Thermal performance and optimum concentration of different nanofluids in immersion cooling in data center servers
Fahad S. Alkasmoul, Ahmed M. Almogbel, Muhammad Wakil Shahzad, Amer Al‐damook
Abstract
• Thermal Performance of Nanofluids in Iceotope Server: Evaluation of heat transfer efficiency using different nanofluids in the Iceotope server cooling system. • Optimal Nanofluid Concentration for Cooling: Identification of the most efficient nanoparticle concentrations for effective cooling in the Iceotope server. • Immersed Cooling of Iceotope Server: Analysis of the immersion cooling system for the Iceotope server using advanced nanofluids. • Improved Thermal Efficiency: Al₂O₃-water nanofluid significantly enhances CPU cooling due to its superior thermal conductivity, reducing operating temperatures. • Viscosity and Pressure Challenges: Higher nanoparticle concentrations result in increased viscosity, leading to greater pressure drops and higher pumping power, particularly in Al₂O₃-water nanofluid. • Cost-Effective Cooling: Despite thermal improvements from nanofluids, pure water remains the most cost-effective coolant, especially at optimal performance with Re = 600. • Balancing Efficiency and Energy Use: The study underscores the importance of balancing thermal performance with energy consumption for an economically feasible cooling solution. • Future Research Directions: Exploration of hybrid nanofluids or alternative cooling methods could further optimize both thermal management and cost efficiency for data center cooling systems. The application of nanofluids in heat transfer systems is widely recognized for enhancing thermal efficiency compared to conventional fluids, representing a pivotal advancement in energy-saving strategies across a range of commercial applications. However, the selection of nanofluid concentration often lacks a systematic approach, with decisions made independently of specific application requirements, type of nanofluid, cost considerations, and economic factors such as energy cost and system lifetime. This article explores the optimization of nanoparticle concentrations in the immersion cooling system of a data center server, utilizing three distinct nanofluids (Al₂O₃-water, TiO₂-water, and CuO-water). The Iceotope immersion cooling designs were examined across various flow conditions. The investigation involved calculating heat transfer rates and pressure drops through the server by simulating a model using Computational Fluid Dynamics (CFD), Finite Element (FE), and COMSOL Multiphysics® Modeling software for varying volume fractions of nanoparticles. The study proposes a novel methodology to assess potential economic trade-offs associated with nanofluid utilization in immersed liquid-cooled data centers. This approach aims to identify the lowest overall cost within the data center and the maximum viable volume fraction of a nanofluid. Results indicate that Al2O3-water nanofluid exhibits superior thermal efficiency due to its higher thermal conductivity, while TiO₂-water and CuO-water are evaluated for comparative performance. However, it also presents challenges, such as higher viscosity leading to increased pressure drop and pumping costs. Economic considerations reveal that, despite its thermal advantages, the most cost-effective fluid is the conventional fluid (pure water). The server incurs a cost of $706.15. These findings underscore the importance of considering both thermal performance and economic factors when determining the optimal nanofluid concentration for practical application.