Analysis of the EMHD nanofluid flow for geothermal pipelines using physics-driven deep learning.

Area of Science:

• Fluid Dynamics
• Heat Transfer
• Machine Learning

Background:

• Hybrid nanofluids exhibit exceptional flow and thermal properties, making them suitable for geothermal energy extraction.
• Understanding electro-magneto-hydrodynamics (EMHD) hybrid nanofluid flow is crucial for optimizing geothermal pipeline applications.

Purpose of the Study:

• To introduce a novel unsupervised deep neural network (DNN) approach for predicting the temperature and velocity behavior of EMHD hybrid nanofluid flow.
• To analyze the influence of electric and magnetic fields on hybrid nanofluid dynamics in geothermal pipelines.

Main Methods:

• A third-grade sodium alginate model was used to examine hybrid nanofluid flow dynamics.
• An unsupervised deep neural network (DNN) was employed to predict the behavior governed by nonlinear differential equations.
• The energy equation incorporated effects of Joule heating and viscous dissipation for fully developed incompressible flow.

Main Results:

• The DNN achieved high accuracy ([Formula: see text] to [Formula: see text]) in predicting fluid behavior.
• Velocity profiles showed symmetry and significant dependence on electric fields and thermal Grashof number.
• Nanoparticles led to a decrease in the overall thermal profile along the pipe length.

Conclusions:

• The study provides a foundational framework for optimizing thermodynamic systems in geothermal applications.
• Findings have practical implications for designing energy-efficient geothermal pipelines with improved heat transfer.
• The DNN approach offers a powerful tool for analyzing complex EMHD hybrid nanofluid dynamics.