properties of phase change materials for diverse applications. The developed model consists of three main layers: input, hidden, and output. Generally, the structure of the developed model to predict latent heat and thermal conductivity was 10-8-1 and 10-9-1, respectively. These numbers show the number of neurons in the input, hidden, and output layers. The number of samples for the thermal conductivity model was 205; for the latent heat model, it was 5408. These samples were divided into training, test, and validation data (70, 15, and 15%, respectively). Validation data showed no overlearning in the models, and the coefficient of determination (R-value) was 0.97213 and 0.99985 for thermal conductivity and latent heat models, respectively. The optimum number in the hidden layer for the developed Artificial Neural Network (ANN) model is to predict thermal conductivity and latent heat, drawing from experimental data. The proficiency of the ANN model and its alignment with experimental data have underscored its predictive prowess. The results demonstrate the potential of adding nano-materials to improve PCM capabilities and the ANN model's ability to predict necessary NePCM attributes correctly.
| Item Type: | Thesis (Masters) |
|---|---|
| Additional Information: | Thesis (Master of Science) -- Universiti Malaysia Pahang – 2024. SV: Dr. Farzad Jaliliantabar, NO. CD: 13514 |
| Uncontrolled Keywords: | differential scanning calorimetry (DSC) |
| Subjects: | T Technology > T Technology (General) T Technology > TJ Mechanical engineering and machinery |
| Faculty/Division: | Institute of Postgraduate Studies Faculty of Mechanical and Automotive Engineering Technology |
| Depositing User: | Mr. Nik Ahmad Nasyrun Nik Abd Malik |
| Date Deposited: | 03 Sep 2024 02:17 |
| Last Modified: | 03 Sep 2024 02:17 |
| URI: | http://umpir.ump.edu.my/id/eprint/42468 |
| Download Statistic: | View Download Statistics |
Actions (login required)
 |
View Item |