Study on the Fluid\u0026#x2013;Structure Interaction of Flexible Printed Circuit Board Electronics in the Flow Environment

$Page 1: Study on the Fluid\u0026#x2013;Structure Interaction of Flexible Printed Circuit Board Electronics in the Flow Environment$
Study on the fluid-structure interactionof flexible printed circuit board motherboard

in personal computer casingsWei Chiat Leong, Mohd Zulkifly Abdullah, Chu Yee Khor and Dadan Ramdan

School of Mechanical Engineering, Universiti Sains Malaysia, Nibong Tebal, Penang, Malaysia

AbstractPurpose – The flexible printed circuit board (FPCB) can be an alternative to the rigid printed circuit board because of its excellent flexibility, twistability,and light weight. Using FPCB to construct personal computer (PC) motherboard is still rare. Therefore, the present study aims to investigate the fluid-structure interaction (FSI) behaviors of the newly proposed FPCB motherboard under fan-flow condition in the PC casings.Design/methodology/approach – The deflection and stress induced, which are usually ignored in the traditional rigid motherboard, are the mainconcern in the current FPCB motherboard studies. Only a few studies have been conducted on the effect of inlet locations, effect of inlet sizes, effect ofmulti-inlets, and effect of a two-fan system. These numerical analyses are performed using the fluid flow solver FLUENT and the structural solverABAQUS; they are real-time online coupled by Mesh-based Parallel Code Coupling Interface (MpCCI).Findings – A smaller inlet size can cause higher deflection and stress fluctuations, but the fluctuations can be reduced by incorporating the multi-inletsdesign. In addition, the inlet locations and two-fan system can prominently affect the magnitudes of the deflection and stress induced.Practical implications – The current study provides better understanding and allows designers to be aware of the FSI phenomenon when dealingwith the FPCB motherboard. Although the present study primarily focuses on the motherboard, the findings could also contribute valuable informationfor other FPCB applications.Originality/value – The present study extends the FSI investigation from the previous novel approach of FPCB motherboard, and uniquely exploresthe behaviors of the FPCB motherboard inside different PC casings.

Keywords Fluid-structure interaction, FPCB motherboard, PC casings, Deflection, Stress, Computers, Circuits

Paper type Research paper

1. Introduction

The flexible printed circuit board (FPCB) can be an alternative

to the rigid printed circuit board. It has been proven

advantageous in several applications because of its excellent

flexibility, twistability, and lightweight. Over the years, FPCBs

have been widely used in numerous electronic devices, but their

application as a personal computer (PC) motherboard is still

rare. Compared with those by traditional rigid PCB, the

deflection and stress induced by the FPCB in the fan-flow

condition are far more prominent; hence, this issue is very

crucial in FPCB in terms of reliability. Many researchers have

focused on this area. A review of previous works, which builds a

background for the current study, is presented as follows.Azar and Russell (1990) experimentally studied the flow

distribution on a circuit pack. They found that the flow in

electronic enclosures is highly 3D. A similar study was reported

by Lee and Mahalingam (1993), who used a computational fluid

dynamics (CFD) tool to evaluate the velocity and temperature

fields of airflow in a computer system enclosure. Dealing with the

component-PCB heat transfer, Rodgers and his co-workers

(Eveloy et al., 2000; Rodgers et al., 2003a, b) performed

numerical and experimental works to assess the predictive

accuracy of CFD tools for the thermal analysis of electronic

systems. They also conducted a detailed characterization of the

airflow patterns around PCB-mounted electronic components

using several complimentary flow visualization techniques

(Lohan et al., 2002).Leung et al. (2000) developed a numerical solution to the

steady-state forced convection for air flowing through a

horizontally oriented PCB assembly under laminar flow

conditions. Shankaran and Karimanal (2002) proposed the

“zoom-in modeling” for accurate and time-efficient CFD design

calculations for a populated system cooled by forced airflow.

Leon et al. (2002) used the FLUENT code to obtain the optimal

layout for cooling fins in forced-convection cooling by uniquely

considering both the heat flux and flow resistance. The modeling

of forced convection cooling was also reported by Baelmans et al.

(2003), whoobserved that fan-induced swirling flows are difficult

to predict using standard k-1 turbulence models. Cole et al.

(2003) investigated the aerodynamic and thermal interactions of

ball grid array packages in a real environment. Furthermore,

extensive experiments and modeling of flow and heat transfer in

fan-cooled electronic systems were reported by Grimes et al.

(2001) and Grimes and Davies (2002, 2004a, b). Based on

observation, the flow in the sucking mode is steady, uniform, and

The current issue and full text archive of this journal is available at

www.emeraldinsight.com/1356-5362.htm

Microelectronics International

30/3 (2013) 138–150

q Emerald Group Publishing Limited [ISSN 1356-5362]

[DOI 10.1108/MI-10-2012-0071]

The authors would like to thank Universiti Sains Malaysia (USM),Penang, Malaysia, for the financial support for this research work underthe USM Fellowship scheme.

138

easily predictable using a laminar model, whereas in blowing,

it is unsteady, swirling, and too complex to be predicted by aturbulent model with reasonable accuracy. However, the heat

transfer is enhanced by shifting the fan from exit (sucking) to theinlet (blowing).

For rigid PCB, Leicht and Skipor (Leicht and Skipor, 2000;Skipor and Leicht, 2001) introduced a method to measure the

reliability of mechanical bending fatigue of area array packages.A global/local modeling in PCB mechanical loading wasproposed by Zhu et al. (2001). Using the finite element method

(FEM) model, Shetty et al. (2001) and Shetty and Reinikainen(2003) investigated the durability of CSP interconnects when

assembled on FR4 substrates. Lau et al. (2006) studied plasticballgrid array package assemblies under a three-point bending

condition by experiment and simulation. In addition, Yu et al.(2010) analyzed the full-field dynamic responses of the PCBs at

the product level using digital image correlation technique.Recently, using FEM, Arruda et al. (2009a, b) have studied aflex-rigid PCB (RFPCB) interface to evaluate the cracking

phenomena during thermal cycling. They found that thedevelopment of stress is due to the design of the PCB, and the

non-uniformity is due to the thermal expansion of various parts.Sun et al. (2009) also found, through FE analysis, that there are

some risk places for potential failures, such as the corners of theconnecting part of the RFPCB. Das et al. (2010) were able to

fabricate a biocompatible FPCB on a polydimethylsiloxanesubstrate.

Only a few researchers have focused on the material and

structural characteristics of FPCB. Li and Jiao (2000) studiedthe effects of temperature and aging on the Young’s moduli of

polymeric-based flexible substrates, such as polyethylenenaphthalate, polyester, and polyimide. Experimentally,

Martynenko et al. (2002) further assessed the fatigueresistance and reliability of polyimide-based FPCB.

Barlow et al. (2002) successfully demonstrated the feasibilityand viability of FPCB for miniaturization purpose on integratedpower modules. In addition, Van Driel et al. (2006) predicted

the reliability problem for packages in FPCB end products. Thefatigue behavior of thin Cu foil for FPCB was experimentally

studied by Han et al. (2007). Considering fluid-structureinteraction (FSI), Arruda and Freitas (2007) proposed a

multiphysics model to describe the interaction between FPCBand its surrounding air. Huang et al. (2009) investigated themechanical reliability of a ball grid array mounted on an FPCB

under drop impact and found that stress could be reduced byincreasing solder joints height. In the work of Rizvi et al. (2010),

Ansys v.11 was used to predict damage accumulations in chipresistor solder joints under a range of thermal cycling

conditions. Interestingly, Siegel et al. (2010) were able tofabricate several low-cost flexible electronic circuits on paper

substrates. Suk et al. (2012) also studied various chip-on-filmpackages using two-metal layer FPCB.

Only a few studies on PC motherboards have beeninvestigated. Using CFD modeling, Chang et al. (2001)identified the minimum air flow design for a PC. Pitarresi et al.(2002, 2004) and Yating et al. (2006) addressed the FEM issueon the dynamic responses of a typical PC motherboard. They

locally modeled the large components as simple blocks to avoidexpensive modeling and found that the predictions correlate

well with the experimental measurements. Darveaux et al.(2009) investigated the temperature distribution in a laptopcomputer under various conditions and observed that the

operating temperature range is approximately 558C-808C.

Tari and Yalcin (2010) suggested that placing the CPU,motherboard, and RAM on the back side of the laptop lid couldcool the laptop passively without the aid of a fan. Recently, theauthors (Leong et al., 2012a, 2012b, 2012c) investigated theFSI issues in relevant with the applications of the FPCB as wellas the FPCB motherboard.

The authors’ recent work (Leong et al., 2012b) has studiedthe few physical design issues on the novel approach ofFPCB motherboard. In the study, the deflection and stressconsiderations were found to bevital for the FPCB motherboard.However, the outcomes inside different PC casings are yet to beexplored, which is particularly important for the FPCBmotherboard. Therefore, different from previous studies, thepresent study extends the study to focus on the FSI of the newlyproposed FPCB motherboard in different configurations of PCcasings. Few cases are investigated such as effects of inletlocations, inlet sizes, multi-inlets and two-fan system. These FSIstudies are performed using the fluid flow solver FLUENTandthe structural solver ABAQUS; they are real-time online coupledby mesh-based parallel code coupling interface (MpCCI).

2. Methodology

2.1 Modeling strategy

The present study utilizes the similar FPCB motherboard testvehicle and also similar numerical technique with the authors’recent work (Leong et al., 2012b). The FSI numericalanalyses are performed using the fluid flow solver FLUENT6.3.26 and structural solver ABAQUS/CAE 6.9; they areonline real-time coupled by MpCCI 3.1.0. In the presentstudy, the analysis consumed approximately 30 h per case on acomputer machine with Pentium (R) Dual-Core processorcomputer machine with E6800, 3.33 GHz and 3 GB of RAM.

2.2 Structural modeling

In ABAQUS, as shown in Figure 1, the components on theFPCB motherboard are modeled as idealized simple blocksrigidly attached to the motherboard. This simplified modelingtechnique has been proven reliable, as demonstrated in theprevious works of Pitarresi et al. (2002, 2004) and authors’recent work (Leong et al., 2012b). In the authors’ recent work(Leong et al., 2012b), the FPCB was tested according to ASTMD638-03 (2003), which is a standard test method for the tensile

Figure 1 Modeling of FPCB motherboard

Study on the fluid-structure interaction of FPCB motherboard

Wei Chiat Leong et al.


Volume 30 · Number 3 · 2013 · 138–150

139

properties of plastics. The FPCB was purchased from PCB

Universe located in the USA, and the effective elastic modulus

was found to be 5.243 GPa in the linear elastic region. In the

present study, the effective elastic modulus, density, and

dimensions of various components are specified in Table I

(Leong et al., 2012b). As shown in Figure 2, the fully constraint

boundary condition (Ux ¼ Uy ¼ Uz ¼ URx ¼ URy ¼ URz ¼ 0)

is assigned to the fixed regions of the motherboard, whereU and

UR denote linear displacement and rotational displacements,

respectively. This structural model is meshed with 28,353

hexagonal elements, as shown in Figure 3.

2.3 Fluid flow modeling

In reality, although there are many different ways of placing

electronic devices in PC casings, the present study considers a

simplified and general type of PC casing, as shown in Figure 4.

As this model involves a wide range of length scale, the details

of the complex structures in the PC casing are eliminated

to minimize the modeling cost and time consumption

concerns. The outlet fan provides an airflow rate of

0.04594 kg/s, which is the typical airflow rate in the

application (Grimes et al., 2001; Grimes and Davies, 2002,

2004a, b). The CD drives/HDD and power supply device are

treated as walls and are not included in the mesh domain for

efficient analysis.In FLUENT, the governing equations describing the fluid

flow are continuity and momentum equations, as given in

equations (1) and (2), respectively:

›rf

›tþ 7 · rf

!uf

� �¼ 0 ð1Þ

where rf is the density of fluid,!uf is the overall velocity vector

of fluid and t is the time:

rf›!uf

›tþ !uf ·7

!uf

� �¼ 27P þ 7 · tþ rf ~gþ ~F ð2Þ

where the P is the static pressure, t is the stress tensor, ~g is

the gravitational acceleration and ~F is the external body

forces.In the analyses, the flow is considered 3D, laminar,

incompressible, and unsteady. The laminar model is suitable

for a fan-sucking system, which was reported by Grimes et al.(2001), Grimes and Davies, 2002, 2004a, b) and authors’

recent work (Leong et al., 2012b). The fluid domain is

meshed with 316,167 hexagonal grids, as shown in Figure 5.

The fluid domain is split into several smaller volumes

appropriately to allow for the assignment of a structured

hexagonal mesh. The boundary conditions used are as

follows:. On motherboard and enclosure wall: ~u ¼ 0.. At the inlet: ~u ¼ ~udesired .. At the outlet: P ¼ 0.

2.4 Case studies

The validation of the present numerical technique had been

demonstrated in the authors’ recent work (Leong et al.,2012b). In the present study, as shown in Figure 6, the inlet is

manipulated at different locations, such as inlets A, B, C, D,

and E, to examine the inlet location effect on the behavior of

the motherboard. To investigate the effect of inlet sizes, the

size of each inlet location is reduced according to Table II.

These inlets are named according to their locations and sizes;

for instance, inlet size 1 at inlet A is denoted as inlet A1,

whereas inlet size 3 at inlet D is denoted as inlet D3.

Sometimes, designers prefer PC casings with a huge inlet A

associated with a few smaller inlets B, C, D, or E. With this

idea in mind, several common combinations of the multi-

inlets are also studied, as shown in Figure 7. Lastly, the effect

of the PC casing with two identical fans (which provide a total

airflow rate of 0.09188 kg/s) is also explored; an example of an

inlet A1 with two fans is shown in Figure 8.

3. Results and discussions

3.1 Effect of inlet locations

An example of the magnified deformation of the FPCB

motherboard is shown in Figure 9. The z-direction

displacement is greatly dominating. Hence, in this paper,

the deflection term is considered the displacement in the

z-direction.Figure 10 shows the deflection and stress contours for inlet

A1 at different times. The magnitudes of the contours slightly

change as the time progresses, but the contours still share the

same distributions. Similarly, as shown in Figure 11, different

inlet cases also result in almost identical deflection and

stress distributions, even though their magnitudes are

different because the physical layout of the components and

the fastening spots of the motherboards remain the same

for the entire study. The critical deflection and critical stress

sites arise from the regions, as shown in Figure 12.

Therefore, these critical sites are appropriate as indicators to

characterize the behavior of the FPCB motherboard in the PC

casings.

Figure 2 Constraint locations for the FPCB motherboard

Red circles representfixed regions

Table I Material properties

Regions

Size

(x mm 3 y mm 3 z mm)

Effective modulus

(Pa)

Density

(kg/m3)

FPCB 190 £ 204 £ 0.1 5.243 £ 109 3,636.4

CPU fan 50 £ 52 £ 35 1.130 £ 109 791.2

Heat sink 38 £ 38 £ 15 1.130 £ 109 1,246.5

Memory 16 £ 141 £ 6 3.130 £ 109 1,034.3

PCI slots 85 £ 9 £ 15 8.460 £ 108 958.6

Pink connector 25 £ 66 £ 30 1.130 £ 109 888.9

USB connector 25 £ 34 £ 30 1.130 £ 109 1,294.1

Yellow connector 25 £ 46 £ 30 1.130 £ 109 956.5

Source: Leong et al. (2012b)




Volume 30 · Number 3 · 2013 · 138–150

140

Figure 3 3D meshed model in ABAQUS

Y

X

Z

Figure 4 Modeling of PC casing environment

Figure 5 3D meshed model in FLUENT

Y

XZ




Volume 30 · Number 3 · 2013 · 138–150

141

In the no-flow condition (still air), purely because of the

weight-induced force, the motherboard critical deflection

is obtained at 0.239 mm, and its accompanied critical stress is

1.591 MPa. Figure 13 shows the deflections and stresses at

the respective critical sites in the motherboard for the

different inlet locations. Inlets B1 and C1 significantly induce

deflections and stresses on the motherboard because the flows

entering from the side of the PC casings contain prominent

z-direction velocity, which can easily affect the motherboard.Conversely, for inlets D1 and E1, which also introduceflows from the side of the PC casings, the flow influences arerather small because a portion of the flow directly bypassestowards the near exits. Although the present study is onlyfocused on the FSI aspect, air bypass is generally not desirableunder thermal consideration. Among them, inlet A1 isconsidered moderate in terms of deflection and stressinduced.

3.2 Effect of inlet sizes

The effect of inlet sizes is also studied for each inlets A, B, C, D,and E. The deflections and stresses induced at respective criticalsites in the motherboard for the different inlet sizes at inlet Aare shown in Figure 14. As the inlet size is reduced,the motherboard deflection and stress fluctuationsremarkably increase, increasing tremendously when inletA5 is used. As shown in Figure 15, as the inlet size becomessmaller, the entering airflow velocity increases, subsequentlyinitiating the vigorous flow. Generally, these fluctuations

Figure 6 PC casings with different inlet locations

Fan outlet

Flow inlet

y

xz

Fan outlet Fan outlet

Fan outletFan outlet

y

xz

y

xz

y

xz

y

xz

Flow inletFlow inlet

Flow inletFlow inlet

Inlet D Inlet E

Inlet B

Inlet A

Inlet C

Table II Dimensions of different inlet sizes

Inlet sizes

Horizontal length 3 vertical

length (mm 3 mm)

1 165 £ 130

2 145 £ 110

3 125 £ 90

4 105 £ 70

5 85 £ 50




Volume 30 · Number 3 · 2013 · 138–150

142

are extremely undesirable under the fatigue concern

because they can lead to reliability failure.To explore further in detail the stress fluctuations, which are

simply the differences between the maximum and minimum

values, all inlets are analyzed and plotted in Figure 16. The

responses of the other inlets are also identicalwith thoseof inlet A,

in which a smaller inlet size gives higher stress fluctuation on

the motherboard. Interestingly, the stress fluctuations are

exponentially increased with the reduction of the inlet area. The

stress fluctuations slightly increase when the inlet area decreases

from 214.5 cm2 at the beginning. However, the fluctuation

increments are significantly boosted when the inlet area is further

reduced from 112.5 cm2. Moreover, inlets D3 and E3 are less

affected by the inlet size effect because a large portion of the flow

directly bypasses towards the exits, in agreement with the finding

in Section 3.1.

3.3 Effect of multi-inlets

As a result of the problem on the high fluctuation

phenomenon, which is attributed to the small inlet size as

discussed in Section 3.2, there is a need to investigate the

behavior of a motherboard in multi-inlet PC casings with

Figure 7 PC casings with different multi-inlets

Inlet A1+B3 Inlet A1+C3

Inlet A1+B3+C3 Inlet A1+B3+E3

Flow inlets Flow inlets

y

xz

y

xz



y

xz

y

xz

Flow inlets Flow inlets

Figure 8 PC casing with two-fan outlets

Two fanoutlets

Flow inlet

y

xz

Figure 9 Magnified deformation scale of the FPCB motherboard

Significant z-directiondisplacement

Y

ZX




Volume 30 · Number 3 · 2013 · 138–150

143

few small inlets because such PC casings are sometimes

preferred. Figure 17 shows the resultant deflections and

stresses induced at the respective critical sites in the

motherboard for different multi-inlets. Significantly, the

fluctuation phenomenon greatly diminishes when the small

inlet is used in conjunction with other inlets. The reason is

that the entering airflows are now contributed by several

inlets; hence, more inlet area is accessible for the PC

casings. Consequently, the concentrated and vigorous flows at

one small inlet can be prevented. Owing to the samereason, the magnitudes of deflection and stress are also

noticeably reduced when the PC casings switch from a two-

inlet design to a three-inlet design, as clearly shown in the

figure.

3.4 Effect of the two-fan system

Occasionally, the PC casing is installed with two fans to

promote airflow. The values of the deflections and stresses for

different inlets are shown in Figure 18. For inlets A1, B1, and

C1, when the airflow rates are doubled, the magnitudes of the

deflection and stress are unavoidably enlarged, with an

average of 36 and 52 percent, respectively. Despite this

condition, a negligible flow influence exists at inlet E1, which

is again due to the bypass of the flow towards the exit.Surprisingly, inlet D1 is affected to a great extent even though

it is located near the exit. The reason is that the two-fan

system provides a very energetic flow that permits the entering

airflow to travel further and impinge on the motherboard

towards the negative z-direction instead of bypass towards the

exit. This situation also causes the unexpected reductions in

the deflection (260 percent) and stress (235 percent) for

inlet D1.

3.5 Discussion on the FPCB motherboard

Unlike the traditional rigid motherboard, the FSI study is far

more vital for the newly proposed FPCB motherboard under

a fan-flow condition. Based on the studies, the maximum

deflection can reach 0.61 mm, whereas the maximum stress

can be attained at 3.092 MPa. At this juncture, the deflection

has to be taken with utmost care to ensure that the deformed

motherboard does not impose any physical contact among the

components as well as with the other surrounding objects,

especially in miniaturized condition. The existence of the

stress fluctuation is also crucial and undesirable, as extreme

fatigue can lead to the failure of the motherboard. In practice,

the reductions in deflection and stress are always favorable for

enhanced reliability; this issue is particularly important for the

FPCB-based motherboard.Based on the results, the PC casing with inlet A1 is the most

promising for the FPCB motherboard because it offers

moderate magnitudes of deflection and stress induced, and it

does not directly bypass the air. Considering the FSI concern,

the designer is recommended to select a huge inlet and avoid an

inlet size smaller than 112.5 cm2to prevent excessive deflection

and stress fluctuations, which can greatly reduce the reliability

of the motherboard due to fatigue. However, if the designer still

insists on selecting a small inlet, the small inlet should be used in

conjunction with other inlets because more inlets can effectively

reduce the fluctuation phenomenon in the FPCB motherboard.

Although using two fans can greatly increase the airflow rate, the

designer has to be aware and make proper compromises because

the enlargements in deflection (36 percent) and stress

(52 percent) induced are critical for inlets A1, B1, and C1.

However, this issue is not a concern if inlets D1 and E1 are used.

Figure 10 Deflection and stress contours for inlet A1 at different times

1 s 1 s

5 s 5 s

10 s 10 s




Volume 30 · Number 3 · 2013 · 138–150

144

In terms of fan-flow influence, the authors believe that theapplication of the newly proposed FPCB motherboard is feasibleprovided that the deflection and stress issues are carefullyconsidered during the design stage. However, more detailedanalyses on the other influential factors of the mechanical andelectronic aspects have to be carried out before the FPCBmotherboard can finally be used commercially.

4. Conclusions

The present study has explored the behaviors of the newlyproposed FPCB motherboard under a fan-flow condition inthe PC casings. The studies are numerically performedusing FLUENT and ABAQUS; they are real-time onlinecoupled by MpCCI. Inlet A1 is the most promising for theFPCB motherboard system considering the FSI concern.Interestingly, as the inlet size is reduced, the deflection andstress fluctuations of the motherboard are amplified, and the

effect is drastically boosted when the inlet area is further

reduced from 112.5 cm2. However, when the small inlet is

used in conjunction with other inlets, the undesirable

fluctuations are greatly diminished. For inlets A1, B1, and

C1, when the airflow rate is doubled, the magnitudes of

deflection and stress are unavoidably enlarged at around

36 and 52 percent, respectively. However, this issue is not a

concern for inlets D1 and E1. The application of the FPCB

motherboard is considered feasible under the fan-flow

condition. However, more detailed analyses of other

influential factors still have to be carried out. The present

study provides a better understanding and allows designers to

be aware of the FSI phenomenon, not only when dealing with

FPCB motherboard but also other FPCB applications. In the

future, the present study can be further extended to include

thermal stress, electrical circuit and electronic signal

concerns.

Figure 11 Deflection and stress contours for different inlet locations at 10 s

Inlet A1 Inlet A1

Inlet B1 Inlet B1

Inlet C1 Inlet C1

Inlet D1 Inlet D1

Inlet E1 Inlet E1




Volume 30 · Number 3 · 2013 · 138–150

145

Figure 12 Critical deflection and critical stress sites

Figure 13 (a) Critical deflections and (b) critical stresses for differentinlet locations

Time (s)

0 5 10 15 20

Def

lect

ion

(m

m)

0.0

0.1

0.2

0.3

0.4

0.5

Inlet A1Inlet B1Inlet C1Inlet D1Inlet E1

(a)

Time (s)

0 5 10 15 20

von

Mis

es s

tres

s (M

Pa)

0.0

0.5

1.0

1.5

2.0

2.5

Inlet A1Inlet B1Inlet C1Inlet D1Inlet E1

(b)

Figure 14 (a) Critical deflections and (b) critical stresses for differentinlet sizes at inlet A

Time (s)

0 5 10 15 20

Def

lect

ion

(m

m)

–0.1

0.0

0.1

0.2

0.3

0.4

0.5

0.6Inlet A1Inlet A2Inlet A3Inlet A4Inlet A5

(a)

Time (s)

0 5 10 15 20

von

Mis

es s

tres

s (M

Pa)

0.0

0.5

1.5

1.0

2.0

2.5

3.0

Inlet A1Inlet A2Inlet A3Inlet A4Inlet A5

(b)




Volume 30 · Number 3 · 2013 · 138–150

146

Figure 15 Flows for different inlet sizes at inlet A

Inlet A1

Inlet A3

Inlet A5




Volume 30 · Number 3 · 2013 · 138–150

147

References

Arruda, L. and Freitas, G. (2007), “Effect of surrounding air

on board level drop tests of flexible printed circuit boards”,

EuroSime 2007: International Conference on Thermal,Mechanical and Multi-Physics Simulation Experiments inMicroelectronics and Micro-Systems, pp. 1-4.

Arruda, L., Chen, Q. and Quintero, J. (2009a), “Failure

evaluation of flexible-rigid PCBs by thermo-mechanical

simulation”, International Conference on Electronic PackagingTechnology & High Density Packaging, pp. 1201-1205.

Arruda, L., Bonadiman, R., Costa, F. and Reinikainen, T.

(2009b), “Cracking phenomena on flexible-rigid interfaces

in PCBs under thermo cycling load”, Circuit World, Vol. 35

No. 2, pp. 28-32.ASTM D638-03 (2003), Standard Test Method for TensileProperties of Plastics, ASTM, Philadelphia, PA.

Azar, K. and Russell, E.T. (1990), “Effect of component layout

and geometry on the flow distribution in electronics circuit

packs”, Sixth IEEE SEMI-THERMe Symposium, pp. 1-9.Baelmans, M., Meyers, J. and Nevelsteen, K. (2003), “Flow

modeling in air-cooled electronic enclosures”, 19th IEEESEMI-THERM Symposium, pp. 27-34.

Figure 16 Stress fluctuations for different inlet sizes

Inlet area (cm2)

20 40 60 80 100 120 140 160 180 200 220 240

Str

ess

flu

ctu

atio

n (

MP

a)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

Inlet AInlet BInlet CInlet DInlet E

Size 3

Size 2 Size 1

Size 5

Size 4

Figure 17 (a) Critical deflections and (b) critical stresses for differentmulti-inlets

Time (s)0 5 10 15 20

Def

lect

ion

(m

m)

0.1

0.2

0.3

0.4

0.5Inlet A1+B3Inlet A1+C3Inlet A1+B3+C3Inlet A1+B3+E3

(a)

Time (s)0 5 10 15 20

von

Mis

es s

tres

s (M

Pa)

1.0

1.5

2.0

2.5Inlet A1+B3Inlet A1+C3Inlet A1+B3+C3Inlet A1+B3+E3

(b)

Figure 18 (a) Critical deflections and (b) critical stresses for single-fanand two-fan systems

Inlets

A1 B1 C1 D1 E1

Def

lect

ion

(m

m)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Single fanTwo fans

(a)

Inlets

A1 B1 C1 D1 E1

(b)

von

Mis

es s

tres

s (M

Pa)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Single fanTwo fans




Volume 30 · Number 3 · 2013 · 138–150

148

Barlow, F., Lostetter, A. and Elshabini, A. (2002), “Low cost

flex substrates for miniaturized electronic assemblies”,

Microelectronics Reliability, Vol. 42, pp. 1091-1099.Chang, J.Y., Yu, C.W. and Webb, R.L. (2001), “Identification

of minimum air flow design for a desktop computer using

CFD modeling”, Journal of Electronic Packaging, Vol. 123,

pp. 225-231.Cole, R., Davies, M. and Punch, J. (2003), “A board level

study of an array of ball grid components – Aerodynamic

and thermal measurements”, Journal of Electronic Packaging,

Vol. 125, pp. 480-489.Darveaux, R., Reichman, C. and Berry, C.J. (2009), “Field

use environment for a FCBGA in a laptop computer

application”, Electronic Components and Technology

Conference, pp. 1234-1239.Das, R.N., Egitto, F.D., Wilson, B., Poliks, M.D. and

Markovich, V.R. (2010), “Development of rigid-flex and

multilayer flex for electronic packaging”, Electronic

Components and Technology Conference, pp. 568-574.Eveloy, V., Lohan, J. and Rodgers, P. (2000), “A benchmark

study of computational fluid dynamics predictive accuracy

for component-printed circuit board heat transfer”, IEEE

Transactions on Components and Packaging Technologies,

Vol. 23 No. 3, pp. 568-577.Grimes, R. and Davies, M. (2002), “The effect of fan

operating point and location on temperature distribution in

electronic systems”, Inter Society Conference on Thermal

Phenomena, pp. 677-684.Grimes, R. and Davies, M. (2004a), “Air flow and heat

transfer in fan cooled electronic systems”, Journal of

Electronic Packaging, Vol. 126, pp. 124-134.Grimes, R. and Davies, M. (2004b), “Optical measurement

of electronic system air flow and temperature distribution”,

Journal of Optics A: Pure and Applied Optics, Vol. 6,

pp. 617-626.Grimes, R., Davies, M., Punch, J., Dalton, T. and Cole, R.

(2001), “Modeling electronic cooling axial fan flows”,

Journal of Electronic Packaging, Vol. 123, pp. 112-119.Han, S., Seo, K., Kim, W., Lee, H., Lee, H., Shin, J. and

Lee, J. (2007), “Fatigue behavior of thin Cu foils for flexible

printed circuit board”, Solid State Phenomena, Vol. 124-126,

pp. 1369-1372.Huang, J., Chen, Q., Xu, L. and Zhang, G.Q. (2009),

“Prediction of drop impact reliability of BGA solder joints

on FPC”, 11th Electronics Packaging Technology Conference,

pp. 663-667.Lau, D., Chan, Y.S., Ricky Lee, S.W., Fu, L., Ye, Y. and Liu, S.

(2006), “Experimental testing and failure prediction

of PBGA package assemblies under 3-point bending

condition through computational stress analysis”, 7th

International Conference on Electronics Packaging Technology,

pp. 1-7.Lee, T. and Mahalingam, M. (1993), “Application of a CFD

tool for system level thermal simulation”, Proceedings of

IEEE 43rd Electronic Components and Technology Conference,

pp. 249-255.Leicht, L. and Skipor, A. (2000), “Mechanical cycling fatigue

of PBGA package interconnects”, Microelectronics Reliability,

Vol. 40, pp. 1129-1133.Leon, O., Mey, G.D. and Dick, E. (2002), “Study of the

optimal layout of cooling fins in forced convection cooling”,

Microelectronics Reliability, Vol. 42, pp. 1101-1111.

Leong, W.C., Abdullah, M.Z. and Abdul Mujeebu, M.

(2012a), “Flow induced deflection and stress on flexible

printed circuit board (FPCB) in fan-cooled electronic

systems: FSI approach”, IEEE Transactions on Components,

Packaging and Manufacturing Technology, Vol. 2 No. 4,

pp. 617-624.Leong, W.C., Abdullah, M.Z. and Khor, C.Y. (2012b),

“Application of flexible printed circuit board (FPCB) in

personal computer motherboards: focusing on mechanical

performance”, Microelectronics Reliability, Vol. 52,

pp. 744-756.Leong, W.C., Abdullah, M.Z., Khor, C.Y. and

Ernest Ong, E.S. (2012c), “Study on the fluid-structure

interaction of flexible printed circuit board electronics in

the flow environment”, IEEE Transactions on Components,

Packaging and Manufacturing Technology, Vol. 2 No. 8,

pp. 1335-1345.Leung, C.W., Chen, S. and Chan, T.L. (2000), “Numerical

simulation of laminar forced convection in an air-cooled

horizontal printed circuit board assembly”, Numerical Heat

Transfer, Part A, Vol. 37, pp. 373-393.Li, R.S. and Jiao, J. (2000), “The effects of temperature and

aging on Young’s moduli of polymeric based flexible

substrate”, The International Journal of Microcircuits and

Electronic Packaging, Vol. 23 No. 4, pp. 456-461.Lohan, J., Eveloy, V. and Rodgers, P. (2002), “Visualization of

forced air flows over a populated printed circuit board and

their impact in convective heat transfer”, Inter Society

Conference on Thermal Phenomena, pp. 501-511.Martynenko, E., Zhou, W., Chudnovsky, A., Li, R.S. and

Poglitsch, L. (2002), “High cycle fatigue resistance and

reliability assessment of flexible printed circuitry”,

Transactions of the ASME, Vol. 124, pp. 254-259.Pitarresi, J., Geng, P., Beltman, W. and Ling, Y. (2002),

“Dynamic modeling and measurement of personal

computer motherboards”, Electronic Components and

Technology Conference, pp. 597-603.Pitarresi, J., Roggerman, B., Chaparala, S. and Geng, P.

(2004), “Mechanical shock testing and modeling of PC

motherboards”, Electronic Components and Technology

Conference, pp. 1047-1054.Rizvi, M.J., Bailey, C. and Lu, H. (2010), “Damage

predictions in a chip resistor solder joint on flexible

circuit board”, Microelectronic Engineering, Vol. 87,

pp. 1889-1895.Rodgers, P.J., Eveloy, V.C. and Davies, M.R.D. (2003a),

“An experimental assessment of numerical predictive

accuracy for electronic component heat transfer in forced

convection – part I: experimental methods and numerical

modeling”, Journal of Electronic Packaging, Vol. 125,

pp. 67-75.Rodgers, P.J., Eveloy, V.C. and Davies, M.R.D. (2003b),

“An experimental assessment of numerical predictive

accuracy for electronic component heat transfer in forced

convection – part II: results and discussion”, Journal of

Electronic Packaging, Vol. 125, pp. 76-83.Shankaran, G.V. and Karimanal, K. (2002), “A multi level

modeling approach with boundary condition import for

system, sub-system or board level thermal

characterization”, 18th IEEE SEMI-THERM Symposium,

pp. 35-41.




Volume 30 · Number 3 · 2013 · 138–150

149

Shetty, S. and Reinikainen, T. (2003), “Three- and four-pointbend testing for electronic packages”, Journal of ElectronicPackaging, Vol. 125, pp. 556-561.

Shetty, S., Lehtinen, V., Dasgupta, A., Halkola, V. andReinikainen, T. (2001), “Fatigue of chip scale packageinterconnects due to cyclic bending”, Journal of ElectronicPackaging, Vol. 123, pp. 302-308.

Siegel, A.C., Phillips, S.T., Dickey, M.D., Lu, N., Suo, Z. andWhitesides, G.M. (2010), “Foldable printed circuit boardson paper substrates”, Advanced Functional Materials,Vol. 20, pp. 28-35.

Skipor, A. and Leicht, L. (2001), “Mechanical bendingfatigue reliability and its application to area arraypackaging”, Electronic Components and TechnologyConference, pp. 606-612.

Suk, K.-L., Choo, K., Kim, S.J., Kim, J.-S. and Paik, K.-W.(2012), “Studies on various chip-on-film (COF) packagesusing ultra fine pitch two-metal layer flexible printedcircuits (two-metal layer FPCs)”, Microelectronics Reliability,Vol. 52 No. 6, pp. 1182-1188.

Sun, H.-b., Shi, M.-Q., Yang, P., Zhong, J., Chen, Q., Xue, T.,Zhao, T., Xu, L. and Salo, A. (2009), “Rigid-flexible printedcircuit structure optimization by simulations”, 10thInternational Conference on Thermal, Mechanical andMultiphysics Simulation Experiment in Micro-Electronics andMicro-Systems, pp. 1-4.

Tari, I. and Yalcin, F.S. (2010), “CFD analyses of a notebookcomputer thermal management system and a proposedpassive cooling alternative”, IEEE Transactions onComponents and Packaging Technologies, Vol. 33 No. 2,pp. 443-452.

Van Driel, W.D., Van der Sluis, O., Yang, D.G., Ubachs, R.,Zenz, C., Aflenzer, G. and Chang, G.Q. (2006),“Reliability modeling for packages in packages in flexibleend-products”, Microelectronics Reliability, Vol. 46,pp. 1880-1885.

Yating, Y., Pingan, D. and Ezhong, G. (2006), “Structuralrobustness study of a computer motherboard”, Proceedingsof the IEEE International Conference on Industrial Technology,pp. 2370-2374.

Yu, D., Kwak, J.B., Park, S. and Lee, J. (2010), “Dynamicresponses of PCB under product-level free drop impact”,Microelectronics Reliability, Vol. 50, pp. 1028-1038.

Zhu, J., Quander, S. and Reinikainen, T. (2001), “Global/localmodeling for PWB mechanical loading”, ElectronicComponents and Technology Conference, pp. 1164-1169.

About the authors

Wei Chiat Leong obtained his Bachelor’s degree in

Mechanical Engineering (First Honours) from Universiti

Sains Malaysia (USM), Malaysia. Currently he is pursuing his

doctoral degree in the field of electronic packaging at USM,

specializing in the area of flexible printed circuit board. He is

an active member of Institution of Engineers Malaysia (IEM),

registering under the Board of Engineers Malaysia (BEM).

His research interests are fluid mechanics, CFD, structural

analysis, FEM, fluid-structure interaction study and

electronic packaging. Wei Chiat Leong is the corresponding

author and can be contacted at: [email protected]

Mohd Zulkifly Abdullah is Professor of Mechanical

Engineering at Universiti Sains Malaysia. He obtained his

Bachelor degree in Mechanical Engineering from University

of Wales, Swansea, UK. His MSc and PhD degrees are from

University of Strathclyde, UK. He has numerous publications

in international journals and conference proceedings. His

areas of research are CFD, heat transfer, and electronic

packaging.

Chu Yee Khor received the Graduate degree in mechanical

engineering and the Master’s degree in computational fluid

dynamics, specializing in integrated circuit packaging from

Universiti Sains Malaysia, Minden, Malaysia, in 2008 and

2010, respectively. He is currently pursuing the PhD degree in

advanced packaging in microelectronics. His current research

interests are electronics packaging, fluid/structure interaction,

fluid mechanics, dynamics, polymer rheology and heat and

mass transfer.

Dadan Ramdan was born in Bandung, Indonesia, in 1964.

He received the BS degree in physics from the Pandjadjaran

University of Bandung, Bandung, and the MSc and MEng

degrees from the Institute Technology of Bandung, Bandung,

and the Toyohashi University of Technology, Toyohashi,

Japan, respectively. He is currently pursuing the PhD degree

with Universiti Sains, Penang, Malaysia. He was previously a

lecturer of mechanical engineering with the Engineering

Faculty, Medan Area University, Medan, Indonesia. His

current research interests include electronic instrumentation,

control systems engineering, and computational flow

dynamics analysis of electronic packaging and casting.




Volume 30 · Number 3 · 2013 · 138–150

150

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints

Study on the Fluid\u0026#x2013;Structure Interaction of Flexible Printed Circuit Board Electronics in the Flow Environment

Documents