Enhanced heat transport behavior of micro channel heat sink with graphene based nanofluids

Highlights

• Functionalized graphene/water nanofluid was prepared and characterized.

• Thermal conductivity of nanofluid increased by ~11% at 0.2 vol% of GnP.

• Heat transfer coefficient and pressure drop of GnP/water nanofluid are reported.

• Heat transfer coefficient enhanced by ~71% with 0.2 vol% of GnP.

• The increase in pressure drop was significantly limited with 0.2 vol% of GnP.

Abstract

This paper investigates the convective heat transfer characteristics of f-GnP (functionalized Graphene Nanoplatelets) suspended in distilled water. The effect of mass flow rate (5 g/s to 30 g/s) and the GnP concentration (0 vol% to 0.2 vol%) on various heat transport parameters such as convective heat transfer coefficient, temperature drop, Nusselt Number, Pressure drop have been discussed in this study. Experiments with GnP based nanofluids in the heat sink decreases the heat sink temperature by 10 °C and enhances the convective heat transfer coefficient and Nusselt number by 71% and 60% with a little increment in the pressure drop of about 12% higher as compared with water. The results prove that the functionalized GnP based nanofluids would effectively replace the existing nanofluids that improves the performance of the electronic chips due to enhanced thermal conductivity and convective heat transfer coefficient.

Introduction

Electronics cooling remains a major concern in all industrial sectors and business organizations dealing with data processing, data communication and data storage due to increase in power dissipation from the electronic chips. The migration of processor chips from one generation to next advanced generation with increasing processing speed leads to dissipation of high power density. Due to ever-increasing computing demand, there is a need for powerful processers. These powerful processors tend to generate heat at a higher rate, and once their core temperature exceeds the safe thermal design temperature, the processors cease to function or at cases may be severely damaged. Thus, there is a constant search for cooling systems that are effective and are also compact in nature.

There are various methods of cooling based on coolants used namely air cooling, liquid cooling and refrigerant cooling respectively. Air cooling is more versatile where the space limitation is more prominent around the processor chips with the absence of sealed enclosures and piping with minimal maintenance [1]. Despite of its minimal maintenance, air has a low thermal conductivity as compared with the liquids. Several researchers analyzed the effects of air as a coolant for electronics system [[2], [3], [4], [5], [6]] and concluded that air is not suitable for high-performance devices dissipating a high power flux. Liquid cooling offers more benefit as compared to air due to its superior thermal characteristics. Past few decades, researchers conducted studies on the development of new coolants with enhanced thermal transport characteristics which resulted in the evolution of newer class of fluids termed as nanofluids. This class of fluids was first reported by Choi et al. [7] by dispersing the metals or its oxide particles of less than 100 nm in size to the traditional coolants viz., distilled water, ethylene glycol. Experiments were conducted on various metal and metal oxide nanoparticles and it was observed that there was a significant improvement in heat transfer coefficient improving the performance of cooling system [[8], [9], [10], [11], [12]].

Several researches were carried out on various metals, its oxides and carbon based nanofluids for enhancing the performance of electronic devices. Sohel et al. [13] experimentally observed the thermal effects of aluminium oxide water nanofluids passed through the custom made copper minichannel heat sink. The experimental results showed that the heat sink temperature reduces with 0.2 vol% of the nanofluids as compared with distilled water and reported that thermal entropy generation rate was found to be decreased by 11.5% as compared with the pure water. Sarafraz et al. [14] experimentally compared the convective performance of copper oxide and gallium dispersed distilled water nanofluids by passing through the Intel core i5 4760 processor liquid cooling system. The experiments were carried out for three different operating conditions of the processor viz., standby, normal and overload conditions. It was reported that the normal temperature of CPU for fan cooling will be around 69–70 °C while for the water, copper oxide and gallium nanofluids, it was found to be 67 °C, 62 °C, and 53.1 °C respectively. It was also concluded that the gallium dispersed nanofluids showed a better heat transfer performance compared to copper oxide nanofluids but has higher pumping power of about 51% as compared to copper oxide nanofluids which is not beneficial.

Cong et al. [15] evaluated the thermo hydraulic performance of CPU coolers with two different nanofluids namely aluminium oxide and titanium oxide dispersed in pure water. The concentrations of aluminium oxide were varied from 0.1 wt% to 2 wt% and titanium oxide from 0.1 wt% to 1 wt%. It was reported that reduction in CPU temperature was about 23.2% and 14.9% respectively for both aluminium oxide and titanium oxide nanofluids respectively. Sarafraz et al. [16] investigated the thermal performance of aqueous carbon nanotubes ranging from 0.05 wt% to 0.1 wt% in a copper made heat sink with rectangular parallel micro channels. It was found that the average heat transfer coefficient increases by 57% at lower heat flux however there was only slight improvement at higher heat flux as CNT particles remains stable in the base fluid at lower heat flux and starts destabilizing with increase in the heat flux applied. Sun et al. [17] have compared the thermal performance of both copper and aluminium oxide nanoparticles dispersed in distilled water by passing the nanofluids through liquid cooled central processing unit with heating radiator. The temperature was found to be decrease by 4–18 °C as compared to water. While comparing the performance of two nanofluids, heat transfer coefficient was found to be 1.1–2 times higher that of distilled water in case of copper based nanofluids. Cong et al. [18] made a comparative study on different arrangement of flow channels viz., aligned and staggered arrangement, respectively in a heat sink operated with titanium oxide nanofluids. It was reported that there is decrement in the CPU temperature of about 10.5% and 12.5% for aligned and staggered arrangement respectively. It was also reported that the titanium oxide nanofluids offers a better thermal performance at 0.4 wt%. Singh et al. [19] studied the effects of aqueous based suspensions of carbon nanotubes on heat transfer characteristics in a triangular shaped microchannel heat sink. It was reported that the heat transfer coefficient enhanced upto three times higher than the pure basefluid. Particle rearrangement, distortion of thermal boundary layer due to the inclusion of CNT, thermal conduction enhancement of CNT and high aspect ratio of CNT were reported to be the possible mechanisms for the heat transfer enhancement.

After rigorous literature survey, we found that there is only one article which discusses the experimental investigation of heat transfer characteristics of graphene-DI water for electronics cooling in microchannel heat sinks. Goodzari et al. [20] performed an numerical analysis on heat transfer behavior of Graphene-Silver/water hybrid nanofluids with volume fraction varying from 0 to 0.1 vol% in a micro channel with different cross sections. For the above mentioned hybrid combination, the concentration of silver remains fixed while the concentration of GnP has been varied. Authors reported that there is a significant increase in heat transfer with increase in the concentration of GnP. Vishnu Prasad et al. [21] conducted experiments on graphene dispersed DI water nanofluids on the performance of the microchannel heat sink and observed an enhancement of heat transfer coefficient by 78.5% at a heat load of 40 W at lower mass flow rate of nanofluids. The research was limited upto 40 W heat dissipation of the electronics chips while the flow rate was varied upto 10 ml/s which remain insignificant for predicting the behavior of GnP based nanofluids in cooling the electronic chips operating at higher heat flux density. Graphene possess a higher thermal conductivity of the order of 3500 W/mK and possess a lower density in the range of 2.2 g/cc. The functionalization process increases the stability of the nanofluids without increasing its viscosity due to the elimination of the surfactant. Only limited studies were reported on the convective heat transfer performance of acid treated GnP nanofluids especially in the field of electronics cooling. This drives the study to examine the behavior of acid treated GnP based nanofluids in cooling of electronic chips operating at higher heat flux densities.