Influence of Solder Condition On Effective Thermal Conductivity of
Influence of Solder Condition On Effective Thermal Conductivity of
Influence of Solder Condition On Effective Thermal Conductivity of
a r t i c l e i n f o a b s t r a c t
Article history: It is indicated that the solder joint of the metal fibrous materials is a critical factor impacting the heat
Received 19 September 2022 conduction. To reveal the mechanism by which solder joint sizes, solder joint skips, solder flux materials,
Revised 31 October 2022
and filling media affect the thermal conductivity of fibres, pore-scale numerical simulation is employed to
Accepted 22 November 2022
study the thermal transport in two-directional (2-D) random fibres. Satisfactory agreement with existing
Available online 30 November 2022
data validates the numerical model. The dimensionless effective thermal conductivity (ETC) of the porous
Keywords: fibres increases with the solder joint sizes. As the solder joint size (i.e., solder joint ratio) increases by
Metal fibre materials 3.06%, the in-plane (ke-in ) and out-of-plane (ke-out ) dimensionless ETC increase by 9.0% and 437.2%, re-
Effective thermal conductivity (ETC) spectively. However, the solder joint skips will weaken the thermal conductivity of the fibres. For the
Numerical simulation same fibre, the ETC of the fibre increases as the thermal conductivity of solders increases. Further, when
Solder joints the dissimilarity in thermal conductivity between the filling medium and the fibre is reduced, the fibre
is less affected by the solder joint skips. Finally, it should be supplemented that the in-plane and out-of-
plane ETC (ke-in and ke-out ) of the fibre without any solder joint are reduced by an average of 14.3% and
98.8%, respectively.
© 2022 The Author(s). Published by Elsevier Ltd.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)
1. Introduction MFM is formed by stacking fibres layer by layer, where each layer
of the fibres is randomly distributed in the plane.
Metal fibrous materials (MFMs) have excellent characteristics MFMs have extremely complex structures due mainly to the
such as lightweight, high porosity, strong damage resistance, heat entanglements and random arrangements in porous structures
transfer, high specific surface area, easy moulding, and low man- [27,28]. This makes it more difficult to determine their transport
ufacturing cost [1–6]. MFMs have attracted extensive attentions in properties. To reveal the transport mechanism in MFM, different
engineering applications, such as thermal management [7–11], bal- methods including theoretical analysis [29–31], numerical simula-
listic protection [12,13], composite reinforcement [14,15], fuel cells tions [32–37], and experimental measurements [38–41] have been
[16–18], catalysts [19], enhanced thermal storage [20–22], and en- taken to characterize the fibre pore structures and to explore their
ergy absorption [23,24]. For these engineering applications, the ef- transport properties.
fective thermal conductivity (ETC) of MFM is a vital parameter that Given the complicated pore structure of metal fibre materials,
determines heat conduction capability [25,26]. However, the fila- most theoretical models simplify the complicated porous struc-
ment orientations of fibrous materials, e.g., two-directional (2-D) tures into ideal periodic unit cells (UCs). Sadeghi et al. [29] pro-
random fibres as the representative, result in different in-plane and posed an analytical model to evaluate the ETC of the fibrous ma-
out-of-plane features for the heat conduction and their ETCs. 2-D terials using cylindrical fibres with an equal spacing and uniform
size. Hooman and Dukhan [42] also established a theoretical model
based on a cubic arrays of cylinders to depict the transport prop-
∗∗
Corresponding author at: State Key Laboratory of Eco-hydraulics in Northwest erties of porous media. The model has good prediction accuracy.
Arid Region, Xi’an University of Technology, Xi’an 710048, PR China. Similarly, Qu et al. [30] chose an octet-truss lattice UC to estab-
∗
Corresponding author.
lish an analytical expression for the ETC of the isotropic fibres. In
E-mail addresses: xiaohuyang@xjtu.edu.cn (X. Yang), K.Hooman@tudelft.nl (K.
Hooman).
addition, a theoretical model for predicting the ETC of sintered fi-
https://doi.org/10.1016/j.ijheatmasstransfer.2022.123715
0017-9310/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
2
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
Fig. 1. 2-D random fibre geometric model: (a) metal fibre [30]; (b) schematic diagram of 2-D random metal fibre; (c) representative unit cell structure; (d) dimensions for
the computational model.
Fig. 2. Boundary conditions for the ETC: (a) out-of-plane; (b) in-plane.
Therefore, it is necessary to select a representative unit cell (RUC) = 0). Therefore, the governing equation is simply the energy
(V
that contains the main characteristics of the MFM. Fig. 1(c) displays equation.
the RUC, which is composed of one full layer fibres and two half- Energy equation for solid phase:
layer fibres. The inherent nature such as random distribution, over-
lap and contact of fibres has been well interpreted by the RUC. The
∇ · (ks ∇ Ts ) = 0 (1)
fibre layers are in contact with each other by the solder joints. Fur- Energy equation for fluid phase:
thermore, there exist overlaps between the randomly-distributed
∇ · k f ∇ Tf = 0 (2)
fibres in the same layer. Fig. 1(b) denotes that the length, width,
and height of the RUC are separately 2.4 mm, 2.4 mm, and 0.4 mm, where, ks and k f represent the ETC of solid phase and fluid phase,
according to the previous study [39] that it is reasonable to set the respectively; Ts and T f denote the temperature of solid phase and
diameter of representative structural fibres to 0.2 mm. fluid phase, in respective. In addition, Fig. 2(a) and (b) demonstrate
the boundary conditions for the out-of-plane and in-plane mod-
2.2. Effective thermal conductivity modelling elling on the ETC. In Fig. 2(a), the temperature is set to 300 K (i.e.,
T1 = 300 K) in the top z-direction, and the bottom surface temper-
ANSYS-Fluent 19.1 is used for the pore-scale simulation, provid- ature in the z-direction is set to 299 K (i.e.,T2 = 299 K). Similarly,
ing better understandings on the heat conduction mechanism of the front surface temperature in the y-direction is 300 K (i.e.,T1 =
metal fibres at microporous scale. Since the fibres are randomly 300 K), and the rear surface temperature in the y-direction main-
distributed in a layer-by-layer pattern, the joint conditions among tains at 299 K (i.e.,T2 = 299 K) in Fig. 2(b). The remaining surfaces
various fibres are critical to heat conduction. In the current study, in Fig. 2(a) and (b) are symmetric. Finally, the residuals of the en-
we consider only the heat conduction of the model and ignore ergy equation are set to 10−10 to ensure the convergence of the
the convection. Furthermore, the fluid in the model is stationary calculation.
3
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
Table 1
Physical parameters for MFM.
Note:
a
Sn63: 106 (Sn63) Sn-Pb Solder Alloy;
b
PCM1: CaCl2 · 6H2 O(95% )+Mg(NO3 ) · 6H2 O(5% ) [52];
c
PCM2: CaCl2 ·6H2 O/EG [53].
tional cost and reliability, the number of mesh for subsequent nu-
merical simulations is not less than the number of Mesh-2.
3. Simulation validation
4
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
fluence upon both ke-in and ke-out . As seen, both . Compared with
ke-in , solder joint size makes greater contribution to ke-out . When
ϕ merely increases from 0.08% to 3.06%, ke-out increase by 437.23%;
while this increment ratio is only 9.04% for ke-in . It evinces that the
solder joint size plays a critical role in determining heat conduc-
tion across the micro structure in MFM. In addition, the influence
of metal materials (copper, aluminum, nickel, and steel) on the
ETC is justified. As demonstrated, except for the steel with much
lower thermal conductivity than other competing metallic materi-
als, copper, aluminum, and nickel fibrous materials have much at
one value of the ETC. Besides, the variation of solder joint size has
the same influence on ke-in and ke-out for porous fibres with differ-
ent metal materials.
To further reveal the impact of the solder joint size on the out-
of-plane heat conduction characteristics, Figs. 7 and 8 compare the
distribution of heat flux and temperature for the out-of-plane heat
conduction in copper porous fibres with different solder joint sizes,
respectively. Three solder ratios (0.37%, 1.55%, and 3.06%) are con-
sidered in Fig. 7, with the red areas in the local enlargement of the
UC in Fig. 7(a1)–(c1) denoting the solder joints. As seen, a larger
solder flux ratio causes a bigger solder joint. With high tempera-
ture imposed on the top of the UC, heat conduction occurs across
the three layers of fibres. As found in Fig. 7, much higher heat flux
is concentrated in the solder joint, compared to the other areas of
the same fibre. The larger the solder joint (the higher solder ra-
tio), the higher the heat flux. Heat conduction is obliged to pass
through the fibre layers along the out-of-plane direction. Hence the
solder joints between each fibre layer play a bridge role in trans-
Fig. 5. Comparison of numerical results with previous measurement data [35,39] porting thermal energy. When the size of the solder joint increases,
and simulation data [35]: (a) in-plane 2-D MFM; (b) out-of-plane 2-D MFM. the cross-sectional area of the solder joint increases accordingly.
According to the thermal resistance formula (R = L/(ks f S )), an in-
crease in the cross-sectional area S of solder joint results in a de-
crease in the thermal resistance R of the solder joint when the
where, VSJ and Vs denote the total volume of the solder joint and length L of the heat conduction and the thermal conductivity ksf of
the fibres, respectively. The solder joint ratio does not exceed 3.5% the solder joint remain constant. With the reduced thermal resis-
in this study for the following reasons: (i) in the engineering fabri- tance of the solder joint, heat is more easily conducted through the
cation routine, previous studies [54,55] have evinced that MFM are solder joint. Therefore, the larger size solder joints in Fig. 7 have
fabricated within 5% solder flux (ϕ = 5%); (ii) excessive solder flux higher heat flux. Comparably, the in-plane heat conduction is pri-
will cause multiple solder joints, forming overlap or weld flash; marily along the direction of the fibres. Therefore, the solder joint
(iii) in the current simulation model (see Fig. 1(b) and (c)), some size exerts a weaker influence on in-plane heat conduction than
solder joints overlap when the sintered flux usage exceeds 3.5%. that for the out-of-plane heat conduction. In Fig. 8, the tempera-
Fig. 6 demonstrates the effect of solder flux ratio on the in- ture distributions of the solder joints with various sizes are almost
plane (ke-in ) and out-of-plane (ke-out ) dimensionless ETC for dif- the same. It indicates that the effect of solder joint size on the
ferent metal fibre materials. A higher solder flux ratio leads to a temperature distribution is not obvious in the out-of-plane thermal
bigger solder joint. The solder joint size causes a remarkable in-
Fig. 6. Effect of solder flux ratio on in-plane (ke-in ) and out-of-plane ETCs.
5
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
Fig. 7. Heat flux distribution for 2-D copper fibre with different sizes of solder joints: (a) ϕ = 0.37%; (b) ϕ = 1.55%; (c) ϕ = 3.06%.
Fig. 8. Temperature distribution for 2-D copper fibre with different solder joint sizes: (a) ϕ = 0.37%; (b) ϕ = 1.55%; (c) ϕ = 3.06%.
conduction. However, as ϕ increases from 0.37% to 3.06%, ke-out of tween the ETC of the MFM and the number of the solder joint
the copper porous fibres increases by 65.5%. skips. Solder joint skips commonly happen when soldering met-
als. It is a kind of weld defect that significantly affects both the
4.2. Effect of solder joint skips mechanical and thermal performance of the MFM with subject to
thermal-mechanical loadings. Fig. 9(a) and (b) separately manifest
To further reveal the impact of the solder joints on the ETC the RUC structure with complete solder joints and the one with
of the MFM, this section focuses on analyzing the correlation be- solder joint skips. In particular, the closest solder joints on the up-
6
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
Fig. 9. Solder joint diagram: (a) complete solder joints; (b) solder joint skips.
Fig. 10. Effect of solder joint skips on dimensionless ETC: (a) in-plane; (b) out-of-plane.
7
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
Fig. 11. Heat flux distribution for 2-D copper fibre with the different volume percentage of solder joint skips: (a) ω = 0%; (b) ω = 23.5%; (c) ω = 51.5%; (d) ω = 76.6%; (e)
ω = 100%.
per and lower layers are skipped together in the z-direction of the noteworthy in Fig. 10 that the impact of solder joint skips on in-
RUC in Fig. 9(b). plane heat conduction is much smaller than out-of-plane heat con-
To determine quantitatively the solder joint skips, a parameter duction. The primary physical reason is that the out-of-plane heat
named solder joint skip ratio (ω) is introduced: conduction between each fibre layer is mainly through the solder
VSJS joints, whereas the in-plane heat conduction is chiefly along the
ω= × 100% (5) direction of the fibres in the same plane. When the solder joint
VSJ
is skipped, the out-of-plane heat conduction is hindered. The low
where, VSJS and VSJ represent the volume of the solder joint skips thermal conductivity medium (typically the filling fluid) at the sol-
and the total solder joints, in respective. When ω is equal to 0% der joint skips is not conducive to heat conduction. Therefore, the
(ω = 0%), all solder joints are intact. With it reaching 100% (ω = out-of-plane dimensionless ETC diminishes remarkably with the
100%), there exist no solder joints. solder joint skips.
Fig. 10 depicts the variation of ke-in and ke-out as a function To supplementary visualize the effect of the solder joint skips
of the solder joint skip ratio. As seen, both ke-in and ke-out de- on the out-of-plane heat conduction, the heat flux distribution and
crease with an increase in ω, demonstrating an approximately lin- temperature field of the copper fibres are compared in Figs. 11 and
ear fashion. ke-in decreases by 15.60%, 15.27%, 14.57%, and 11.67% 12, respectively. An increase in the number of solder joint skips
for copper (Cu), aluminum (Al), nickel (Ni), and steel fibres as ω in- results in an increase in the solder joint skip ratio (ω). In Fig. 11,
creases from 0% to 100%, respectively. However, ke-out of the Cu, Al, the left column demonstrates the computational models with and
Ni, and steel fibres separately decreased by 99.84%, 99.70%, 99.34% without solder joint skips. Five solder joint skip ratios are com-
and 96.38% as ω increases from 0% to 100%. It displays that the pared regarding the heat flux distribution on the metal fibres and
impact of solder joint skips on the dimensionless ETC is weak- the corresponding filling fluids. The heat flux in the areas where
ened when the conductivity of the fibre material is smaller. It is the solder joint is intact is much greater than the areas with the
8
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
Fig. 12. Temperature distribution for 2-D copper fibre with the different volume percentage of solder joint skips: (a) ω = 0%; (b) ω = 51.5%; (c) ω = 100%.
solder joint skips (see the 2nd and 3rd column in Fig. 11). For main reason is that there are some solder joint skips in area A,
the areas of solder joint skips, the heat flux in the correspond- while the solder joints in area B are unimpaired (here ω = 51.5%).
ing areas is significantly reduced, and there is no obvious differ- Similarly, without any solder joint (i.e., ω = 100%), the temperature
ence from the surrounding areas. It further indicates that the sol- gradients of the middle-layer fibre in area A and B are remark-
der joints play a substantial role in enhancing out-of-plane heat ably reduced compared with the intact solder joints (i.e., ω = 0%).
transfer, in consistence with the continuously-decreasing trend of This phenomenon is caused by the constant temperature of the
ke-out as a function of ω. Especially, when the solder joints are top and bottom surfaces in the z-direction during the simulation.
intact (ω = 0%), the heat flux in area A is much larger than that The higher thermal conductivity leads to a larger local temper-
in area B in Fig. 11(a3). This is due to the fact that the upper ature gradient in the middle layer of the fibres. Therefore, the
and lower solder joints in area A are closer than those in area B, temperature gradient in the area where the solder joint exists is
thereby providing an easier way for heat conduction. Moreover, the larger.
solid solder joint is an essential factor affecting the out-of-plane Meanwhile, it should be specially noted that the heat flux
heat conduction. When the solder joints are skipping, the medium distributions in different solder joint regions are various in
at the solder joint skips becomes the filling fluid. Thus, the ther- Fig. 11 without the solder joint skips (ω = 0%). The thermal resis-
mal conductivity at the solder joint skips decreases dramatically. tance of the different solder joints are calculated to quantify the
Based on the thermal resistance equation (R = L/(ks f S )), an appar- dissimilarity in the contribution of the various solder joints. The
ently decrease in the thermal conductivity ksf of solder joint and thermal resistance formula is expressed as:
an corresponding increase in the length L of the heat conduction
lead to an increase in the thermal resistance R of the solder joint Tt−i − Tb−i
Ri = (6)
when the cross-sectional area S of the solder joint is fixed. With Pi
the increased thermal resistance of the solder joint, heat is more
difficult to conduct through the solder joint skips. Therefore, the where, Tt and Tb are the top surface temperature of the upper sol-
heat flux decreases sharply at the solder joint skips (see Fig. 11). der joints and the bottom surface temperature of the lower solder
The temperature distribution of the solder joint skips also displays joints in the z-direction; P is the power of the top solder joints
apparent dissimilarity in Fig. 12. In the regions where the solder on the top surface along the z-direction; i denotes the ith solder
joints exist, the temperature gradient of the intermediate layer fi- joint. As illustrated in Fig. 13(a), St−i represents the ith top sol-
bres in the z-direction is larger. In contrast, the temperature gradi- der joint and Sb−i denotes the ith bottom solder joint. The detailed
ent in the regions with the solder joint skips decreases (see areas numbering of the top and bottom solder joints are documented in
A and B in Fig. 12(a2)). When the solder joint skips ratio reaches Fig. 13(b) and (c), in respective. In addition, Ri is the thermal re-
51.5% (ω = 51.5%) (see Fig. 12(b2)), the temperature gradient of sistance between the ith top solder joint and the ith bottom solder
the middle-layer fibre in area A with the solder joint skips is dra- joint (see Fig. 13(a)). For instance, R2 depicts the thermal resistance
matically changed, while that in area B remains almost unchanged between the 2nd top solder joint (St−2 ) and the 2nd bottom solder
compared with the intact solder joints (ω = 0%) in Fig. 12(a2). The joint (Sb−2 ) in Fig. 13(a). Further, Fig. 13(d) demonstrates the distri-
9
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
Fig. 13. Fibre solder joints: (a) schematic diagram of solder joints distribution; (b) distribution and numbering of top solder joints; (c) distribution and numbering of bottom
solder joints; (d) thermal resistance distribution between top and bottom solder joints; (e) thermal resistance of solder joints.
bution of each thermal resistance for the solder joints. The grada- a bit lower. The major reason for this difference is that the dis-
tion and range of color indicate the value of the thermal resistance. tances between the corresponding solder joints of the upper and
To analyze the thermal conductivity between the top and cor- lower layers are different. It can be noticed that the distance be-
responding bottom solder joints, Fig. 13(e) compares the thermal tween the upper and lower solder joints of red number areas (1,
resistance of the different solder joints. The thermal resistance of 5, 7 and 9) is smaller, and thus the distance of heat conduction is
different solder joints is diverse in Fig. 13(e). However, it can be shorter. In addition, the heat conduction distance of the upper and
divided into three parts according to the different thermal resis- lower solder joints for the blue number areas (2, 3, 14, and 15) is
tances: the first part of the solder joints (red number areas 1, 5, much greater than that of the other upper and lower solder joints.
7, and 9 in Fig. 13(d)) with thermal resistance less than 21 K/W Therefore, the heat flux in blue number areas (2, 3, 14, and 15) is
(denoted as solder joint I); then the second part of solder joints remarkably reduced.
(blue number area 2, 3, 14, and 15 in Fig. 13(d)) with thermal re-
sistance greater than 31 K/W (marked as solder joint II); the last 4.3. Thermal conductivity of solder flux
part is the rest of the solder joints with thermal resistance in the
range of 21 K/W to 31 K/W (labelled as solder joint III). The heat The solder joints size and quality for the fibres are vital fac-
flux distribution diagram without the solder joint skips (ω = 0%) tors affecting the out-of-plane heat conduction. The solder joints
in Fig. 13(d) demonstrates that the heat flux in the solder joint discussed in the previous sections are the same material as the
I is much higher, and the heat flux in the solder joint II area is fibres. In practice, the solder of the fibre is often not the same
10
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
Fig. 14. (a) Out-of-plane dimensionless ETC of the copper fibres with various solder flux; (b) Out-of-plane dimensionless ETC of the fibres as a function of the solder joint
skips.
composition as the fibre; hence the thermal conductivity of sol- ponent of the solder joint is copper, ke-out of the copper fiber with
ders and fibres may be inconsistent. This section discusses the in- the solder joint skips decays the fastest (Fig. 14(b)). Conversely, as
fluence of different thermal conductivity of solder joints upon the the solder joint is mainly composed of Plumbum, the declining rate
out-of-plane heat conduction. of ke-out with the solder joint skips is the slowest. For solder joints
Fig. 14 indicates the effect of solder joint with different ther- with Cu, Sn, Sn63, or Pb, the declining rate (slope of the curve)
mal conductivity on the out-of-plane dimensionless ETC of the of ke-out caused by the solder joint skips is approximately 0.079,
fibre (ke-out ). Although ke-out increases steadily with the solder 0.032, 0.027, and 0.020, respectively.
joint thermal conductivity for the intact solder joints in Fig. 14(a), To completely analyze the influence of the solder joints with
the relationship is nonlinear. When the conductivity of the sol- various thermal conductivities on the out-of-plane heat conduc-
der joint material is less than 150 W · m−1 · K−1 (i.e., ks f 150W · tion, Figs. 15 and 16 compare the heat flux distribution and tem-
m−1 · K−1 ), the variation range of ke-out occupies 66.8% of the en- perature distribution of the RUCs. Fig. 15(a1)–(c1), and (d1) are the
tire variation range for the copper fibre. It demonstrates that it has schematic diagrams of solder joints with different thermal conduc-
a distinct impact on ke-out when ksf is less than 150 W · m−1 · K−1 . tivities, wherein solder joints with different colors represent differ-
Therefore, the solder flux is a non-negligible element that affects ent thermal conductivities. The blue, green, yellow, and red solder
the out-of-plane heat conduction of the MFM. Although ke-out of joints represent the thermal conductivities from low to high (see
the copper fibre increases with the thermal conductivity for the the first column of Fig. 15). The RUC has complete solder joints in
solder joint, the influence of the solder joint thermal conductivity Figs. 15 and 16. The heat flux of the solder joint areas is not appar-
on the overall out-of-plane heat conduction of the fibres is weak- ently distinct from the surrounding areas when the thermal con-
ened. For four different solder joint materials, the influence of the ductivity of the solder joint material is equal to 50 W · m−1 · K−1
solder joint skips on ke-out is shown in Fig. 14(b). ke-out of the cop- (ks f = 50W · m−1 · K−1 ). Currently, only the thermal conductivity of
per fibres with different solder joint materials decreases with the the solder flux is different, while the length L of the heat con-
increase in the solder joint skips ratio (ω). Referring to Table 1, duction and the cross-sectional area S of the solder joint remain
it can be uncovered that the smaller the thermal conductivity of constant. By the thermal resistance formula (R = L/(ks f S )), solder
the solder joints, the smaller the ke-out for copper fibre. Besides, fluxes with different thermal conductivity (ksf ) lead to variations in
it should also be noted that the impact of solder joints with low the thermal resistance of the corresponding solder joint. Therefore,
thermal conductivity on ke-out is weakened. When the main com- when ksf increases, the heat flux of the solder joint areas increases
11
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
Fig. 15. Heat flux distribution for 2-D copper fibre with different thermal conductivity of solder joints: (a) ks f /ks = 0.125; (b) ks f /ks = 0.375; (c) ks f /ks = 0.625; (d) ks f /ks =
0.875.
accordingly, exacerbating the difference between the heat flux of skips have less impact on ke-out (see Fig. 17(d)). On the one hand,
the solder joint areas and the surrounding areas. This demonstrates without solder joints, ke-out of the copper fibres saturated with
that the thermal conductivity of solder joints is one of the vital air decreases by 99.84% in Fig. 17(a). On the other hand, ke-out of
factors influencing the out-of-plane heat conduction of the fibres. the steel fibre saturated with PCM2 is reduced by only 10.07% in
Therefore, improving the thermal conductivity of the solders is a Fig. 17(d). To be conclusive, Fig. 17 reveals that the filling fluids
meaningful measure to promote the out-of-plane heat conduction have a grave effect on ke-out for low-conducting fibres. When the
of fibres. Furthermore, the temperature gradient of the interme- discrepancy in thermal conductivity between the fibre and the fill-
diate layer fibres between the upper and lower solder joints in- ing medium is reduced, the solder joint skips may become a sec-
creases as the thermal conductivity of the solder joints approaches ondary factor affecting out-of-plane heat conduction.
that of the fibre material (see region A in Fig. 16). This is chiefly Furthermore, Fig. 18(a)–(d) compare the association between
because the high thermal conductivity of the solder joints facili- ke-out and the solder joint skip ratio, where four kinds of fluids,
tates heat conduction between the fibre layers. including air, water, PCM1, and PCM2 are filled in the MFMs, re-
spectively. As seen, an increase in ω is often accompanied by a
4.4. Different filling fluids decrease in ke-out . Low-conducting fibre (steel) leads to a moderate
decreasing trend for ke-out with ω, with compared to other metal
MFMs filled with different fluids such as air, water, and phase fibres in Fig. 18. Moreover, this phenomenon becomes more pro-
change material (PCM) have distinctive engineering scenarios. To nounced when the thermal conductivity of the filling medium in-
address this issue, this section focuses on studying the impact of creases. On the one hand, the variation trend of ke-out of the four
different filling fluids upon ke-out of the MFMs with solder joint MFMs (Cu, Al, Ni, and steel) is consistent and the difference is
skips. For the solder joint skips, Fig. 17(a)–(d) severally compare slight when the filling medium is air (kair = 0.0242W · m−1 · K−1 ).
the relationship between ke-out of the fibres and the filling fluids. On the other hand, the difference in the variation trend of ke-out
Four kinds of MFMs including copper, aluminum, nickel, and steel of the fibres becomes visible when the filling medium is PCM2
are considered. As demonstrated in Fig. 17, there exists a steady (kPCM2 = 3W · m−1 · K−1 ). For instance, when the filling medium is
decline trend of ke-out as ω increases. Moreover, the filling fluid air with low thermal conductivity, ke-out of the copper and steel
with higher thermal conductivity can considerably promote ke-out fibre without solder joints are reduced by 99.8% and 96.4% (see
for the same fibre. The smaller difference in thermal conductivity Fig. 18(a)), respectively. Likewise, when the filling medium is re-
between the fibres and the filler medium makes the solder joint placed by PCM2 with high thermal conductivity (Fig. 18(d)), ke-out
12
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
Fig. 16. Temperature distribution for 2-D copper fibre with different thermal conductivity of solder joints: (a) ks f /ks = 0.125; (b) ks f /ks = 0.375; (c) ks f /ks = 0.625; (d)
ks f /ks = 0.875.
Fig. 17. Comparing the out-of-plane ETC of fibres with different filling fluids: (a) copper; (b) aluminum; (c) nickel; (d) steel.
of the same MFMs decreases by 83.3% and 10.1%, in respective. The conductivity of fibres decreases, the impact of the solder joint skips
influence of the solder joint skips on ke-out decreases with the in- on ke-out also weakens. In general, the ratio of the thermal conduc-
crease in the conductivity of the filling medium. As the thermal tivity of the filling medium to that of the fiber increases, and the
13
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
Fig. 18. Comparing the out-of-plane ETC of different fibres with the same filling fluids: (a) air; (b) water; (c) PCM1; (d) PCM2.
effect of the solder joint skips on ke-out weakens. conductivity of the solder joints enhances heat conduction be-
tween the fibre layers.
(5) The influence of the solder joint skips on ke-out is weakened as
5. Conclusion
the ratio of the thermal conductivity of the filling medium to
that of the fiber increases. Without any solder joint, the out-of-
In this study, the influence mechanisms of the solder joint
plane ETCs of the copper fibre saturated with air and the steel
sizes, the solder joint skips, solder flux materials, and filling flu-
fibre saturated with PCM2 are reduced by 99.8% and 10.1%, re-
ids on the heat conduction of 2-D random porous fibres are stud-
spectively. At this point, the ratio of the thermal conductivity of
ied. A numerical model is built based on the reconstruction of the
the copper fibre to air is 6.08 × 10−5 , but the ratio of the steel
microstructure. The numerical model is in satisfactory agreement
fibre to PCM2 is 0.18.
with the existing data, which verifies the rationality of the numer-
ical simulation. The following conclusions can be drawn: The current study comprehensively considers the factors affect-
ing the thermal conductivity of 2-D random porous fibers. Quanti-
(1) The dimensionless effective thermal conductivity (ETC) of the tative relationships between influencing factors (including the sol-
porous fibres increases with the solder joint size. As the solder der joint sizes, the solder joint skips, solder flux materials, and fill-
joint ratio increases by 3.06%, the in-plane (ke-in ) and out-of- ing fluids) and heat conduction are established by pore-scale sim-
plane (ke-out ) dimensionless ETC increase by 9.0% and 437.2%, ulation. Therefore, the current numerical model can provide refer-
respectively. In contrast, the solder joint size mainly affects the ence and guidance for engineering applications.
out-of-plane heat conduction, and the influence on the in-plane
heat conduction is subordinate. Declaration of Competing Interest
(2) An increase in the solder joint skip ratio causes an apparent
decline in the ETC of the fibres. Without any solder joint, ke-in The authors declare that they have no known competing finan-
and ke-out are reduced by an average of 14.3% and 98.8%, re- cial interests or personal relationships that could have appeared to
spectively. Similarly, the solder joint skips dramatically affect influence the work reported in this paper.
the out-of-plane heat conduction, and their effect on in-plane The authors declare the following financial interests/personal
heat conduction is secondary. relationships which may be considered as potential competing in-
(3) The intact solder joint is a vital factor affecting the out-of-plane terests.
heat conduction of the fibres. Furthermore, the distance be-
tween the adjoining solder joints of the upper and lower layers CRediT authorship contribution statement
is the key factor that cannot be neglected in determining the
out-of-plane heat conduction. Tian Xiao: Formal analysis, Validation, Writing – original draft.
(4) The out-of-plane ETC ke-out of the fibre increases with the ther- Qiaoling Zhang: Data curation, Visualization, Writing – review &
mal conductivity of the solder joints. The increased thermal editing. Xiaohu Yang: Conceptualization, Methodology, Writing –
14
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
review & editing, Supervision. Kamel Hooman: Supervision, Writ- [25] T. Xiao, J. Guo, X. Yang, K. Hooman, T.J. Lu, On the modelling of heat and fluid
ing – review & editing. Guodong Li: Data curation, Supervision. transport in fibrous porous media: analytical fractal models for permeability
and thermal conductivity, Int. J. Therm. Sci. 172 (2022) 107270.
[26] Y. Mahmoudi, K. Hooman, K. Vafai, Convective Heat Transfer in Porous Media,
Data availability CRC Press, 2019.
[27] J. Xie, J. Fang, L. Chen, W. Jiao, Z. Yang, L. Chen, Micro-scale modeling of 3D
Data will be made available on request. needled nonwoven fiber preforms, Compos. Struct. 281 (2022) 114995.
[28] P.H. Jadhav, T. G, N. Gnanasekaran, M. Mobedi, Performance score based mul-
ti-objective optimization for thermal design of partially filled high porosity
Acknowledgments metal foam pipes under forced convection, Int. J. Heat Mass Transf. 182 (2022)
121911.
This work was supported by the Open Research Fund Program [29] E. Sadeghi, M. Bahrami, N. Djilali, Analytic determination of the effective ther-
mal conductivity of PEM fuel cell gas diffusion layers, J. Power Sources 179 (1)
of State Key Laboratory of Eco-hydraulics in Northwest Arid Region, (20 08) 20 0–208.
Xi’an University of Technology (2020KFKT-5), and the National Nat- [30] Z. Qu, T. Wang, W. Tao, T. Lu, A theoretical octet-truss lattice unit cell model
ural Science Foundation of China (51976155). One of the authors for effective thermal conductivity of consolidated porous materials saturated
with fluid, Heat Mass Transf. 48 (8) (2012) 1385–1395.
(Xiaohu Yang) gratefully acknowledged the support of K. C. Wong [31] T. Xiao, X. Yang, K. Hooman, L. Jin, C. Yang, T.J. Lu, Conductivity and perme-
Education Foundation. ability of graphite foams: analytical modelling and pore-scale simulation, Int.
J. Therm. Sci. 179 (2022) 107706.
References [32] A.N. Volkov, L.V. Zhigilei, Thermal conductivity of two-dimensional disordered
fibrous materials defined by interfiber thermal contact conductance and in-
[1] J.J. Remmers, R. de Borst, A. Needleman, The simulation of dynamic crack trinsic conductivity of fibers, J. Appl. Phys. 127 (6) (2020) 065102.
propagation using the cohesive segments method, J. Mech. Phys. Solids 56 (1) [33] M. Wang, J. He, J. Yu, N. Pan, Lattice Boltzmann modeling of the effective
(2008) 70–92. thermal conductivity for fibrous materials, Int. J. Therm. Sci. 46 (9) (2007)
[2] A. Vlot, J.W. Gunnink, Fibre Metal Laminates: An Introduction, Springer Science 848–855.
& Business Media, 2011. [34] R. Arambakam, H.V. Tafreshi, B. Pourdeyhimi, A simple simulation method for
[3] R. Alderliesten, R. Benedictus, Fiber/metal composite technology for future pri- designing fibrous insulation materials, Mater. Des. 44 (2013) 99–106.
mary aircraft structures, J. Aircr. 45 (4) (2008) 1182–1189. [35] X. Huang, Q. Zhou, J. Liu, Y. Zhao, W. Zhou, D. Deng, 3D stochastic model-
[4] G. Wu, J.M. Yang, The mechanical behavior of GLARE laminates for aircraft ing, simulation and analysis of effective thermal conductivity in fibrous media,
structures, JOM 57 (1) (2005) 72–79. Powder Technol. 320 (2017) 397–404.
[5] E.C. Botelho, R.A. Silva, L.C. Pardini, M.C. Rezende, A review on the develop- [36] L. Ni, Z. Chen, P. Mukhopadhyaya, X. Zhang, Q. Wu, Q. Yu, G. Miu, Numer-
ment and properties of continuous fiber/epoxy/aluminum hybrid composites ical simulation on thermal performance of vacuum insulation panels with
for aircraft structures, Mater. Res. 9 (3) (2006) 247–256. fiber/powder porous media based on CFD method, Int. J. Therm. Sci. 172 (2022)
[6] G. Wu, Y. Tan, J.M. Yang, Evaluation of residual strength of notched fiber metal 107320.
laminates, Mater. Sci. Eng. A 457 (1–2) (2007) 338–349. [37] M. Siavashi, S.M. Miri Joibary, Numerical performance analysis of a coun-
[7] T. Lu, H.A. Stone, M. Ashby, Heat transfer in open-cell metal foams, Acta Mater. ter-flow double-pipe heat exchanger with using nanofluid and both sides
46 (10) (1998) 3619–3635. partly filled with porous media, J. Therm. Anal. Calorimetry 135 (2) (2019)
[8] L. Tadrist, M. Miscevic, O. Rahli, F. Topin, About the use of fibrous materials in 1595–1610.
compact heat exchangers, Exp. Therm. Fluid Sci. 28 (2–3) (2004) 193–199. [38] C. Veyhl, T. Fiedler, O. Andersen, J. Meinert, T. Bernthaler, I. Belova, G. Murch,
[9] S. Mahjoob, K. Vafai, A synthesis of fluid and thermal transport models for On the thermal conductivity of sintered metallic fibre structures, Int. J. Heat
metal foam heat exchangers, Int. J. Heat Mass Transf. 51 (15–16) (2008) Mass Transf. 55 (9–10) (2012) 2440–2448.
3701–3711. [39] N. Haruki, A. Horibe, K. Nakashima, Anisotropic effective thermal conductivity
[10] C. Zhao, M. Opolot, M. Liu, F. Bruno, S. Mancin, K. Hooman, Numerical study measurement of various kinds of metal fiber materials, Int. J. Thermophys. 34
of melting performance enhancement for PCM in an annular enclosure with (12) (2013) 2385–2399.
internal-external fins and metal foams, Int. J. Heat Mass Transf. 150 (2020) [40] W. Li, Z. Qu, Experimental study of effective thermal conductivity of stainless
119348. steel fiber felt, Appl. Therm. Eng. 86 (2015) 119–126.
[11] H. Pourrahmani, M. Moghimi, M. Siavashi, Thermal management in PEMFCs: [41] P. Yi, L. Peng, X. Lai, M. Li, J. Ni, Investigation of sintered stainless steel fiber
the respective effects of porous media in the gas flow channel, Int. J. Hydrogen felt as gas diffusion layer in proton exchange membrane fuel cells, Int. J. Hy-
Energy 44 (5) (2019) 3121–3137. drogen Energy 37 (15) (2012) 11334–11344.
[12] J.H. Lin, Novel compound cushion layer for reinforcement of ballistic resis- [42] K. Hooman, N. Dukhan, A theoretical model with experimental verification
tance, Textile Res. J. 75 (5) (2005) 431–436. to predict hydrodynamics of foams, Transp. Porous Media 100 (3) (2013)
[13] F. Martínez-Hergueta, A. Ridruejo, F. Gálvez, C. González, J. LLorca, Influence 393–406.
of fiber orientation on the ballistic performance of needlepunched nonwoven [43] M. Sheng, D.R. Cahela, H. Yang, C.F. Gonzalez, W.R. Yantz Jr, D.K. Harris,
fabrics, Mech. Mater. 94 (2016) 106–116. B.J. Tatarchuk, Effective thermal conductivity and junction factor for sintered
[14] M.A. Choi, M.H. Lee, J. Chang, S.J. Lee, Permeability modeling of fibrous media microfibrous materials, Int. J. Heat Mass Transf. 56 (1–2) (2013) 10–19.
in composite processing, J. Non Newton. Fluid Mech. 79 (2–3) (1998) 585–598. [44] H. Moghimi, M. Siavashi, M. Mousavi Nezhad, A. Guadagnini, Pore-scale com-
[15] B.T. Åström, R.B. Pipes, S.G. Advani, On flow through aligned fiber beds and putational analyses of non-Darcy flow through highly porous structures with
its application to composites processing, J. Compos. Mater. 26 (9) (1992) various degrees of geometrical complexity, Sustain. Energy Technol. Assess. 52
1351–1373. (2022) 102048.
[16] J. Feser, A. Prasad, S.G. Advani, Experimental characterization of in-plane per- [45] P. Poureslami, M. Siavashi, H. Moghimi, M. Hosseini, Pore-scale convection–
meability of gas diffusion layers, J. Power Sources 162 (2) (2006) 1226–1231. conduction heat transfer and fluid flow in open-cell metal foams: a three-di-
[17] J.T. Gostick, M.W. Fowler, M.A. Ioannidis, M.D. Pritzker, Y.M. Volfkovich, mensional multiple-relaxation time lattice Boltzmann (MRT-LBM) solution, Int.
A. Sakars, Capillary pressure and hydrophilic porosity in gas diffusion layers Commun. Heat Mass Transf. 126 (2021) 105465.
for polymer electrolyte fuel cells, J. Power Sources 156 (2) (2006) 375–387. [46] M. Ramezanpour, M. Siavashi, A.Q. Raeini, M.J. Blunt, Pore-scale simulation of
[18] M.M. Heyhat, S. Mousavi, M. Siavashi, Battery thermal management with ther- nanoparticle transport and deposition in a microchannel using a Lagrangian
mal energy storage composites of PCM, metal foam, fin and nanoparticle, J. approach, J. Mol. Liquids 355 (2022) 118948.
Energy Storage 28 (2020) 101235. [47] E. Sepehri, M. Siavashi, Pore-scale direct numerical simulation of fluid dy-
[19] D.M. Burke, M.A. Morris, J.D. Holmes, Chemical oxidation of mesoporous car- namics, conduction and convection heat transfer in open-cell Voronoi porous
bon foams for lead ion adsorption, Sep. Purif. Technol. 104 (2013) 150–159. foams, Int. Commun. Heat Mass Transf. 137 (2022) 106274.
[20] G. Liu, Z. Du, T. Xiao, J. Guo, L. Lu, X. Yang, K. Hooman, Design and assess- [48] M. Shirbani, M. Siavashi, M. Hosseini, M. Bidabadi, Improved thermal energy
ments on a hybrid pin fin-metal foam structure towards enhancing melting storage with metal foam enhanced phase change materials considering various
heat transfer: an experimental study, Int. J. Therm. Sci. 182 (2022) 107809. pore arrangements: a pore-scale parallel lattice Boltzmann solution, J. Energy
[21] T. Xiao, G. Liu, J. Guo, G. Shu, L. Lu, X. Yang, Effect of metal foam on improving Storage 52 (2022) 104744.
solid–liquid phase change in a multi-channel thermal storage tank, Sustain. [49] T. Fiedler, N. White, M. Dahari, K. Hooman, On the electrical and thermal con-
Energy Technol. Assess. 53 (2022) 102533. tact resistance of metal foam, Int. J. Heat Mass Transf. 72 (2014) 565–571.
[22] M. Pan, Y. Zhong, Experimental and numerical investigation of a thermal man- [50] M. Opolot, C. Zhao, M. Liu, S. Mancin, F. Bruno, K. Hooman, Investigation of the
agement system for a Li-ion battery pack using cutting copper fiber sintered effect of thermal resistance on the performance of phase change materials, Int.
skeleton/paraffin composite phase change materials, Int. J. Heat Mass Transf. J. Therm. Sci. 164 (2021) 106852.
126 (2018) 531–543. [51] S. Maneemuang, N. Kammuang-Lue, P. Terdtoon, P. Sakulchangsatjatai, Effect
[23] Z. Chen, G. Fang, J. Xie, J. Liang, Experimental study of high-temperature tensile of pipe flattening on pressure drop in vapor core and thermal characteristic of
mechanical properties of 3D needled C/C–SiC composites, Mater. Sci. Eng. A miniature round and flat-shape heat pipe with sintered fiber wick, Int. J. Heat
654 (2016) 271–277. Mass Transf. 176 (2021) 121416.
[24] T.T. Li, R. Wang, C.W. Lou, J.H. Lin, Evaluation of high-modulus, puncture-re- [52] J.M. Charles, Performance and Stability of CaCl2 ˙ 6H2 O-Based Phase Change
sistance composite nonwoven fabrics by response surface methodology, J. Ind. Materials, Lehigh University, 2019.
Textiles 43 (2) (2013) 247–263. [53] R. Ye, W. Lin, X. Fang, Z. Zhang, A numerical study of building integrated
15
T. Xiao, Q. Zhang, X. Yang et al. International Journal of Heat and Mass Transfer 202 (2023) 123715
with CaCl2 · 6H2 O/expanded graphite composite phase change material, Appl. 11787–11793.
Therm. Eng. 126 (2017) 480–488. [55] J.T. Gostick, M.W. Fowler, M.D. Pritzker, M.A. Ioannidis, L.M. Behra, In-plane
[54] J. Sun, Z. Hu, J. Li, H. Zhang, C. Sun, Thermal and mechanical properties of and through-plane gas permeability of carbon fiber electrode backing layers, J.
fibrous zirconia ceramics with ultra-high porosity, Ceram. Int. 40 (8) (2014) Power Sources 162 (1) (2006) 228–238.
16