Content uploaded by Yixiang Gan

Author content

All content in this area was uploaded by Yixiang Gan on Jul 28, 2016

Content may be subject to copyright.

Effective thermal conductivity of submicron powders: A numerical

study

Weijing Dai1,a, Yixiang Gan1,b*, Dorian Hanaor1,c

1The School of Civil Engineering, The University of Sydney, Sydney, Australia

aweijing.dai@sydney.edu.au, byixiang.gan@sydney.edu.au, cdorian.hanaor@sydney.edu.au

Keywords: effective thermal conductivity, gas heat transfer, finite element analysis, compacted

powders

Abstract: Effective thermal conductivity is an important property of granular materials in

engineering applications and industrial processes, including the blending and mixing of powders,

sintering of ceramics and refractory metals, and electrochemical interactions in fuel cells and Li-ion

batteries. The thermo-mechanical properties of granular materials with macroscopic particle sizes

(above 1 mm) have been investigated experimentally and theoretically, but knowledge remains

limited for materials consisting of micro/nano-sized grains. In this work we study the effective

thermal conductivity of micro/nano powders under varying conditions of mechanical stress and gas

pressure via the discrete thermal resistance method. In this proposed method, a unit cell of contact

structure is regarded as one thermal resistor. Thermal transport between two contacting particles

and through the gas phase (including conduction in the gas phase and heat transfer of solid-gas

interfaces) are the main mechanisms. Due to the small size of particles, the gas phase is limited to a

small volume and a simplified gas heat transfer model is applied considering the Knudsen number.

During loading, changes in the gas volume and the contact area between particles are simulated by

the finite element method. The thermal resistance of one contact unit is calculated through the

combination of the heat transfer mechanisms. A simplified relationship between effective thermal

conductivity and loading pressure can be obtained by integrating the contact units of the compacted

powders.

Introduction

In energy systems, granular media are commonly used to store, convert, capture and produce

energy, including lithium-ion batteries, solid oxide fuel cells, and thermal storage systems. Heat

transport in these energy systems is a key issue necessitating extensive research to 1) establish a

fundamental understanding of the thermo-mechanical properties and 2) provide useful knowledge

for process optimisation [1-3]. Due to heterogeneous material properties, complex packing

structures and various inter-particle interactions, different theoretical methods have been developed

to study the heat transport mechanisms and the influencing external factors [4, 5].

Most research in this field has focused on the effective thermal conductivity (ETC) of powder

beds, employing three types of models to predict this parameter. Type I models are based on the

materials’ constitution and porosity and can be used to calculated the ETC by considering powder

beds as dispersion media [4]. Type-II models regard powder beds as thermal circuits in order to

obtain the effective thermal resistance through an analogy with electrical circuits [4, 6-9]. Type-III

models discretise powder beds into many unit cells and integrate the cells’ properties to derive the

ETC [5, 10, 11]. The former two types of models can provide relatively simple equations to predict

the ETC while the type-III models usually involve considerable numerical calculation. However, it

is more convenient to study the effects of external factors on the ETC and to approximate the real

situation in practical application by applying type-III models. To realize a useful type-III model, a

proper description of the unit cell combing different heat transfer mechanism and external factors is

required. An analytical solution for a contact unit involving two spheres was proposed by Batchelor

and O’Brien in order to describe the relationship between ETC and contact area [10]. Baharami et

al. employ the thermal contact resistance model to include the effects of rough contacts in the

Applied Mechanics and Materials Online: 2016-07-25

ISSN: 1662-7482, Vol. 846, pp 500-505

doi:10.4028/www.scientific.net/AMM.846.500

© 2016 Trans Tech Publications, Switzerland

All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of Trans

Tech Publications, www.ttp.net. (#68764661, Karlsruhe Research Center, Karlsruhe, Germany-22/07/16,06:21:07)

spherical powder beds [7]. The particle size of the powder beds considered in these studies has

typically been above 1 mm, thus the heat transfer through the interparticle gap is regarded as the

same as through a continuum gas phase. Furthermore, Gusarov et al. considered the effect of the

Knudsen number on the ETC of powder beds and the results shows that for millimetre- and

micrometre-sized particles the increase of gas pressure increases the ETC with saturation occurring

above a certain gas pressure [5]. However, these studies have not addressed the powder beds

containing submicron or nano particles which are widely used in sintering process, thermal

insulation, and solid oxide fuel cells. In this study, the finite element method (FEM) is utilised to

calculate heat transport processes in the submicron units with different particle sizes. By altering the

gas phase pressure and mechanical deformation, the relationship between the ETC and loading

condition can be obtained.

Simulation Model

A two-hemisphere contact unit cell is used to evaluate the effective thermal conductivity of

individual element as shown in Figure 1. The hemispheres, with constant material properties, form a

contact along the axial direction concentrically. The gap between the hemispheres is filled with air,

thus completing a cylindrical unit. In this simulation unit, a constant temperature difference

between the top surface and bottom surface is applied, and uniform temperatures of the upper and

lower hemispheres are established at the initial state as and , respectively. An axial

compression is applied from the top surface and the bottom surface is fixed. In this study, the

commercial finite element package ABAQUS is utilised to perform the FEM calculation. The

temperature difference ∆ is maintained to be 1K and −=∆ is created at initial state. The

properties of the air between the gap of the two hemispheres are assumed to be independent of the

temperature distribution of the air phase because ∆T is relatively small. Thus, in the numerical

model, a distance-dependent the gap thermal conductance (*GAP CONDUCTANCE in ABAQUS)

is used to approximate the thermal conductance via the gas gap.

(a) (b) (c)

Figure 1: Unit cell model: (a) schematic drawing of two spherical particles nearly in contact;

(b) Two particles in contact; (c) The FEM model with temperature profile with an arbitrary unit.

Here, ∆T, ∆x, hg and hs denote the temperature difference, particle deformation, heat flux through

the gas phase and through the solid contact, respectively.

Gas heat transfer at submicron-scale depends on the the ratio of the mean free path, , of gas

molecules to the gap distance, as given by the Knudsen number, . For the powder beds

comprised of submicron particles, the gaps between particles become comparable to the mean free

path of the gas molecules, leading to a transition of the heat transfer modes of gas phase. In other

words, the equation ℎ=/ to calculate the heat conductance of gas is invalid in this range,

where is the thermal conductivity of gas and is the gap distance. An approximation based on

∆T hg ∆T

∆x

hg hs

Applied Mechanics and Materials Vol. 846 501

the Knudsen number is used to predict the gas phase heat conductance. The mean free path is

calculated based on the dynamic viscosity

=

(1)

and the dynamic viscosity is calculated by

=

(

)

(2)

where µ0 is dynamic viscosity at =296.15K. At a certain temperature , the mean free path

depends on gas pressure and gas constitution, where is the molar mass of the gas. When the

Knudsen number is very large ( ≫100), the heat conductance is approaching the free

molecular limit [5]

ℎ=

(

)

(3)

Here is the gas constant and is the adiabatic exponent. For 0.01<< 100, the heat

conductance at different Knudsen number is approximated by

ℎ=

(∗)

⁄+

∗)

⁄, (4)

= √

. (5)

It can be concluded that the gas heat conductance at ≫100 depends on not only the

temperature but also the gas pressure while the heat conductance is not affected by the gas pressure

for small values of Knudsen number.

Figure 2: Air heat conductance as a function of gap distance at different gas pressure, at = 298.

The solid line indicate the conductance calculated based on the continuum equation, ℎ=/.

502 Advances of Computational Mechanics in Australia

Further to the heat transfer of gas phase, the heat transfer through the contact region of two

hemispheres in one unit cell needs to be defined. Thermal contact resistance is commonly used to

describe the heat transfer across the solid contact between two relatively large hemispheres. In such

situation, the contact region consists of a finite number of nanoscale contact spots. However, this

contact model is not applicable for situation where the contact region itself falls into nanoscale. Due

to the consistent material properties of the two contact hemispheres, the nano contact region is

considered to have the same thermal properties as the bulk material. Thus the heat conductance of

the contact region is defined as the heat conductance of 1nm thick material in this simulation.

In ABAQUS, the gap conductance is varied with Knudsen number to simulate the gas heat

transport at different gas pressure. Deformation of the unit cell can be achieved by applying

displacement as the boundary condition. Table 1 summarizes the materials and model parameters

used in the ABAQUS simulations. The hemispheres are assumed to be SiO2 powders and the gas

phase is air. The negative sign of axial deformation ratio means that the two hemispheres are

separated and a finite gap is left between two hemispheres.

Table 1: Materials and model parameters used in ABAQUS simulation

Powder material: SiO2

Thermal conductivity ks=1.3 nW/nm·K, specific heat c=7*1011 nJ/kg·K, density

ρ=2.65*10-24 kg/nm3, Young’s modulus=7*10-8 N/nm2, Poisson’s ratio=0.17.

Gas: Air

Molar mass M=0.028966 kg/mol, dynamic viscosity µ

0

=1.827*10-5 Pa·s,

Axial deformation ratio -1% 0% 1% 2% 5%

Particle diameter (nm) 100 200 500 1000

Gas pressure (Pa) 1 100 10000 100000 1000000 1000000

Result and discussion

In these models, the reaction heat fluxes are calculated according to the temperature boundary

conditions. At the equilibrium state, the effective thermal conductivity of individual unit cells can

be calculated based on the Fourier’s law

=∆

(6)

where is the reaction heat flux, denotes the effective thermal conductivity of unit cell, is the

axial -length of the unit cell and ∆ is the temperature difference.

Figure 3 presents the change of ETC of the unit cells versus the change of the gas pressure for

different particle sizes. When the two hemispheres are not in contact (-1% deformation), the heat

transfer inside the unit cell takes place entirely through the gas phase. It is clear that increasing the

gas pressure continuously increases the ETC of the unit cells which is consistent with the trend in

the heat conductance of air versus pressure in Figure 2. For conditions where the heat transfer occur

solely through the gas phase, larger particle sizes result in a larger ETC of the unit cells. Note that

the gap size is relative to the size of the unit cell, and in Figure 3(a) a displacement of -1% is

considered. When the two hemispheres approach to each other to form a point contact, the effect of

gas pressure diminishes at low gas pressure region. The gas pressure begins to affect ETC of unit

cells at value above 10 kPa and this effect becomes more significant as the gas pressure increases.

However, for conditions of particle contact the relationship between the particles size and ETC is

opposite to the situation of separated hemispheres shown in Figure 3(a). Compared with the ETC of

non-contact group, the ETC of this point contact group is much larger due to the presence of solid-

solid contact conductance. The contact point provides a thermal shortcut for the heat transport

because the heat conductance of the contact region is several orders (103~109) larger than the gas

heat conductance.

Applied Mechanics and Materials Vol. 846 503

(a) (b)

(c) (d)

Figure 3: The effective thermal conductivity as a function of air pressure in different sized powders,

at T=298K: (a) non contact, -1% deformation, (b) point contact, 0% deformation,

(c) 2% deformation, and (d) 5% deformation

When the deformation increases as shown in Figure 3 (b-d), with an expanding contact region,

the ETC of the unit cells rises and follows a similar pattern as a function of the gas pressure. The

magnitude of the increase in the ETC due to the change in pressure in the three contact groups (0%,

2%, 5%) actually corresponds to the increment of ETC in the non-contact group (-1%). Only when

the gas heat transfer is comparable to the heat conduction through the contact region, the gas

pressure starts to significantly alter the ETC. This phenomenon indicates these two heat transport

mechanisms, i.e., heat conduction by solid-solid contact and heat transfer through gas phase, are

parallel, which is a general consideration in many other methods used to predict the ETC of powder

beds. Figure 4 shows the effect of deformation on the ETC of unit cells. As the deformation

increases, the ETC of unit cells of different converges. The reason is that the increasing contact area

dominates the heat transport in the unit cell, and the gas gap becomes less significant.

Figure 4: The effective thermal conductivity as a function of axial deformation in different sized

powders, at P=100 kPa and T=298 K

504 Advances of Computational Mechanics in Australia

Conclusion

The effective thermal conductivity of compacted submicron powders was studied for various

loading conditions, including gas pressure, temperature and deformation. A microscopic unit cell

was established to provide the basis for evaluating overall conductivity of the packed beds and was

analysed using finite element method. The Knudsen number has been incorporated in the gas

conduction model to include the influence of gap distance, gas pressure and temperature. The

numerical result shows a clear dependency of the effective thermal conductivity on parameters of

particle size, gas pressure and deformation. Increasing gas pressure tends to elevate the overall

conductivity at high pressure. When interparticles contacts are form, smaller sizes lead to larger

overall conductivity of unit cells. With increasing compaction, the overall conductivity increases

and the solid-solid contact conduction dominates the heat transfer. This study provides a theoretical

basis to consider the heat transfer in compacted beds with submicron-sized particles under a wide

range of loading conditions.

Acknowledgement

Financial support for this research from the Australian Research Council through grants

DE130101639 is greatly appreciated.

References

[1] Y. Gan, Francisco Hernandez, Dorian Hanaor, Ratna Annabattula, Marc Kamlah and Pavel

Pereslavtsev, "Thermal Discrete Element Analysis of EU Solid Breeder Blanket Subjected

to Neutron Irradiation," Fusion Science & Technology, vol. 66, pp. 83-90, 2014.

[2] A. M. Abyzov, A. V. Goryunov, and F. M. Shakhov, "Effective thermal conductivity of

disperse materials. I. Compliance of common models with experimental data," International

Journal of Heat and Mass Transfer, vol. 67, pp. 752-767, 2013.

[3] A. M. Abyzov, A. V. Goryunov, and F. M. Shakhov, "Effective thermal conductivity of

disperse materials. II. Effect of external load," International Journal of Heat and Mass

Transfer, vol. 70, pp. 1121-1136, 2014.

[4] E. Tsotsas and H. Martin, "Thermal conductivity of packed beds- A review," Chemical

Engineering Process, vol. 22, pp. 19-37, 1987.

[5] A. V. Gusarov and E. P. Kovalev, "Model of thermal conductivity in powder beds,"

Physical Review B, vol. 80, pp. 024202(1)-024202(12), 2009.

[6] R. A. Crane and R. I. Vachon, "A prediction of the bounds on the effective thermal

conductivity of granular materials," International Journal of Heat and Mass Transfer, vol.

20, pp. 711-723, 1977.

[7] M. Bahrami, M. M. Yovanovich, and J. R. Culham, "Effective thermal conductivity of

rough spherical packed beds," International Journal of Heat and Mass Transfer, vol. 49, pp.

3691-3701, 2006.

[8] Y. Liang and X. Li, "A new model for heat transfer through the contact network of

randomly packed granular material," Applied Thermal Engineering, vol. 73, pp. 984-992,

2014.

[9] G. Weidenfeld, Y. Weiss, and H. Kalman, "A theoretical model for effective thermal

conductivity (ETC) of particulate beds under compression," Granular Matter, vol. 6, pp.

121-129, 2004.

[10] G. K. Batchelor and R. W. O'Brien, "Thermal and Electrical Conduction Through a

Granular Material," The Royal Society Proceding A, vol. 355, pp. 313-333, 1977.

[11] H. W. Zhang, Q. Zhou, H. L. Xing, and H. Muhlhaus, "A DEM study on the effective

thermal conductivity of granular assemblies," Powder Technology, vol. 205, pp. 172-183,

2011.

Applied Mechanics and Materials Vol. 846 505