Statistical approach for prediction of thermal properties of CNC and CNC-CuO nanolubricant using Response Surface Methodology (RSM)

• Main Authors: Sakinah, Hisham, K., Kadirgama, D., Ramasamy, M., Samykano, W. S., Wan Harun, R., Saidur
• Format: Conference or Workshop Item
• Language: English
• Published: IOP Publishing 2019
• Subjects: TJ Mechanical engineering and machinery
• Online Access: http://umpir.ump.edu.my/id/eprint/29819/1/1.%20Statistical%20approach%20for%20prediction%20of%20thermal%20properties%20of%20CNC.pdf; http://umpir.ump.edu.my/id/eprint/29819/; https://doi.org/10.1088/1757-899X/788/1/012016

In the present work, response surface methodology (RSM) using the miscellaneous design model was performed to optimize thermal properties of Cellulose nonocrystal (CNC) and hybrid of cellulose nanocrystal-copper (II) oxide (CNC-CuO) nanolubricant. Influence of temperature, concentration and type of nanolubricant is used to develop empirical mathematical model by using Response Surface Methodology (RSM) based on Central Composite Design (CCD) with aid of Minitab 18 statistical analysis software. The significance of the developed empirical mathematical model is validated by using Analysis of variance (ANOVA). In order to produce second-order polynomial equations for target outputs including thermal conductivity and viscosity, 26 experiments were performed. According to the results, the predicted values were in sensible agreement with the experimental data. In other words, more than 80% of thermal conductivity and specific heat capacity variations of the nanolubricant could be predicted by the models, which shows the applied model is precise. The predicted optimized value shown in the optimization plot is 0.1463 for thermal conductivity and 1.6311 for specific heat capacity. The relevant parameters such as concentration, temperature and type of nanolubricant are 81.51, 0.1, and the categorical factor is CNC-CuO. The composite shown in the plot is 0.6531. The validation result wit experimental as shown in indicate that the model can predict the optimal experimental conditions well.