TRANSACTIONS ON ELECTRICAL AND ELECTRONIC MATERIALS
Vol. 15, No. 4, pp. 201-206, August 25, 2014
pISSN: 1229-7607
Regular Paper
eISSN: 2092-7592
DOI: http://dx.doi.org/10.4313/TEEM.2014.15.4.201
Study on the Development of High-efficiency, Long-life
LED Fog Lamps for the Used Car Market
Sang Jun Park
Department of Mechanical Engineering, Kongju National University, Kongju 314-701, Korea
Young Lim Lee†
Department of Mechanical and Automotive Engineering, Kongju National University, Kongju 314-701, Korea
Received March17, 2014, Revised April 11, 2014; Accepted June 2, 2014
LED lighting,considered to be a new growth industry, has attracted a great deal of attention due to its higher
illumination and longer life time than existing light sources. In this study, high-efficiency and long-life LED fog lamps
for automobiles were developed, which can substitute the existing 27 W halogen fog lamps for a used car market. For
this purpose, the number of LED modules, the body, heat sink, and the output of the fog lamp were first optimized
through a numerical analysis. Then, a 10 W-class LED fog lamp was prototyped based on the optimized numerical
model, and the performance of the fog lamp was successfully verified through the experiments.
Keywords: LED, Fog lamp, Thermal design, Numerical analysis, Used car market
1. INTRODUCTION
LED lighting is considered to be a growth industry and this
can be regarded as an important factor for solving the environment and energy problems according to the global policy. It has
higher illumination, longer life, and higher energy efficiency
than other light sources such as halogen bulbs and high intensity
discharge(HID) lamps, etc. Table 1 shows the characteristics of
light sources, and an LED lamp has excellent life and light efficiency in comparison with a halogen lamp or an HID lamp. In
addition, the LED lamp consumes about half of the HID lamp
power.
For automotive fog lamps, a 27 W-halogen lamp is generally
used, and it is known that the power consumption can be reduced to as low as 10 W if using an LED lamp. While the life time
is about 2,000 hours for halogen lamps, this can be extended to
about 50,000 hours for LED lamps.
†
Author to whom all correspondence should be addressed:
E-mail: ylee@kongju.ac.kr
Copyright ©2014 KIEEME. All rights reserved.
This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial
License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted noncommercial use,
distribution, and reproduction in any medium, provided the original work is properly cited.
Copyright
2011 KIEEME. All rights reserved.
201
In summary, LED lamps have many advantages such as lower
power consumption, longer life, higher color efficiency, no ultraviolet and infrared emission, and stronger durability against
vibration and impact. However, LED lamps need to be thermally
well managed to ensure their normal operation. In particular,
high-power LED accelerates the decomposition between the
parts due to phenomena such as the light efficiency reduction of
the light beam, the reduction of the forward voltage, and the increase of the fundamental wavelength of the emitted light due to
the heat occurring in the junction during operating; accordingly,
the LED lifespan is reduced [1,2]. For LED lights, if the junction
temperature reaches 135~150℃, the chip can become disconnected. Also, problems such as color-temperature variation and
the reduction of the lifespan of LED can occur. A high-efficiency
heat sink that can lower the temperature of the junction is therefore required [3].
Among LED heat sinks, the plate-fin heat sink [4,5] has typically been used because it does not require a high cost. Also, a
heat pipe [6] is used for high-power LED lights due to its high
heat transfer rate. In addition, many advanced cooling technologies that can greatly decrease high-power LED junction temperature are being studied. As shown in Fig. 1, the soldering method
is used to remove the thermal resistance between the LED mod-
http://www.transeem.org
Trans. Electr. Electron. Mater. 15(4) 201 (2014): S. J. Park et al.
202
Table 1. Characteristics of lights.
Life time(hour)
Electric power (W)
Luminous efficiency (lm/W)
Halogen
330
55~58
26
HID
2,000
35
91
LED
50,000~100,000
18
100
ule and the aluminum body. The micro jet array cooling system
has also been suggested [7,9], and similar cooling methods such
as the use of a synthetic jet [10,11], a micro channel cooler [12],
thermoelectric cooling [13], and a piezoelectric fan [14], etc. have
been studied.
In order to increase the efficiency of LED lamps, Kim et al.
[15] compared the performance of an LED lamp with and without heat pipes. In addition, Liu et al. [16] studied forced cooling
technology using a micro jet, and reported that the junction
temperature of LED could be further reduced by 23℃. Chen et
al. [17] reported the effect of the contact resistance for LED with
a metal core printed circuit board(MCPCB) through numerical
analysis and experiment. Kang et al. [18] manufactured a heat
sink for LED headlamps and studied the performances of cooling
and illumination intensity by varying the fin-base area and fin
length. Weng [19] numerically studied the relationship between
thermal resistance and junction temperature under various
conditions such as the material properties of the printed circuit
board (PCB), the cooling condition, and chip size with a power of
0.1~1 W. He identified interactions between parameters and suggested a thermal design rule for the development of LED lights.
Hwang, et al. [20,21] predicted the variation of junction temperature according to the fin shape, PCB material properties, and the
number of LED modules for a 10 W LED light. They showed that
the life time of a fan could be extended up to 164.5% by on-off
control of the fan.
The lighting systems of some automobiles have already been
replaced with LED lighting since it has high efficiency and long
life. However, an LED lamp that can replace a halogen bulb of a
used car has not yet been developed. Therefore, in this study, a
high-efficiency and long-life LED fog lamp for a used-car market
has been developed by optimizing thermal performance. First,
to optimize the thermal performance of the LED fog lamp, the
number of LED modules and the properties of the body, heat
sink, and power were optimized. A prototype of a 10 W LED fog
lamp was then manufactured and its thermal performance and
illumination intensity were successfully verified from experiments.
Fig. 1. Schematic of a 1-dimensional thermal resistance model.
Fig. 2. Schematic for components of the numerical model.
Fig. 3. Specifications of the numerical model.
2. NUMERICAL ANALYSIS AND
EXPERIMENTAL METHOD
2.1 Numerical analysis
Thermal analysis was conducted to predict the junction temperature of an LED fog lamp. A steady-state thermal analysis
module in Ansys [22] was adopted and, for the material properties and boundary conditions, the LED chip is assumed to be
silicon, the lens is silicone, and the body and PCB are both aluminum.
The heat conversion rate of LED power is about 70%. Furthermore, it is assumed that the outside air temperature was 25℃,
the convective heat transfer coefficient is 2 W/m2K, the thermal
resistance, and θj-s is 2℃ /W. An LED chip was modeled as a heat
source for simplification of the grid and reduction of the analysis
time in the numerical analysis. Figure 2 shows a schematic of the
numerical model, and Fig. 3 shows the specification of the numerical analysis model.
The number of LED modules plays an important role in decreasing the junction temperature due to the effect of power dispersion. Thus, in order to optimize the number of LED modules,
the number of LED modules was increased from 1 to 6 as shown
in Fig. 4. In addition, the LED thermal performance according
to the shape of the heat sink was investigated, and the platefin heatsink and the radial-fin heatsink are shown in Fig. 5. The
thermal performance was optimized by varying the fin number
of the heatsink from 4 to16, as shown in Fig. 6. Table 2 shows a
summary of the numerical analysis models considered in this
study. Catia [23] was used for the three-dimensional shape design.
2.2 Experimental method
Figure 7 shows a schematic diagram of the heat radiation
experiment. The thermal performance of an LED fog lamp was
verified by applying power from 8 W to 12 W using a power sup-
Trans. Electr. Electron. Mater. 15(4) 201 (2014): S. J. Park et al.
(a)
(b)
(c)
(d)
(e)
(f)
203
(a)
(b)
(c)
Fig. 4. Numerical models with the different numbers of LED modules
(a) model 1, (b) model 2, (c) model 3, (d) model 4, (e) model 5, and (f)
model 6.
Table 2. Summary of numerical models with model number.
Model No.
1
2
3
4
5
6
7
8
9
10
11
(a)
Number of LED module Fin type and number
1
2
3
4
5
6
6
Plate / 12 ea
6
Radial / 12 ea
6
Radial / 4 ea
6
Radial / 6 ea
6
Radial / 16 ea
Fig. 6. Numerical models with the different numbers of heatsink fins
(a) model 9, (b) model 10, and (c) model 11.
Fig. 7. Schematic diagram of heat radiation experiment.
(b)
Fig. 5. Numerical models with the different shapes of heatsink fins (a)
model 7 and (b) model 8.
ply. A T-type thermocouple was used to measure the temperature, and the data was obtained by using a recorder. The junction
temperature (Tj) of the LED fog lamp was not directly measured
but was calculated by using the solder point temperature (Tsp)
and thermal resistance (θj-sp) given by the manufacturer. The relationship between Tj and Tsp is as follows:
Tj = Tsp+ θj-sp*Power
(1)
When measuring the solder point temperature, the effect of
radiation was minimized by wrapping the thermocouple with
a radiation shield tape. The thermal equilibrium of the LED fog
lamp was reached within 40 minutes after the LED light was
turned on.
Fig. 8. Variation of junction temperature with the number of LED
modules.
3. RESULTS AND DISCUSSION
3.1 Junction temperature according to the number
of LED modules
The case of LED modules mounted on the aluminum body
without the heatsink was first considered. The junction temperature was numerically predicted by varying the number of LEDs,
204
and the result is shown in Fig. 8. In the case of model 1,where a
single LED module of high power is mounted, the junction temperature was as high as about 359℃. The target junction temperature should be below 100℃. When the number of modules
increases with low-power chips, the junction temperature exponentially decreases, but the rate of decrease is gradually reduced.
For model 6, where six LED modules are used, the junction temperature reaches about 257℃, showing a temperature decrease
of 102℃ compared to model 1. Therefore, using multiple lowpower chips is more favorable for reducing junction temperature
than using a single high-power LED due to the thermal dispersion effect. In general, phenomena such as chip disconnection,
color temperature variation, and sudden life-shortening occur
if the LED junction temperature reaches higher than 135~150℃.
Thus, since the
Thermal management with the aluminum body alone is not
sufficient, an additional thermal device such as a heat sink becomes necessary.
Trans. Electr. Electron. Mater. 15(4) 201 (2014): S. J. Park et al.
Fig. 9. Variation of junction temperature with the shape of heatsink
fins.
3.2 Junction temperature according to the fin
shape of the heat sink
In this study, two types of heat sinks mounted on the aluminum body are considered to further improve the thermal performance of the LED light: the plate-fin type and the radial-fin type.
Since the heat sink is assumed to be attached to the body by a
soldering process, additional thermal contact resistance was not
included. Figure 9 shows the variation of LED junction temperature with the shape of the heat sink. It was assumed that each
heat sink has twelve fins. The junction temperature of model
7(the plate-fin type), is found to be about 140℃, and the junction temperature of model 8(the radial-fin type), is about 99℃.
Thus, it can be seen that the radial-fin type is considerably more
efficient in thermal management. This is mainly because the
surface area of the radial-fin type is about 2.4 times larger than
that of the plate-fin type. Although the junction temperature is
below 100℃ for the radial-fin type, it can be further lowered by
optimizing the number of radial fins.
Fig. 10. Variation of temperature at junction and heatsink fin according to the number of heatsink fins.
3.3 Junction temperature according to the number
of radial fins and amount of input power
Figure 10 shows the variation of the LED junction temperature with the number of radial fins. The junction temperature of
model 6 without a heat sink is 257℃. When the heat sink with
four fins is mounted, the junction temperature lowers to about
149℃. This was expected since the surface area for heat release
increases with the number of fins. The junction temperature of
model 11 that has 16 fins decreased to about 87℃, and this corresponds to about a 42% decrease in the junction temperature
compared to model 8. The ambient temperature of a car fog
lamp can increase up to 45℃ in some places, such as in middleeastern countries. In this case, the junction temperature can
increase from 87℃ to 107℃. However, it is still lower than 120℃,
which is the maximum allowable junction temperature recommended by the manufacturer. Thus, if 6 LED modules and a heat
sink with 16 radial fins are used, the thermal performance of a 10
W LED fog lamp can be sufficiently secured.
Figure 11 shows the contour plot of the surface temperature
of model 11. While the LED junction temperature is 87℃, the
temperatures of the body and heat sink are 67℃ and 64℃, respectively. Although there is a difference between the junction
temperature and heat sink temperature, the tendency of temperature variation with the number of fins is similar.
Currently, the power of commercial LED fog lamps ranges
from 7.5 to 12 W. Thus, the variation of junction temperature
Fig. 11. Contour of temperature for an LED fog lamp with 10 W.
with LED input power was studied. Figure 12 shows a variation of the junction temperature of model 11withinputpower.
The junction temperature is linearly proportional to the input
power. The junction temperature is 75℃ at 8 W of input power,
whereas if the input power is increased to 12 W, the junction
temperature increases to 99℃.For an ambient temperature
of 45℃, the junction temperature becomes 119℃, which is
still lower than the maximum allowable temperature of 120℃.
This suggests that a 12 W LED fog lamp can be developed with
Trans. Electr. Electron. Mater. 15(4) 201 (2014): S. J. Park et al.
205
6 LED modules and 16 radial fins without any thermal problems.
3.4 Experiments on the thermal performance of an
LED fog lamp prototype
The prototype of the LED fog lamp was manufactured based
on model 11, as shown in Fig. 13. Model 11 has an aluminum
body with 6 LED modules and a heat sink with 16 radial fins.
First, the variation of the junction temperature with input power
was measured as shown in Fig. 14, where input power varies as
8 W, 10 W, and 12 W. The junction temperature was calculated
from the measured temperature of the soldering point with the
thermal resistance supplied by the manufacturer. At 8 W, the
soldering point temperature was measured as 62.0℃, and the
corresponding junction temperature was 78.0℃. As the input
power gradually increases, the junction temperature also linearly
increases with the power. At 10 W, the junction temperature
reached 85.8℃ and increased to 94.6℃ at 12 W. Therefore, it is
confirmed that the LED fog lamp developed in this study can be
used with up to 12 W of input power. The illumination variation
of the LED fog lamp with input power was also investigated. The
intensity of illumination was measured at the distant point of 20
cm, perpendicular to the LED lens. The intensity of illumination
was measured as 2,688, 3,181, and 3,610 lux for 8, 10, and 12 W,
respectively. The typical illumination intensity of a 27 W halogen
fog lamp is 1,010 lux, so the intensity of a 12 W LED lamp is 357%
greater than that of a halogen lamp.
Fig. 12. Variation of junction temperature with input power at ambient temperature, 25℃.
4. CONCLUSIONS
In this study, the properties of the LED module, the body, heat
sink, and power of an LED fog lamp were optimized in order to
develop an LED fog lamp that can replace halogen fog lamps
in used cars. In addition, the performance of the fog lamp was
verified with a prototype of the LED fog lamp. The conclusions
drawn from this study are as follows:
(1) Multiple low-power LED’s are more advantageous in reducing the junction temperature of a fog lamp in comparison
with a single high-power LED. The junction temperature
with6 LED modules can be greatly reduced by 105℃ compared to a single LED module.
(2) The radial-fin type heatsink is considerably more efficient
than the plate-fin type for thermal management since it
has 2.4 times more surface area for heat release. The radialfin type can lower the junction temperature by about 40℃
compared to the plate-fin type.
(3) 6 LED modules, a heatsink with 16 radial fins, a 6-faces
light body, and 12 W were finally determined as effective
properties for LED fog lamps. The prototype of the LED fog
lamp showed sufficient thermal performance, where the
LED junction temperature is 94.6℃ for an input power of 12
Wand an ambient temperature of 25℃. In addition, the illumination intensity of the 12 W LED lamp was 357% greater
than that of a typical halogen lamp.
Fig. 13. Prototype of the LED fog lamp.
Fig. 14. Variation of junction temperature with power of the LED fog
lamp.
REFERENCES
ACKNOWLEDGMENT
This work was supported by the Human Resources Program in
Energy Technology of the Korea Institute of Energy Technology
Evaluation and Planning(KETEP) with financial resource granted
from the Ministry of Trade, Industry & Energy, Republic of Korea
(No. 20134030200230).
[1] D. G. Todorov and L. G. Kapisazov, Proc. Electronics(Sozopol,
Bulgaria, 2008) p.139.
[2] J. Fan, K.C. Yung, and M. Pecht, Device and Materials Reliability,
IEEE Trans., 11, 407 (2011).
[3] J. Hu, L. Yang and M. W. Shin, J. Phys. D: Appl. Phys., 41, 35
(2008).
[4] X. B. Luo, W. Xiong, T. Cheng, and S. Liu, IET Optoelectron., 3,
206
225 (2009)
[5] A. Christensen, M. Ha, and S. Graham, 7th International Conference on Solid State Lighting, 6669 (San Diego, 2007).
[6] L. Kim, J.H. Choi, S. H. Jang, and M. W. Shin, 6th Symposium of
the Korean Society of Thermophysical Properties, 455, 21 (2006).
[7] S. Liu, J. H. Yang, Z. Y. Gan, and X. B. Luo, Int. J. Therm.
Sci., 47, 1086 (2008) [DOI: http://dx.doi.org/ 10.1016/
j.ijthermalsci.2007.09.005].
[8] X . B . L u o , W. C h e n , R . X . S u n , a n d S . L i u , H e a t
Transf. Eng., 29, 774 (2008) [DOI: http://dx.doi.org/
10.1080/01457630802053777].
[9] X. B. Luo and S. Liu, IEEE Trans. Adv. Packag., 30, 475 (2007)
[DOI: http://dx.doi.org/10.1109/TADVP.2007.898522].
[10] M. Arik, Appl. Thermal Eng., 27, 1483 (2007) [DOI: http://
dx.doi.org/ 10.1016/j.applthermaleng.2006.09.027].
[11] B. M. Song, B. Han, A. Bar-Cohen, R. Sharma, and M. Arik, IEEE
Trans. Compon. Packag. Technol., 33, 728 (2010) [DOI: http://
dx.doi.org/ 10.1109/TCAPT.2010.2051034].
[12] L. L. Yuan, S. Liu, M. X. Chen, and X. B. Luo, Proc. of 7th International Conference on Electronics Packaging Technology, 1-5,
26-29 (2006).
[13] J. H. Chen, C. K. Liu, Y. L. Chao, and R. M. Tain, Proc. of 24th
International Conference on Thermoelectrics (ICT), 53-56, 19-23
Trans. Electr. Electron. Mater. 15(4) 201 (2014): S. J. Park et al.
(2005).
[14] T. Acikalin, S. V. Garimella, J. Petroski, and A. Raman, Proc. of
9th Intersociety Conference on Thermal and Thermomechanical
Phenomena in Electronic Systems, 1, 663-671, 1-4 (2004) [DOI:
http://dx.doi.org/ 10.1109/ITHERM.2004.1319239].
[15] L. Kim, J. H. Choi, and M. W. Shin, Thermochimica Acta, 455, 21
(2007).
[16] S. Liu, J. Yang, Z Gan, and X. Luo, International Journal of Thermal Sciences, 47, 1086 (2008) [DOI: http://dx.doi.org/ 10.1016/
j.ijthermalsci.2007.09.005].
[17] H. Chen, Y. Lu, Y. Gao, H. Zhang, and Z. Chen, Thermochimica Acta, 488, 33 (2009) [DOI: http://dx.doi.org/ 10.1016/
j.tca.2008.12.019].
[18] B. D. Kang and H. W. Kang, and H. W. Kang, KASE, 7, 1147 (2007).
[19] C. J. Weng, International Communications in Heat and Mass
Transfer, 36, 245 (2009) [DOI: http://dx.doi.org/ 10.1016/j.ichea
tmasstransfer.2008.11.015].
[20] S. H. Hwang, S. J. Park, and Y. L. Lee, Journal of KAIS, 7, 2317
(2010).
[21] S. H. Hwang and Y. L. Lee, Journal of KSME, 35, 569 (2011).
[22] Ansys, Version 12.1 2005
[23] Catia, V5R17, Dassault Systems, 2006.