Artificial intelligence analysis of thermal energy for convectively heated ternary nanofluid flow in radiated channel considering viscous dissipations aspects

Abstract

Thermal management is very important in engineering applications to improve the systems’ performance and to reduce the environmental impact. This research works to establish the convective heat transfer coefficient (CHTC) of new improved ternary hybrid functionalized nanofluids with CuO, Fe2O3, and SiO2 nanoparticles in polymeric fluid. The investigations are aimed at the use of the state-of-the-art AI techniques for predicting and simulating the heat transfer processes in radiated channels as well as incorporating the effects of viscous dissipation and radiation. A new computational process tools up Python, Mathematica, and MATLAB to solve the transformed system of PDEs a LMNNA. These results support the qualitative understanding regarding flow rate dependency on R but dependency of flow rate on γ. Likewise, temperature profiles increase with increase in Eckert number (Ec) and Prandtl ratio (Pr) but decreases as radiating parameter (Rd) increases. The use of AI in creating the simulations is more accurate for prediction than traditional numerical methods with an improved MSE of up to 10−14 through the Python model. With focus on technological advancements in the field of thermal heat, these studies show great promise of THF in enhancing rate of heat transfer-issues which complete several energy storage systems, cooling techniques in aeronautics as well as electric vehicle operational convenience via thermal layout.A synergetic composition of three distinct nanomaterial oxides of Copper, Iron and Silicon in engine oil, contributes unique thermophysical character in thermal management. Advance computational technique with combination of AI with Python, Mathematica and Matlab (AIPMM) employing Levenberg Marquardt Neural Network Algorithm (LMNNA), is used for solving a transformed system of ODEs, which was obtained from the system of PDEs of present model. Dataset generated from Python and Mathematica is filtered and embedded into LMNNA for evaluation and comparison of results.Temperature and flow rate profile are analyzed against variations in sundry characteristics. The profile of flow rate shows it increases with fluidity parameter R and decreases with increasing deviation parameter γ. Temperature outline shows it enhances with Eckert Ec and Prandtl Pr ratio but decreases with increase in radiating parameter Rd.

Keywords

• Artificial intelligence
• Levenberg-Marquardt algorithm
• Nanofluids
• Velocity slip
• Neural network
• Tri-nanoparticle hybrid fluid