A predictive model for boiling heat transfer coefficient of dielectric fluids on metal foams

• Autor(a) principal: Manetti, Leonardo L. [UNESP]
• Data de Publicação: 2020
• Outros Autores: Oliveira Henriques Moita, Ana Sofia, Cardoso, Elaine Maria [UNESP]
• Tipo de documento: Artigo de conferência
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.1615/TFEC2020.boi.032028; http://hdl.handle.net/11449/205201
• topic: Heat transfer coefficient; HFE-7100; Metal foams; Pool boiling; Predictive model

Pool boiling is a suitable technique for direct immersion cooling in electronic devices coupled with dielectric fluids. However, these fluids have relatively poor thermophysical properties in contrast to water, and extremely small contact angle that causes temperature overshooting at the boiling incipience. So, the use of surface enhancement techniques such as porous surfaces has been widely reported to enhance heat transfer performance and meet the cooling requirements. The porous thickness and pore size are the most important parameters of a porous surface, and their optimal values mainly depend on the fluid properties. This work aims to investigate the performance of metal foams of nickel and copper, with different pore diameter and thicknesses on pool boiling, using HFE-7100 as working fluid. A predictive model was proposed for the heat transfer coefficient (HTC) based on the Buckingham π theorem and experimental database. Additional data were taken from the literature for comparative purposes. The dimensionless numbers showed a greater contribution of the transient heat conduction and single-phase convection than the latent heat. In addition, as the pore diameter decreases the HTC increases. The thickness presents a variable exponent, which is a function of the heat flux, due to the balance of heat transfer area and vapor bubble resistance. The developed model accurately predicts 93% of the experimental data within an error band of ± 30% and absolute mean deviation of 13%; moreover, the developed model predicts 68% (within the ± 30% error band) of data from the literature for different working fluids and foams parameters.