Proceedings of the International Conference on Vibration, Sound and System Dynamics Penang, 2 August 2017 21 Structural Analysis on Stretchable Electronics Joints of Automotive LED Lighting under Vibration Loading M.F.M. Sharif A.A. Saad N. Mustafa School of Mechanical Engineering, Universiti Sains Malaysia, Penang, MALAYSIA F.C. Ani M.Y.T. Ali Jabil Circuit Sdn Bhd, Bayan Lepas Industrial Park, 11900 Penang, MALAYSIA The purpose of this paper is to investigate the effect of vibration loading on structural integrity of LED joints strength in manufacturing of automotive lighting. Ten LEDs were separately assembled on printed electronic circuit by dispensing a conductive adhesive. There were five different build matrices constructed for the conductive adhesive and each matrix had two samples. A platform was fabricated and mounted to the shaker to run the samples altogether in vertical and horizontal positions. The samples were run under random vibration in frequency range of 3 Hz to 500 Hz for 4 hours. The assembly of LED joints was further analyzed using finite element simulation to understand their stress-strain behavior under vibration loading. The quality and integrity of LED joints were investigated using microscopy observation and x-ray analysis at the LED joints. The results show there is no failure generated from the vibration test of applied period. The experimental result was supported by finite element analysis on detail stress-strain behavior during the test. Keywords: LED joints; Stretchable electronics; Vibration fatigue. 1 INTRODUCTION Random vibration is the most related vibration condition subjected in the environment conditions due to destabilization which created by rotational and tumbling masses. Automobiles and portable electronic components are mostly exposed to random vibration. A system can vibrate in a standard manner that is known as vibration modes representing the degree of freedom of the vibration in the system with certain natural frequency. Range of frequency varied according to its application, it had been tested that 2.5 Ton truck provide acceleration level 15 G and 19 G with frequency 15 Hz – 40 Hz at speed of 10 mph - 15 mph (Steinberg & Wiley, 2001). In the automobiles vibration frequency range is varied with the system, vibration while driving, suspension system, and steering system, which the vibration range from 3 Hz – 1 KHz (Izumi, n.d.). A range of frequency is excited at the same time, which means that frequency loading correspond to all masses are represented at the same time which produce large displacement amplitude to produce impact of failure to analysis system. To quantify random frequency, Power Spectral Density (PSD) input is needed. PSD is the input curve of acceleration units per Hz² versus range of frequencies. PSD curve input is translated to the Root Mean Square (RMS) value which the value of each curve is formulated to a single quantity. As the random vibration is the nature response statistic it is represented in zero mean distribution where the result is presented in three standard deviations of RMS value in the presented value of stress, strain, acceleration and velocity (Kumar & Engineer, 2008). Several types of interconnection materials have been used in electronic packaging such as a lead-based solder, a lead-free solder and a conductive adhesive. The lead-based solder was the main interconnection material in electronics industry as it is requires low temperature and less issues regarding to its quality. However due to restriction in law and new requirements for non-toxic materials, lead free solder and conductive adhesive are used as alternative (Li & Wong, 2006). Lead free solder is a tin based alloy, common composition of lead free solder used are Sn- Ag, Sn-Cu, Sn-Zn and Sn-Ag-Cu. Conductive adhesives are a composite of adhesive epoxy and a conductive metal such as silver, silver, tin oxide or indium. Concept of the composition is that soft material epoxy filler promote good contact within and other particles as it deform and shrink during curing process. Most common material for the epoxy filler is silver as it is a good conductor, commercially available and has moderate cost. Conductive adhesive is applied as the electronics interconnections by using
8
Embed
Structural Analysis on Stretchable Electronics Joints of ...thevibrationlab.eng.usm.my/images/ICVSSD/FullPaper/... · solder, a lead-free solder and a conductive adhesive. The lead-based
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Proceedings of the International Conference on Vibration, Sound and System Dynamics Penang, 2 August 2017
21
Structural Analysis on Stretchable Electronics Joints of Automotive LED
Lighting under Vibration Loading
M.F.M. Sharif
A.A. Saad
N. Mustafa School of Mechanical Engineering, Universiti Sains Malaysia, Penang, MALAYSIA
F.C. Ani
M.Y.T. Ali Jabil Circuit Sdn Bhd, Bayan Lepas Industrial Park, 11900 Penang, MALAYSIA
The purpose of this paper is to investigate the effect of vibration loading on structural integrity of LED joints strength in
manufacturing of automotive lighting. Ten LEDs were separately assembled on printed electronic circuit by dispensing a
conductive adhesive. There were five different build matrices constructed for the conductive adhesive and each matrix had two
samples. A platform was fabricated and mounted to the shaker to run the samples altogether in vertical and horizontal
positions. The samples were run under random vibration in frequency range of 3 Hz to 500 Hz for 4 hours. The assembly of
LED joints was further analyzed using finite element simulation to understand their stress-strain behavior under vibration
loading. The quality and integrity of LED joints were investigated using microscopy observation and x-ray analysis at the LED
joints. The results show there is no failure generated from the vibration test of applied period. The experimental result was
supported by finite element analysis on detail stress-strain behavior during the test.
Keywords: LED joints; Stretchable electronics; Vibration fatigue.
1 INTRODUCTION
Random vibration is the most related vibration
condition subjected in the environment conditions
due to destabilization which created by rotational and
tumbling masses. Automobiles and portable
electronic components are mostly exposed to random
vibration. A system can vibrate in a standard manner
that is known as vibration modes representing the
degree of freedom of the vibration in the system with
certain natural frequency. Range of frequency varied
according to its application, it had been tested that 2.5
Ton truck provide acceleration level 15 G and 19 G
with frequency 15 Hz – 40 Hz at speed of 10 mph -
15 mph (Steinberg & Wiley, 2001). In the
automobiles vibration frequency range is varied with
the system, vibration while driving, suspension
system, and steering system, which the vibration
range from 3 Hz – 1 KHz (Izumi, n.d.). A range of
frequency is excited at the same time, which means
that frequency loading correspond to all masses are
represented at the same time which produce large
displacement amplitude to produce impact of failure
to analysis system. To quantify random frequency,
Power Spectral Density (PSD) input is needed. PSD
is the input curve of acceleration units per Hz² versus
range of frequencies. PSD curve input is translated to
the Root Mean Square (RMS) value which the value
of each curve is formulated to a single quantity. As
the random vibration is the nature response statistic it
is represented in zero mean distribution where the
result is presented in three standard deviations of
RMS value in the presented value of stress, strain,
acceleration and velocity (Kumar & Engineer, 2008).
Several types of interconnection materials have been
used in electronic packaging such as a lead-based
solder, a lead-free solder and a conductive adhesive.
The lead-based solder was the main interconnection
material in electronics industry as it is requires low
temperature and less issues regarding to its quality.
However due to restriction in law and new
requirements for non-toxic materials, lead free solder
and conductive adhesive are used as alternative (Li &
Wong, 2006). Lead free solder is a tin based alloy,
common composition of lead free solder used are Sn-
Ag, Sn-Cu, Sn-Zn and Sn-Ag-Cu. Conductive
adhesives are a composite of adhesive epoxy and a
conductive metal such as silver, silver, tin oxide or
indium. Concept of the composition is that soft
material epoxy filler promote good contact within
and other particles as it deform and shrink during
curing process. Most common material for the epoxy
filler is silver as it is a good conductor, commercially
available and has moderate cost. Conductive adhesive
is applied as the electronics interconnections by using
Proceedings of the International Conference on Vibration, Sound and System Dynamics Penang, 2 August 2017
22
similar method as the solder paste, where it is
dispensed to the circuit before placement of
electronic component. Conductive adhesive does not
need a specified pressure for curing and curing
temperature is between 130ºC - 180ºC which suitable
for the thermal sensitive electronic components
(Siow & Lin, 2016).
The finite element analysis is used to predict stress
strain behaviour of the solder joint and fatigue
lifetime. There are two models that have been used in
other research which are detailed FE models and
smeared FE models. Detailed FE models however are
not recommended since for a commercial case as the
time required to build and solve such a model is
excessive when simplified models produce data of
appropriate accuracy much more quickly and with
less effort (Amy & Aglietti, 2009). The simulation of
FEA used to check the stresses on solder balls when
conducting a fatigue life assessment of the
component, the boundary condition setting that
identical to the vibration test is used in the analysis.
The variation in mesh densities is applied in the
model in order to examine the convergence of the
analysed frequency results. For verification of FEA
model, natural frequencies test vehicle are examined
experimentally with modal testing method (Chen,
Wang, & Yang, 2008). Besides, FEA is used to
obtain stresses for the fatigue estimation. The
analysis then was validated with experimental modal
analysis to determine natural frequencies and mode
shapes (Yu, Al-Yafawi, Nguyen, Park, & Chung,
2011). Researcher also, employed FEA to get
numerical deflection of solder joints similar to the
out-of-plane displacement measured with the
scanning vibrometer. The package is accurately
modelled. The modal analysis is first performed on
the global model and compared with experimental
results. Equivalent strain in the critical area of solder
joints is then determined with global-local approach
(Libot et al., 2016).
Aim of this paper is to investigate the effect of
vibration loading on structural integrity of LED joints
strength in manufacturing of automotive lighting.
The experimental work is supported by simulation
study in order to determine the stress generated due
to the vibration loading.
2 METHODOLOGY
2.1 Experimental procedure
The conductive adhesive (Loctite Ablestik ABP
2032S) that contained in syringe was taken out from
freezer and left in room temperature at least around
20-30 minutes for thawing process which depended
on volume of the adhesive in the syringe (reference).
The syringe with needle size 0.5 mm in diameter was
mounted to Liquid Dispenser (JB1113N) machine
and the machine was set at pressure 50 psi with
duration timing 1 s.
The adhesive was dispensed using five build matrices
and each build matrix had two samples in order to
investigate strength of the interconnections between
LEDs joints and the circuit as depicted in Fig. 1. The
circuit printed on polycarbonate sheet was cut into
small pieces and each piece consisted of one LED.
All of the samples were attached on a customized
platform using Araldite epoxy adhesive where five
samples were in vertical position and another five
samples were in horizontal position as illustrated in
Fig. 2. The platform was mounted on shaker to run
the vibration loading on the samples altogether. The
shaker was set at frequency range of 3 Hz – 500 Hz
using random vibration with running time duration 4
hours. Before the samples were tested, the shaker’s
amplitude was calibrated by mounting an
accelerometer on it and gets the desired amplitude.
The bonding between the joints and the circuits was
observed at every time interval 30 minutes and the
final inspection was done by using Hi-Scope
Advance (Hirox) and x-ray scanning. The assembly
of LED joints was further analyzed using finite
element simulation to understand their stress-strain
behavior under vibration loading. The quality and
integrity of LED joints were investigated using
microscopy observation and x-ray analysis at the
LED joints.
2.2 Simulation Procedure
Two 3-dimensional (3D) CAD models were
developed based on the assembly prototype model
that had been used in experiment as shown in Fig. 1.
CAD model of LED was developed by referring to
the actual component specifications from
manufacturer. The actual dimension of the adhesive
cannot be determined as it was not a geometry
structure, therefore, the adhesive 3D CAD model was
developed approximately as the actual prototype as
illustrated in Fig. 3. Whereas, models of the circuit
and polycarbonate base was modified to cater the
needs of simulation as shown in Table 1. The
complete assembly model as shown in Fig. 4.
Proceedings of the International Conference on Vibration, Sound and System Dynamics Penang, 2 August 2017
23
(a) (b)
(c) (d)
(e)
Figure 1: The build matrices of a) 3 dots at bottom + 1 dot
at top of the joints b) 2 dots at bottom + 1 dot at top of the
joints c) 1 dot at bottom + 1 dot at top of the joints d) 3
dots surround the joints e) 2 dots surround the joints
Figure 2: The vibration test setup for LEDs joint
(a) (b)
Figure 3: Assemble model for LED with adhesive for (a)
build matrix no. 1 and (b) build matrix no. 4.
Table 1: Dimension for each parts of geometry model
Parts Dimensions (mm)
Base of the lead 5.5×2.5×0.5
Adhesive geometry
model 1
6.5×3.5×0.5
Adhesive geometry
model 2
6.5×3.5×1.2
Silver circuit pad 20×15×0.08
Polycarbonate base 150×150×1
Figure 4: Complete assembly model
Materials properties were defined based on the
material used in the experimental prototype model.
Materials’ mechanical properties were provided from
the ANSYS materials library and manufacturers’
datasheets. All materials were assumed to be elastic-
plastic materials as summarized in Table 2.
2.3 Random Vibration Parameter Input
The solution data history from the modal analysis
was used in random vibration analysis. The number
of 12 mode shapes used and 0.02 of constant
damping ratio need to be set up in the analysis
settings. Random vibration is a non-deterministic
motion. The vibration pattern would be varied, to
quantify the frequency excitation PSD input need to
be assigned in the simulation. In this analysis band
limited white noise has been used which the spectral
density has a constant value over a quantified
frequency range as shown in Figure 5. Frequency
range from 3 Hz - 500Hz would be excited at the
same time. The PSD input was referred from the
JEDEC standard for vibration and variable frequency
to test the reliability of package devices under
various levels of application vibration to which
component can be exposed (Lai, 2009). The PSD
input applied in this analysis was based on level
where the component can be exposed to the most
Proceedings of the International Conference on Vibration, Sound and System Dynamics Penang, 2 August 2017
24
severe condition. PSD excitation was applied on the
model’s fixed support in Y-axis direction which was
perpendicular to the model plane.
Figure 5: Power Spectral Density input pattern and values
Table 2: Mechanical material properties
Parts Density
(kg/m3)
Young’s
modulus
(MPa)
Poisson
ratio
Yield
strength
(MPa)
Tangent
modulus
(MPa)
Ultimate
tensile
strength
(MPa)
Circuit pad 8900 130000 0.34 120 125 210
Polycarbonate
base 1200 2506 0.38 63 0.05 65
Adhesive 4500 4140 0.32 24.1 38.62 34.5
Lead 8300 110000 0.34 280 1150 491
3 RESULT AND DISCUSSION
3.1 Visual Inspection of LED Joints
Based on visual inspection, there was no
disconnection occurred at adhesive/circuit interface
after the vibration loading for both LEDs in vertical
and horizontal positions as shown in Fig. 6 and Fig.
7. Neither LED joints nor adhesive peeled off from
the circuit. It shows that strength of the adhesive was
enough to withstand the vibration loading even the
LEDs were positioned horizontally where more
internal force loading exerted on the adhesive
bonding at the LEDs joints since mass of the LED’s
head included as gravitational force.
Figures 8 and 9 show x-ray images of bottom view of
the adhesive bonding at the LEDs joints. There was
no defect such as crack propagated during the
vibration loading at the adhesive bonding. However,
some small voids were visualized in the adhesive for
a certain bonding. These voids formed during the
dispensing process of the adhesive at the joint and
they stayed in the adhesive after curing process. The
formation of the void did not contribute a significant
effect to the bonding as the bonding had not
experienced any failure after vibration loading.
Figure 9: X-ray images of bottom view for (a) build