CFD analysis of hotspots copper metal foam flat heat pipe for electronic cooling applications

https://doi.org/10.1016/j.ijthermalsci.2020.106583Get rights and content

Highlights

  • Investigate the potential use of copper metal foam.

  • Highlights the importance hotspot locations on FHPs operation condition.

  • Effect of heat sources distance on liquid pressure drop.

  • Comsol Multiphysics as useful tool to predict FHPs hydrodynamic behavior.

Abstract

Hotspots have become one of the most limiting factors in the conception of more and more performing electronic chips. Although many wicking materials are used as porous media in used as porous media in flat plate heat pipe (FHP), copper metal foams are not well studied with these devices. In the present study, a three-dimensional FHPs with multi heat sources is investigated numerically using COMSOL Multiphysics. The potential use of copper metal foam as a wick structure is investigated. The effect of heat flux, as well as hotspot positions on heat pipe performance, is analyzed and results show that copper metal foam can be an alternative wick used for FHPs. Visualization is made taking into account the hot spot position effect at different heat flux. Results show that top hot spot positions especially close to the wall are the most critical locations where dry out can occur. Hot spot dimensions, as well as their distances and effective pore radius, have a significant effect on heat pipe operation.

Introduction

Heat pipes emerged as the most appropriate technology and cost-effective thermal design solution due to its excellent heat transfer capability, high efficiency, and structural simplicity. The thermal performance of a heat pipe can be characterized by both its overall thermal resistance and its maximum power in horizontal and vertical positions. These characteristics depend mainly on the capillary structure, which is usually made of grooves, meshes, sintered powder or a combination of them [1].

In recent years, there is a clear trend in the market towards the development of more compact electronic devices. Thermal management has emerged as the primary challenge in this scenario where the heat flux dissipation of electronic chips is increasing exponentially. One of the major problems arises is due to chip thermal hotspots, which can be as large as eight times the average power density. To keep device junction temperatures within the safe operating limit, there is an urgent requirement for ultrahigh-conductivity thermal substrates that not only absorb and transport large heat fluxes but can also provide localized cooling to thermal hotspots. Hotspots with high heat flux are prevalent in electronic systems with severe impacts on their performance and reliability. Thermal management of hotspots is often complicated by their unsteady spatial distribution, e.g., in high-speed microprocessors with constantly changing computing tasks and high-density power electronics with rapidly evolving output demands. Mobile hotspots are inadequately addressed with existing hotspot [2].

For modern electronic cooling, heat pipe shall dissipate ~100 W heating loads on a square centimeter area or heat flux attaining ~500 W/cm2 for hotspot (for instance, one square millimeter area). The chip temperature should be lower than 85 °C and critical heat flux shall not occur [3,4].

A flat plate heat pipe is expected to satisfy the challenging requirements. Many researchers investigated the characteristics of heat pipe both experimentally and analytically. Most of them are performed assuming a grooved wick structure and studied are focused mainly in two main categories: 1) the numeric or analytic simulation of steady-state or transient operation of FHPs [[5], [6], [7], [8], [9]], and 2) the transport limits (mainly capillary and boiling) and dry-out lengths [[10], [11], [12]]. Complex capillary structures such as meshes or sintered powder are also used in FPHP.

Avenas et al. [13] evaluated the thermal performance of a flat heat pipe and found that 40% of the thermal resistance of the heat pipe is reduced by using pure copper as a wick structure.

Xuan et al. [14] studied the performance and mechanism of a flat plate heat pipe under different heat fluxes, orientations and amounts of the working fluid by means of both experimental and theoretical approaches.

Yahushi et al. [15] suggested a mathematical model simulate a vapor chamber to cool the new generation of electronics. The center heated vapor chamber was studied mathematically and experimentally and the compared results showed close agreement.

Xiao and Faghri [16] developed a detailed 3-D model to analyze the thermal hydrodynamic behaviors of flat heat pipes without empirical correlations.

Do et al. [17] investigated the thermal performance of screen mesh wick heat pipes using water-based Al2O3 nanofluids. A significant decrease in thermal resistance was observed. This enhancement of the thermal performance is due to the thin porous coating layer formed by nanoparticles at the wick structure.

Wang [18] reports on the theoretical, simulation and experimental studies of a light-emitting diode (LED) vapor chamber using the illumination-analysis method. Results show that the thermal performance of the LED vapor chamber-based plate is better than that of the LED copper-based plate with an input power above 5 W. However, the wick structure of the vapor chamber is not finely defined.

Wong et al. [19] studied the evaporation process in the wick of an operating flat plate heat pipe with a sintered two-layer 100 + 200 mesh copper wick (0.26 mm thick) and three different working fluids such as deionized water, methanol, and acetone. Their results indicate that the maximum heat loads for water are far greater than those of methanol and acetone.

Recently, Dillig et al. [20] carried out experiments with planar heat pipes using sodium as working fluid and different types of wick structures (screen mesh, sintered plates, and grooves). The capillary limitation was identified as one of the main challenges in their study, underlining the importance of the wick structure optimization.

The heat pipe used for electronic cooling purposes is usually a wicked capillary force-driven heat pipe, with a working temperature range of about 50–120 °C. For different types of FPHP wicks, researchers have studied the numerical as well as experimental performance. The main results indicated that high permeability and small pore diameter of the wick material increase the heat transfer due to low fluid flow resistance and high capillary pumping advantages respectively.

Different materials are proposed as a wick for a heat pipe and among them, recently invented Bi-porous metal foams exhibit a very significant performance improvement, i.e. high transport limit in comparison with competing materials. Metal foams are becoming attracting wick materials for heat. Metal foams have high permeability and many numbers of small diameters pores that make the fluid transport and capillary actions increase to maximize the heat transferred by the heat pipe.

Metal foams are porous media with low density and novel thermal, mechanical, electrical, lower weight and acoustic properties [21]. They can be categorized as open-cell or closed-cell foams, but only open-cell metal foams appear to have promise for constructing heat exchangers. Usually, the goal is to achieve high capillary pressure and high permeability, but the weight of the wick structure can also be a parameter for optimization [22].

Unmodified foams usually have larger pore sizes which are too large for heat pipe wick applications. Additional data are also needed, especially related to the wettability and the flow characteristics of the different wick and working fluid combinations.

The key parameters for the flow characteristics are the permeability and the effective pore radius of the wick. The effective pore radius depends on the wettability, commonly described by the contact angle [23]. To reduce pore size compression is used. Compression of nickel foam, as described by Sheehan et al. [24] and Queheillalt et al. [25], is one method of producing wicks that have so far not been given much attention. The method is particularly interesting because the degree of compression, and thereby the pore size and the wick properties can be tailored to best fit the requirement of the application at hand.

Numerous studies on the metal foams for thermal management of electronics by single-phase forced convection using different coolants [[26], [27], [28]].

Phillips [29] studied the permeability, capillary pressures and evaporative performance of different porous materials including high porosity metal foams in order to evaluate their performance as wicks in heat pipes. He concludes that from the capillary pumping point of view, foam wicks are most desirable, felts rank second, and screens are least desirable. But for the onset of nucleate boiling, felts are more convenient than foams. Peterson et al. [30], Carbajal et al. [31] and Queheillalt et al. [32] have studied transient response of a large flat heat pipe structure employing nickel foam as a wick, while exposed to a non-uniform localized heat flux. Studies to date on metal foams do not indicate significant efforts, especially for electronic cooling applications. Thus, most of them focused on the wick properties or specific applications, but the relationship between these properties and operating limits remains to be investigated. The thermal performance of a flat heat pipe with metal foam wick is experimentally investigated by Somasundaram et al. [33]. They evaluated the effect of heat input, cooling water flow rate and temperature at the condenser side, and the fill ratio of working fluid on the thermal efficiency of copper flat heat pipe with a wick with 0.77 porosity and 1.114e-10 mm2 and water as the working fluid. They concluded that the addition of one and two wick columns to the setup lead to 2.1% and 3.1% reduction in thermal resistance respectively.

The aspect of Metafoam copper foams is their particular microstructure compared with other metal foams and sintered copper powder. It is characterized by having two-pore scales and a high surface area. This bi-porous microstructure may lead to interesting capillary and evaporation properties that have not yet been fully characterized [34,35]. Recently, Xianbing et al. [36] proposed a novel strategy for the improvement of vapor chamber heat pipe (HP) performance. The idea is to maintain a nucleation mechanism but suppresses convection mechanism. Holding the nucleation mechanism, the developed heat pipe behaves excellent thermal sensitivity with respect to heating loads.

Numerical and Computational Fluid Dynamic (CFD) analysis of heat pipes also has been evaluated by researchers to gain a better insight into the heat pipe thermal behavior. Davarnejad and Jamshidzadeh [37] developed a CFD model to simulate the thermal behavior of MgO nanofluid with various particle concentrations in a tube. k-e Method was used to model the turbulent flow. The friction factor and Nusselt number were studied at different Reynolds numbers and particle concentrations in nanofluid. CFD model showed by increasing the concentration of the nanoparticle, heat transfer and pressure drop directly related to the friction factor increased. Mohammed et al. [38] studied the CFD analysis of the effect of nanoparticle in thermal and heat transfer performance of acetone/ZnBr2 solution. A two-phase CFD methodology with standard k-e method and SIMPLE scheme for pressure/velocity coupling was used by ANSYS Fluent to simulate the Eulerian-Eulerian two-phase flow. The validity of the model was examined by other experimental studies and which showed a good agreement between the CFD model and experimental results. Asmaie et al. [39] developed a CFD model to predict the thermal performance of a nanofluid in a thermosyphon heat pipe. A volume of fluid (VOF) model is used to solve the momentum equation. VOF considers two or more fluid mixtures and solves a single momentum equation for all of them and follows the footprint of each fluid by their corresponding volume fraction. Temperature distribution at the wall, heat transfer coefficients at condenser and evaporator, and effect of nanoparticle concentration on heat pipe thermal performance was presented by the VOF computational dynamic model. The results showed that the thermosyphon maximum heat transfer via nanofluid was 46% more efficient than that of pure water. According to the above-mentioned literature review, the field of computational fluid dynamics coupled with heat transfer is still rather undiscovered. There are a lot of potential subject matters that could be addressed during the CFD modeling of heat pipes. Therefore, in this work, in addition to discussion about the analytical solutions explaining the thermal behavior of the heat pipes, a CFD model evaluating a rectangular heat pipe is modeled [40]. The analytical model enables to solve the thermal and hydrodynamic equations involved in a heat pipe with a low computation time. As a consequence, it could easily be used to optimize the location of the heat sources on the electronic card by considering the different constraints of the users. This type of problem can indeed be time consuming using a numerical approach. As the boundary condition has to be homogeneous to solve the heat transfer equation analytically.

The objective of this research is therefore to simulate the thermal behavior of copper metal foam flat plate heat pipe under a localized heat source. Unlike previous reports with uniform heat fluxes, one or two-dimensional heat transfer can no longer be assumed in the present study focused on hotspot cooling. Consequently, many geometrical details are now important and are accounted for by a numerical model (Comsol 5.3). This model is particularly convenient for engineers since it enables the calculation of both the 3D temperature field and the capillary limit of an FPHP used to cool several electronic components (or any kind of heat source) and several heat sinks, whatever their locations and their heat transfer rate. It can be also an alternative tool to simulate heat pipes under several conditions.

Section snippets

Mathematical model

The numerical model employed here is adapted from previous work by the authors [35]. A Brinkman-Forchheimer extended Darcy model is employed for fluid in the wick. The vapor flow is assumed to be laminar and incompressible. The phase-change mass flow rate due to evaporation/condensation and the temperature and pressure at the liquid-vapor interface are determined using an energy balance at the interface in conjunction with kinetic theory and the Clausius-Clapeyron equation. The energy balance

Result discussion

The numerical model is tested for copper metal foam PPI 250 (pore per inch) with water as working fluid at an operating temperature of 60 °C which is the typical operating temperature in practice [42]. A comparison is also made for other wicks. The configuration consists of a flat rectangular cross-section with wicking layers on both the top and bottom. The thermo-physical properties of these porous structures are presented in Table .1.

At present, the heat fluxes in high-end server applications

Conclusion

Hotspot cooling is critical to the performance and reliability of electronic devices, but existing techniques are not very effective in managing mobile hotspots. 3D numerical model using (COMSOL 5.3) is used giving a solution for both the liquid and vapor flows inside an FPHP and temperature distribution inside its sections. The numerical model could easily be used to optimize the location of the heat sources on the electronic card by considering the different constraints of the users. This

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (46)

  • K.H. Do et al.

    Thermal resistance of screen mesh wick heat pipes using the water-based Al2O3 nanofluids

    Int. J. Heat Mass Tran.

    (2010)
  • J.C. Wang

    Thermal investigations on LED vapor chamber-based plates

    Int. Commun. Heat Mass Tran.

    (2011)
  • S.C. Wong et al.

    Visualization and evaporator resistance measurement in heat pipes charged with water, methanol or acetone

    Int. J. Therm. Sci.

    (2012)
  • Raffaele Savino et al.

    Comparative study of heat pipes with different working fluids under normal gravity and microgravity conditions

    Acta Astronaut.

    (2008)
  • Raffaele Savino et al.

    Comparative study of heat pipes with different working fluids under normal gravity and microgravity conditions

    Acta Astronaut.

    (2008)
  • K. Boomsma et al.

    Metal foams as compact high performance heat exchangers

    Mech. Mater.

    (2003)
  • G. Carbajal et al.

    A quasi-3D analysis of the thermal performance of a flat heat pipe

    Int. J. Heat Mass Tran.

    (2007)
  • D. Somasundaram et al.

    Experimental investigation of thermal performance of metal foam wicked flat heat pipe

    Exp. Therm. Fluid Sci.

    (2017/04/01/2017)
  • M. Shafahi et al.

    An investigation of the thermal performance of cylindrical heat pipes using nanofluids

    Int. J. Heat Mass Tran.

    (2010/01/15/2010)
  • Xianbing Ji et al.

    Switchable heat transfer mechanisms of nucleation and convection by wettability match of evaporator and condenser for heat pipes: nanostructured surface effect

    Nanomater. Energy

    (2017)
  • R. Davarnejad et al.

    CFD modeling of heat transfer performance of MgO-water nanofluid under turbulent flow

    Engineering Science and Technology, an International Journal

    (2015/12/01/2015)
  • H.I. Mohammed et al.

    CFD assessment of the effect of nanoparticles on the heat transfer properties of acetone/ZnBr2 solution

    Appl. Therm. Eng.

    (2018/01/05/2018)
  • Grissa Kods et al. Investigations of the thermal performance of a cylindrical wicked heat pipe January...
  • Cited by (29)

    View all citing articles on Scopus
    View full text