thermal management of outdoor led lighting systems and streetlights--variation of ambient cooling conditions.
According to the US Department of Energy SSL manufacturing roadmap 2011 [
Basil\'s consultation and other 2011]
The relative manufacturing cost of mechanical structure and thermal management of LED Downlight fixtures, although the luminous efficiency of LEDs has been significantly improved, is expected to grow from about 15% to 25% in the ten years from 2010 to 2020.
In 2011, the manufacturing cost of mechanical and thermal components for outdoor LED lamps accounted for about 40% of the total manufacturing cost. [
Basil\'s consultation and other 2011]
The cooling element significantly affects the size, weight and shape of the fixture, thus affecting the cost, design flexibility, material consumption and overall environmental impact of the LED fixture, being promoted as the most eco-friendly
Ten years of friendly lighting solutions.
In fact, the cooling of LED lamps is a compromise.
This is an issue of selecting acceptable working levels for LED junction temperatures, designing cooling elements capable of maintaining this temperature level and not exceeding the maximum rated junction temperature given by the assembly manufacturer.
However, without proper knowledge of the heat transfer conditions applied to the installation environment of the lamps, it is difficult to make such a compromise. A worst-
Case-Based Design methods do not adversely affect the thermal performance of the system.
However, this approach lacks complexity in optimization.
In this study, outdoor weather conditions around the world were studied to determine global and local changes in night temperature and wind speed.
Dew point temperature, precipitation and snowfall data were also studied.
The main assumption is that the design parameters of thermal management of outdoor LED lighting systems will vary greatly between different geographical locations, and global and general design parameters such as ambient temperature 50 [degrees]
C, or assuming that there is no natural convection of the wind, does not reflect the actual situation.
This paper suggests that the cooling design method and temperature limit of LED lamps should be evaluated more carefully and the influence of wind and heat radiation should be solved.
In addition, it is recommended that the cooling design of outdoor LED lamps is twofold. A worst-
The case method should be used to check that the LED junction temperature never exceeds the maximum rated.
In addition, average-condition-
A cooling-based approach should be adopted.
This study aims to provide tools for both methods.
The values given in this paper can serve as a basis for further research on thermal management of outdoor LED lamps and the development of more optimized cooling solutions.
2 Heat transfer between outdoor LED lamps and the surrounding environment.
1 The heat transfer method transmits heat from the surface of the thermal shell to the surrounding environment through two different heat transfer mechanisms of convection and radiation.
Through the installation of the surrounding structure, conduction is also possible.
Newton\'s law of cooling can be used to represent convection heat flow [q. sub. c](W)
From the surface of the [area]A. sub. s]([m. sup. 2])
Can be written :【q. sub. c]= [h. sub. c][A. sub. s]([T. sub. s]-[T. sub. [infinity]]/1/([h. sub. c][A. sub. s])= [T. sub. s]-[T. sub. [infinity]]/[R. sub. c](1)where [h. sub. c][(W/m. sup. 2]K)
Is the convection heat transfer coefficient, and [T. sub. s]and [T. sub. [infinity]](K)
Surface and ambient air temperature, respectively.
The equation can also be expressed by convection thermal resistance [R. sub. c](K/W). [ASHRAE 2009]
The heat transfer coefficient or the thermal resistance of the convection is not a constant, but is actually a very complex parameter.
They are strongly influenced by the air flow properties, which depend on the environmental conditions, and so is the temperature difference between the surface and the ambient air.
Relatively small area [A. sub. s][(m. sup. 2])
In a large constant temperature fenceT. sub. a](K)
, From the surrounding environment to the surface is equal to the black body emission power of the surrounding environment.
Therefore, the equation of net radiation heat transfer [q. sub. r]
From the surface, when [T. sub. s]+++++ [T. sub. a]
Can be written :【q. sub. r][epsilon][sigma][A. sub. s](T. sub. s. sup. 4]]-[T. sub. a. sup. 4])(2)where [sigma](5. 67 * [10. sup. -8]W/[m. sup. 2][K. sup. 4])is the Stefan-
Constant and [epsilon]
Radiation rate on the surface. [T. sub. s]
Again, the temperature of the surface being cooled and [T. sub. a]
Ambient temperature of radiation. [ASHRAE 2009]2.
2 Special features of outdoor cooling changes in temperature and wind speed affect heat transfer.
Influence of humidity on convection heat transfer [
Or wait for 1998].
Locally, especially globally, these outdoor parameters are expected to change significantly, resulting in changes in heat transfer and LED junction temperature, which will affect the reliability and optical performance of LED lighting systems.
Many studies have been carried out in areas such as thermal management of buildings, outdoor telecom cabinets and enclosures with battery systems, such as Baer and Harrison , McKay 
Kelley and others. 1995], Estes , Marongiu and ASHRAE .
However, many of them highlight the solar load applied to the housing.
In the case of a lighting system, the solar thermal load can be considered less important, especially in the case of low lamps
Weight, surface with low radiation absorption coefficient, used only at night. McKay 
List the phenomena that affect the internal temperature of the outdoor shell: heat generated by the components in the Shell, ambient temperature, solar radiation, wind speed, shadows and reflections caused by the surrounding environment, and shell design. Estes 
Interesting observations related to radiation dissipation.
He specifies two ambient \"radiators\", ambient air and a remote sky, the former being primarily a convection environment and the latter being a radiation environment.
The remote sky temperature depends on the dew point and dry ball temperature and is always below the ambient air temperature.
According to Estes 
However, in the actual cooling design, it should be assumed that the remote sky temperature is the same as the ambient air temperature, although it can be much lower in a dry desert climate, for example.
It should also be noted that only the top of the shell is facing the sky radiation, while the bottom is facing the ground. [
Bell and Harrison 2008.
The temperature of the sky is calculated by Eicker, Dalibard as: [T. sub. sky]= [[epsilon]. sub. sky. sup. 0. 25][T. sub. [infinity]](3)where [T. sub. sky]
What is the temperature in the sky ,[[epsilon]. sub. sky]
The radiation rate of the sky and [T. sub. [infinity]]
Ambient air temperature.
The radiation rate of the sky varies with the amount of cloud and dew point temperature.
Models of various sky radiation rates can be found in the literature.
Most of them, however, apply only to clear sky conditions [Eicker 2011].
When it comes to convection cooling of outdoor housing, McKay 
It has been observed that in the case of better heat dissipation in the Shell, higher wind speed will lead to better convection heat transfer.
In winter, the effect of wind speed is less significant when the housing temperature is low [McKay 1988]. 2.
3 temperature and wind speed range and cooling of outdoor fixture Marongiu and Estes 
The outdoor temperature range of the specified electronic cooling design is-40 to +55 [degrees]C, and -40 to +46 [degrees]
C in North America.
Cowsley and others 
Ambient temperature range for use-30 to +55 [degrees]C.
Cole and others. 2010]
The temperature limit of different types of lighting systems is studied and it is concluded that most lighting systems are designed for the ambient temperature of 25 [degrees]
C, and the limit of LED operation comes from-50 to +40 [degrees]C.
For metal halides and high
They suggest that the start-up temperature of the lamp is not less-35 and -40 [degrees]C, respectively.
For the upper limit temperature of these lamps, they recommend that the temperature be between 40 and 65 [degrees]C.
Finnish standard sfs6 000-7-
714 listing special requirements for outdoor fixtures, it is recommended to consider the following temperature ranges40 to +40 [degrees]C. [SESKO ry 2007]
In recent years, LED refrigeration has been widely studied.
However, most studies focus on LED components or thermal interface materials and radiators used with LED.
For example, LED Yang et al [have covered the cooling of LED street lamps]2010, 2011]
Luo Xiaobing and others [2008, 2009]
Marosy and others 
Luo MOU and others [2007, 2009]
Kai Keding and others 
Wang and others. 2010].
Typical features of these studies are that ambient temperatures, sometimes even called room temperature, are set to specific values such as 11, 20, 25, 35, or 40 [degrees]C.
Another typical assumption is that the effects of radiation are negligible.
None of the above mentioned wind speed
It mainly involves the study of convection LED street lamp cooling.
In addition, if the convection heat transfer coefficient is mentioned, they are given as constant values without explaining their selection or considering their variation.
American Society of Heating, Refrigeration and Air
Air conditioning engineer (ASHRAE)Handbook 
Provides a wide range of thermal design Climate data for thousands of locations around the world, including wind speed data.
However, even if multiple references for night cooling of buildings are provided, the night cooling conditions are not specifically addressed. 2.
A new ideal method for cooling LED lamps 
It is pointed out that although it is important to limit the peak temperature inside the outdoor electronic housing, the long-term average temperature and temperature gradient of the outdoor housing should also be considered.
Therefore, in addition to knowing the upper and lower limits of temperature, more attention should be paid to the changes and typical values of temperature and wind speed.
Except ASHRAE 
Climate data and Chen Gu et al 
They have incorporated a large amount of outdoor temperature data into the reliability function used by outdoor electronic products, and few of the papers mentioned earlier consider the changes in these parameters.
The Led is strongly affected by the temperature, and the thermal management of these lamps is critical.
The worst thing, however, is that
The case method, that is, selecting the maximum ambient temperature and junction temperature as the main design parameters, can be considered problematic for two reasons.
First of all, the LED component manufacturer guarantees a longer life and usually specifies the warranty at a temperature much lower than the maximum rated junction temperature of its component.
Therefore, it is doubtful to use the highest junction temperature as one of the main cooling design parameters.
Second, design the lamps for the ambient temperature of for example 40 or 50 [degrees]
Natural convection conditions without wind may be far from reality.
In addition, it should include heat radiation heat transfer depending on the surface of the shell and the surrounding Kelvins to its fourth power, so it is highly dynamic in terms of temperature change, which cannot be ignored in the thermal design practice of LED lamps and lanterns.
The goal of this article is to find answers to the following questions.
What is the actual average ambient night temperature and wind speed, what are the differences between them and daytime conditions, and what are the differences between them at different locations?
How does the assumption of zero wind speed reflect the true cooling conditions of outdoor LED lamps and whether there is a position-
Difference in average wind speed?
Overall, should the cooling of outdoor LED fixtures be different from the current common practice?
In this study, global weather conditions were studied to determine what kind of temperature and wind are susceptible to LED outdoor fixtures.
Weather data for 2007, 2008 and 2009 were obtained from the National Climate Data Center of the National Oceanic and Atmospheric Administration (NOAA)
Secretary of Commerce. [
NOAA 2008, 2009, 2010]
50 geographic locations showing a large number of records were selected for further study.
A total of 150 measuring stations per year, more than 2.
The study selected 7 million lines of weather data.
In addition to the various elevations, countries and continents, these sites were selected to represent highly illuminated areas with different climatic conditions, coastline and inland areas.
Figure 1 photo pollution chart [
Cinzano and other 2001]
, Universal lighting use indicator added with the selected location.
In addition to the NOAA weather data, sunrise and sunset data for selected locations were obtained from the US Naval Observatory [USNO 2010].
With the help of these data, the NOAA climate data is divided into night and day data sets.
Assuming that outdoor lighting is only used between sunset and sunrise, this simplification is not always fully applicable despite the various control systems installed outdoors.
Because the main goal is to study changes in weather conditions between different locations, not for long periods of time
The word climate change
Select the annual survey period.
Calculate the following from the night Data Sheet: local and global arithmetic mean temperature and wind speed, minimum and maximum, standard deviation, standard error, temperature and wind speed range.
The daytime average is also calculated.
In addition to the temperature and wind speed, precipitation, snowfall and dew point temperatures were studied.
In addition, each location is divided into a location set or area to be evaluated, where the area-
A specific cooling design can be adopted.
4 RESULTS 4.
1 day and night time duration is calculated based on USNO data, and daily day and evening time duration for each study site is calculated.
In areas with higher latitudes, the summer days are long, while in the vicinity of the equator, the length of the day does not change much during the year.
At higher latitudes, a fraction of the total service life of LED lamps used at warm summer temperatures.
Figure 2 shows the monthly average night duration of four locations representing different latitudes.
Research location in the northernmost part of hemaven, Sweden (65\'46\" North)
The average night duration in June was only one hour.
This change is taken into account by separating the weather data at night and during the day.
Therefore, data, especially in the northern region, have a greater impact on winter records. 4.
2 the arithmetic average night temperature of the 50 locations selected by the temperature is 12. 5 [degrees]C.
Day and 24-
The average hourly temperature is 17. 6 and 15. 1 [degrees]C, respectively.
The average temperature during the day is 5. 1 [degrees]
C is higher than night temperature.
The biggest difference in the average day and night temperature is 10. 0 [degrees]
C. In Astana, ha, the difference is the smallest. 1 [degrees]
C in Hong Kong, China.
The average temperature at night is-3. 2 [degrees]
C in Edmonton, Canada.
The average temperature at night is 28. 8 [degrees]
C. Causing 32 in Khartoum, Sudan [degrees]
C within the average night temperature range of 50 selected locations.
The maximum temperature in the night data is 50 [degrees]
C. Measured in Jakarta, Indonesia, May 31, 2009.
The lowest night maximum temperature of 17 in the selected location was measured. 2 [degrees]
C. La Paz, Bolivia, at an altitude of about 4060.
The second lowest local maximum temperature at night was measured at 18. 3 [degrees]
C. Hemavan, Sweden, at an altitude of about 790.
The highest night temperatures in all 50 locations averaged 31. 9 [degrees]C.
Figure 1 shows the difference between mean and maximum night temperature3.
In the same picture, AHSRAE 
Climate data, especially
Average daily temperature, average 0.
The hottest temperature for comparison is 4%.
The difference between the average and the maximum night temperature at all locations was 19 on average. 4 [degrees]C.
Calculate the maximum difference 5. 5 [degrees]
C. The highest 34 in Côte d\'Ivoire and Côte d\'Ivoire [degrees]
C. In Astana, Kazakhstan.
Table 1 shows the average night temperature at seven selected locations.
These sites are divided into three groups of latitude ranges.
The other is made up of 21 cities in Europe, and the other two are located in inland and seaside areas.
The number of locations within the collection ranges from 12 to 50.
Table 1 also lists the highest average night temperature for each group and their respective locations. 4.
3 the arithmetic average night wind speed of the selected 50 stations is 3. 1 m per second.
Day and 24-
The average hourly wind speed is 4. 1 and 3.
6 metres per second, respectively.
In all 50 locations, the average daytime wind speed is higher than the average nighttime wind speed.
The difference between the average wind speed of the day and night is 1. 0 m per second.
The maximum difference in wind speed between the average day and night is 2.
Alice Springs, Australia, 1 m per second, the minimum difference is 0.
Swedish hemavin is 2 metres per second.
The average wind speed at night is 1.
Singapore 1 m per second.
The average wind speed at night is up to 5.
Dublin, Ireland, 4 metres per second, resulting in 4.
Within the average nighttime wind speed range of 50 selected locations, 3 m per second.
The maximum wind speed in the night data is 50.
1 m per second measured in Khartoum, Sudan, on February 15, 2007.
The highest wind speed at night is second, 36.
On June 1, 2009, 2 metres per second was measured in Jakarta, Indonesia.
The lowest wind speed at night at all locations is 0 m/s.
Calculate the percentage of night time records showing zero wind at each location. On average, 8.
Wind speed data records of 8% at night show no wind.
The highest percentage is 32.
Singapore is 2%, with a minimum of 0.
04% in Hong Kong
The mean value of the maximum wind speed at night in all 50 locations is 19. 8 m per second.
Figure 1 shows the difference between the average wind speed and the maximum wind speed at night
4. With ASHRAE 
Average of 1% of the highest wind speed.
The difference between the average wind speed at night and the maximum wind speed at all locations is 16. 7 m per second.
Calculate the maximum difference 6.
Kuala lumpur, Malaysia, is 8 metres per second, up to 45 metres.
In Khartoum, Sudan, 9 metres per second.
Table 2 shows the average wind speed at night at seven groups of locations.
Table 2 also lists the areas with the lowest average wind speed at night and their respective locations. 4.
4 Other parameters night temperature and wind speed are the most important variables in the thermal management of LED outdoor lamps.
In addition to these two parameters, dew point temperature, precipitation and snowfall are also factors that affect the performance of outdoor LED lamps.
The latter not only has an impact on the thermal management of LED lamps, but also on road lamps.
Figure 5 shows a comparison of mean night air and dew point temperature, which can be used to estimate the humidity and sky temperature of the heat transfer from radiation.
Hourly data for precipitation and snow depth are harder to find than temperature and wind speed data.
For some places, these data are not recorded from time to time in the NOAA data set [
NOAA 2007, 2008, 2009].
Figure 6 shows the average monthly snow depth (cm) at two sites with the highest snow depth record ).
With regard to precipitation, only data from Prague, Czech Republic, show that there are records of precipitation recorded almost every hour, with a total of 22,956.
Figure 7 shows the monthly percentage of night precipitation records in Prague, Czech Republic.
In this paper, solar radiation has not been studied since it only focuses on night use.
Sun data and instructions for calculating the Sun effect can be found in [ASHRAE 2009].
5 The simplified LED light fixture is simulated to provide a practical example of the simplified LED light fixture is simulated and measured in the laboratory. 5.
1 Case of lamp structure the lamp consists of six high-power LEDs (Bridgelux BXRA-C2002)
And a radiator, which is a 400mm x 4mm aluminum sheet (AlMg3-5754).
Led is 22. 8 mm x 25.
The size is 8mm, and the coordinates from the corner of the aluminum plate are 163: 136; 163:198; 163:260; 242:136;
242: 198 and 242: 260mm.
According to the data sheet, the thermal resistance of the LEDs is 0.
5 K/W, nominal luminous flux of 2000 lm in the case of 60 [degrees]
C. Typical voltage 16.
2v and 1500 mA current, power 24. 3 W. [Bridgelux 2012]
The luminous efficiency of LED arrays and lamps in the case of 60 [degrees]
C, the current is 1.
5 A, calculated as 82. 3 lm/W.
Assuming the ideal luminous efficiency of using this method and 683 lm/W, at least 12% of the power is emitted in the form of light.
The remaining 88% is assumed as heat load and used in the simulation. 5.
2 Laboratory measurement first, the measurement is carried out under laboratory conditions to determine the natural convection heat transfer coefficient of the lamp structure.
6 LEDs are connected with 3 LED arrays in 2 series.
Both series are driven by a laboratory power supply with a constant current of 1. 05 mA.
The power supply is a programmable power supply for TTi CPX200 dual 35V 10A PSU and HAMEG instruments HM8143.
The power, current and voltage of the 2 LED series are measured with the Metrix PX110 watt-hour meter, either at the beginning of the measurement or at the end of the steady state.
The first measurement was made on a transparent aluminum sheet without coating.
Using the Optris LS portable infrared thermometer, it is determined that the specific radiation rate is less than 0. 1.
Unfortunately, in order to define the radiation rate more accurately, the device does not allow the radiation rate value below that value.
The ambient temperature measured for reflection is 24. 7 [degrees]
C, set on the measuring device.
At the steady state measured by the transparent plate, the first series shows a voltage of 44. 8 V, 1. 048 A and 46. 8 W.
The second series of Led displays 44. 9 V, 1. 048 A and 47. 0 W.
The voltage between the steady state at the beginning and end of the measurement is reduced to 1.
There are 5v in both series.
Temperature of Tc-
Measuring the point of the LED array with Meilhaus electronic ME
RedLab data logger using K-
Type of thermocouple. A thermo-
Pasted several wires on top of each LED near Tc
Use point marking of electronic tape.
A thermocouple is also attached at the aluminum plate angle at coordinates 50: 50mm, and a thermocouple is attached near the wall of the measuring room to measure room temperature.
At room temperature [degrees]
C. The average temperature of the two intermediate LED components at steady state is 88. 6 [degrees]
C, and four corner assemblies 85. 0 [degrees]C.
Repeat measurements on white paint-covered plate (
Professor Meston spraypaint White).
The measured specific radiation rate is about 0. 9.
In this case, the power in a stable state is measured at 47. 4 and 47. 5 W.
At room temperature [degrees]
C. The average temperature of the two intermediate LED components is 66. 0 [degrees]
C, and four corner components 64. 1 [degrees]C.
During laboratory measurement, it was noted that increasing the specific radiation rate of the plate would result in a lower temperature, which means that the sample structure can handle higher power inputs when painting.
The current situation of future research has increased to 1.
40 A leads to 65. 4 and 65.
Two Channels 9 W.
Therefore, the simulated LED heat load given in the later part is 0. 88*131. 3 W=115. 5 W. 5.
3 determine the natural convection coefficient using COMSOL Multiphysics 4.
3 simulation program, heat transfer studies were carried out to determine the natural convection coefficient and check the compatibility of the simulation model with the real model
Room temperature 25degrees]
C, heat load 82. 5 W (
88% measure the thermal loss of power 93. 8 W)
On the 6 rectangular shapes, select the size of the sample LED array as the analog input.
The simulation of different convection coefficients was carried out until the simulated model led reached the same temperature as the real temperature
The life model of the measured transparent aluminum plate.
In this case, the specific radiation rate used for heat transfer by thermal radiation is 0. 1.
The same is true for painted shells with a radiation rate of 0. 9.
The convection coefficient is 4. 7 W/[m. sup. 2]
In both cases, K is the closest option, resulting in only 1 to 3 degrees difference between the measurement model and the simulation model. 5.
4 The influence of solar radiation as mentioned earlier, in the case of low radiation, the influence of solar radiation can be ignored
Weight LED lamps used at night only.
To support this, simulations were conducted and 24-hour cycle. A worst-
Case scenarios are studied.
In this case, the outdoor temperature is maintained at 50 [degrees]
C throughout the day, convection is natural using the measured heat transfer coefficient (4. 7 W/[m. sup. 2]K)only.
The projected surface area is not calculated, but it is assumed that the area under the sun is the full top area of the fixture.
The absorption coefficient is assumed to be 1.
At the bottom of the lamp, the heat load is 0.
The solar heat load is applied 15 times.
The ground reflectivity of the black top surface is assumed to be 0. 15 [
Navy Department 1990
, And assume that the reflected heat flow from the ground will be absorbed by the lamps.
There is no heat load applied on the side of the lamp.
As the solar thermal load, the calculation function of the equatorial position during the autumn equinox is used [
Honsberg and Bowden 2012].
The Led turns on at sunset (6 PM)
Cars at sunrise (6 AM).
Solar heat flow function and ladder function of 115.
Figure 5 shows the thermal load of 5w LED8. In Fig.
The temperature of the hottest LED array is introduced.
For example, lamps, you can see that the lamps cool fast enough before turning on the LED, the sun does not need extra heat transfer in the sample structure, the weight is about 1. 7 kg.
The effects of solar radiation will vary in different locations, seasons, and different types of lamps with different shapes and surface features. Also in Fig.
9, another result is the simulation performed on a heavier structure, that is, 0. 4 m x 0. 4 m x 0.
04 m plate weighing about 17 kg.
It can be seen that the response of the lamp to changing the heat load is much slower, and at 24-
Than the original light hour cycle-
Weight example structure.
The temperature is still lower than the temperature in the light.
Weight structure, but the temperature during the day is higher than at night, which can cause potential problems if not considered.
In the structure of heavy-duty lamps, the influence of solar radiation needs to be considered.
It must be noted that the cooling area in the sample fixture is 0. 3264 [m. sup. 2]
, And the heavier simulation example 0. 384 [m. sup. 2].
The value of one launch is 0.
9 was used in the simulation.
The transparent aluminum plate used in the measurement will have lower heat radiation heat transfer to the surrounding environment, but the absorption of the solar heat load on the plate will also be lower.
It is also important to remember that the simulation parameters used are the worst parameters assumedCase parameters.
For these types of research, it is important to know the real outdoor conditions. 5.
5 Comparative simulations were conducted using outdoor data comparison simulations, and the effects of location and outdoor conditions on sample lamps with a heat load of 115 were studied.
5w and radiation rate 0. 9.
The two location sets specified earlier in the study were compared, with a hotter climate position between 30 degrees in Europe and North latitude. Both worst-
Case and average
Based on the simulation. In the worst-
The heat transfer coefficient is 4. 7 W/[m. sup. 2]K was used.
Radiation is considered in all models.
The temperature difference between the shell and the junction is calculated as about 11 [degrees]C.
In Table 3, the simulation results are given.
Simulations were performed using COMSOL Multiphysics 4.
3 modules for heat transfer and fluid flow.
In the flow simulation, place the sample fixture structure in the center of 0. 6 m x 0. 6 m x 0.
6 m box of Air, one side to choose the inflow, the other side to choose the outflow wall. A no-
Sliding conditions are applied on all surfaces of the fixture.
The temperature of the air is the ambient temperature.
The problem is modeled as a coupled heat transfer and fluid flow model where for fluid flow (spf)
Physical model of COMSOL 4. 3 was used.
The addition of the flow analysis will not only cause the Tc temperature to be much lower, but also change the temperature distribution of the lamps and lanterns.
It is important to note the results shown in the figure.
All LED fixtures cannot be summarized as 10 and 11;
When fluid flow is considered, different types of lamps perform differently.
In the case of smaller analog lamps, the spacing of LED components is relatively uniform on the board.
This affects the temperature distribution of the plate and therefore the heat transfer.
It is also important to note that for fluid flow analysis, the characteristic length of the smaller plate is different from that of the larger plate.
The natural convection coefficient of smaller plates may also be slightly different from 4. 7 W/[m. sup. 2]
K is measured for larger plates.
In this simplified example study, these phenomena and their impact on the results have not yet been taken into account.
The simulation results show that the cooling area can be from 0 in this special case. 3264 [m. sup. 2]to 0. 1848 [m. sup. 2]
, Weight from 1. 73 kg to 0. 97 kg.
When designing a specific cooling structure for European conditions, this can be done without sacrificing LED performance, rather than being applicable globally.
Save about 44% of the material and reduce the cooling area by about 43%.
For different types of lamps, the savings of weight and area may vary depending on the structure of the lamps.
These results should not be extended to all outdoor LED fixtures.
6 discuss outdoor convection heat transfer depending on airflow properties and velocity, ambient temperature and surface temperature, and the shape and direction of the surface being cooled.
In the field of fluid dynamics research, a lot of work has been done in the development of calculation and simulation methods for the flow of turbulent fluid on different shapes and oriented surfaces, simulation programs such as COMSOL Multiphysics 4.
Has been improving.
However, these research methods cannot be used correctly without parameter information that affects the results of calculations, simulations or measurements.
The study of wind speed alone is not enough to calculate the convection heat transfer coefficient, but together with the simulation tool and the flow calculation method, the most important is the experimental study, which can be used to better estimate these coefficients and their variations.
The results of this paper show that, on average, the content of outdoor lamps and lanterns at 90% of the operating time is susceptible to wind.
Therefore, the type of convection in practice is usually forced, not free, and the convection coefficient should be calculated or selected accordingly.
The results also show that the maximum wind speed is much higher than the average wind speed.
Occasional high wind speeds are less sensible to consider in cooling designs, but for mechanical designs that are usually synchronized with thermal management, information about high wind speeds is essential.
However, one problem to consider wind data when designing convection cooling is that, especially in urban street canyons, the pattern and speed of wind is affected by the surrounding structure.
Also in the valley, the wind pattern may be different from the plain area.
The survey stations chosen for this study are mostly airport areas and therefore open plains.
In most outdoor fixture cooling studies, cooling by thermal radiation is ignored and there is not much explanation for the reason.
One possible explanation for this is to assume that the ambient temperature is higher than the actual average temperature.
Assume that the sky temperature is the same as the air temperature and give a value of 0 for the specific radiation rate, area and temperature of the surface. 8, 0. 2 [m. sup. 2]and 50 [degrees]
C. Power emitted from the surface by radiation at an average European night temperature of 7 °c. 9 [degrees]
C, will be about 40 W, if calculated (2).
In addition, measurements of the sample structure show that radiation heat transfer is important and cannot be considered negligible in the case of 100 to 200 w led lamps, for example.
The radiation heat exchange between the upper surface of the lamp and the sky or shading structure, as well as the radiation between the lower surface of the lamp and the ground, road surface or other surfaces, is a complex phenomenon and needs further study.
At the time of these phenomena, precipitation and snowfall are expected to have an impact on the cooling properties of lamps and lanterns.
However, the data set of the study lacks these data and, in addition to these parameters, conclusions are not allowed, which is difficult to explain in the cooling design.
However, for the reliability, Life of LED lamps, especially the study of optical properties, the data of precipitation and snowfall should also be collected, and precipitation and snow significantly affect the reflective properties of the road surface and the performance of the overall street lamp.
The Prague data shown here shows that about 5% of the service life of outdoor LED lamps may be spent under \"natural\" water cooling, and the wet surface conditions apply to roads.
Broader data on all of the above factors will help to design intelligent lighting control systems for variable driving conditions.
It is recommended that the cooling design of outdoor lamps is twofold.
Except for the worst-case (
Zero wind, maximum temperature)
The based approach should be considered, especially in the case of LEDs, with a longer lifetime at the maximum allowable junction temperature.
In addition, it is recommended to pay more and more attention to the convection and radiation heat transfer coefficient between the surface of the LED lamp and its environment.
The complexity and variation of these coefficients should also be considered, especially in the case of outdoor lighting systems.
In the past, the general lighting market was dominated by a handful of large international companies.
With the rise of the LED industry, there are now some small and medium-sized enterprises to join in
Large new companies entering the lighting market.
Most LED applications such as street lamps can be considered fixed installation, which means that the life of the lamps is spent in a specific location.
Since smaller companies typically target specific market areas, zone-specific cooling designs can significantly save material costs when targeting cooling areas, and on the other hand, improve reliability and luminous flux, when designing cooling for the hottest places.
In order to optimize and complex LED Lamp cooling methods in practice, it is recommended to focus on the market areas of related lamps when dealing with environmental conditions.
In the global market, the use of different-
Dimension cooling elements for areas with different environmental conditions.
Different companies are already using different types of methods designed to solve the problem of variable environmental conditions.
Whether it is to give different maintenance factors for different areas, or to use intelligent lighting control, or to design different cooling elements for different areas, data on real outdoor conditions are required.
Through regional and optimized cooling design, the thermal management cost of LED lamps can be greatly reduced without significantly affecting the life of lamps.
In addition, if the cooling performance matches the requirements of the installation environment more accurately, the stability of the optical performance of the LED lamps can be better guaranteed in different environments.
7 conclusion the thermal environment of outdoor LED lamps is neither fixed temperature nor zero wind speed environment.
This paper presents outdoor temperature, wind speed, precipitation and snow data for 50 locations and 7 locations worldwide.
The main conclusions are as follows.
* The lowest measured maximum temperature at night is 17. 2[degrees]
C. Maximum 50 [degrees]
C, showing significant changes in the worst case
Housing temperature between different positions.
* The global average night temperature between all locations is 12. 5 [degrees]
C, the average maximum temperature is 31. 9 [degrees]
C. The worst-
Housing and average conditions during lamp life.
* 90% of lamps are used under forced rather than natural convection cooling conditions.
* In addition to convection, heat transfer can be significantly enhanced by considering radiation heat dissipation.
* The cooling design in a specific location has over 40% savings potential in terms of weight and cooling surface area.
* Solar radiation does not pose a serious threat to low radiation
A weight light fixture that responds quickly to changing heat loads. In heavy-
The study of weight structure in relation to time is necessary.
These results provide a knowledge base for the reliability, life, optical properties and road brightness studies of LED lamps for outdoor use at night.
A new method of cooling and lamp manufacturing is proposed in this paper.
If this method is applied in practice, it is likely to bring better reliability, cost savings, lower material consumption, and therefore more eco-friendly outdoor LED lamps.
The work was done under the highlight project of the University of Alto MIDE project, with strong support from Oy MTGMeltron Ltd. [c]
2013 North American Society of Lighting Engineering: 10. 1582/LEUKOS. 2013. 09. 03.
001 reference materials [ASHRAE]
American Society of Heating, Refrigeration and Air
Air conditioning engineer2009.
2009 ASHRAE manual-Fundamentals (SI edition). Atlanta (GA)
: American Society of Heating, Refrigeration and Air
Air conditioning engineer Limited986 p.
Bell SC, Harrison special zone. 2008.
Temperature control in electronic products. INTELEC[R]
2008. IEEE International Conference on telecom energy. INTELEC 2008. 5 p.
Basil Consulting, navigator Consulting, Radcliffe Consulting, SB Consulting2011. Solid-
National Lighting Research and Development: manufacturing roadmap. [Internet]. eere. energy. gov: U. S.
Department of Energy, Office of Energy Efficiency and Renewable energy, construction technology projects, lighting R & D. 2011 [
Reference 2011 10/24.
From: 79 p. Bridgelux. 2012.
DS11 product data sheet for the Bxgelux Es Series [Internet]. Bridgelux. com; [
Reference 2012 11/14.
Available from: Chen Gu, Kwasnick RF, Mielke NR, Monroe EM, Shirley CG. 2006. Ambient use-
A conditional model for reliability assessment.
Proc of IEEE 44 annual conference on Reliability Physics 2006. 299 p.
Falkie F, El Vigi CD. 2001.
The first World Atlas of artificial night sky brightness.
There is no R astronomy system chip (Soc) on Monday ). 328(3):689 -707.
Cole, Martin, Walton, rockmorton. 2010.
Lighting in extreme temperature environments: high and low.
Record of PCIC Europe 2010 meeting. 1-8.
Mr. cowsley, MP Garcia, Grzesik LK, Webster J, marunji Jordan. 1995.
Heat dissipation development of modular outdoor cabinets.
1995 International Conference on telecom energy. INTELEC \'95. 384 p.
Navy, Naval Facilities Engineering Command.
Military Manual (MILHDBK)
1011/2. Natural ventilation 【Internet]. 1990. www. civil. waterloo. ca; [Reference 2012 12/4.
Available from Eicker U, Dalibard A. 2011. Photovoltaic-
A heat collector for radiation cooling at night in a building. Sol Energy. 85(7):1322-1335. Estes R.
Thermal management of telecom cabinets 【Internet]. Coolingzone. com. 2005 [
Reference 2011 10/20.
Available from: Bowden S. PVCDROM [Internet]. pveducation. org; [
Reference 2012 11/28.
From: Yang Kai, Chen Yuqing, Lin Bin, building, June. 2011.
Heat dissipation design based on heat siphon radiator of high power LED street lamp.
2011 International Conference on electronics and optoelectronic (ICEOE). 374 p.
Yang Kai, Chen Yuqing, Lin Bin, Song Jidong. 2010.
Thermal analysis and design
LED street lamp power supply.
2010 seminar on photon and photoelectric (SOPO).
Kai Keding, Feng Shiwei, Guo Chunsheng, Zhang Guangchen and Chen Gao. 2011.
Structure Optimization of high-power LED street lamp radiator.
2011 International Conference on electronics and optoelectronic (ICEOE). 2:75-78.
Luo x, bear W, Cheng t, Liu S. 2009.
Temperature Estimation at high temperature
LED street light through multiple power supplies
Chip analysis solution.
Let the photoelectric element. 3(5):225-232.
Luo X, Cheng T, Xiong W, gan Z, Liu S. 2007.
Thermal analysis of 80w lamps
LED street light.
Let the photoelectric element. 1(5):191-196. Marongiu MJ. 2006.
Cooling outdoor housing]Internet]. process-cooling. com. [
Reference 2011 10/20.
Available from: Marosy G, Kova cs Z, Molna r G, Poppe A. 2010.
Diagnosis of LED
Street lighting based on thermal transient method.
Minutes of the 16 international symposium on ICs and systematic hot investigations. THERMINIC 2010. October 2010. Barcelona (Spain). 18 p. McKay JR. 1988.
Effects of solar radiation and wind speed on air temperature rise in outdoor cabinets containing telephone equipment.
Minutes of 1988 International Conference on telecom energy. INTELEC \'88. 285 p. [NOAA]
National Oceanic and Atmospheric Administration. 2008.
Comprehensive ground hourly observation 2007 [Data DVD]. U. S.
Ministry of Commerce. [NOAA]
National Oceanic and Atmospheric Administration. 2009.
Comprehensive ground hourly observation 2008 [Data DVD]. U. S.
Ministry of Commerce. [NOAA]
National Oceanic and Atmospheric Administration. 2010.
Comprehensive ground hourly observation 2009 [Data DVD]. U. S.
Ministry of Commerce. [SESKO ry]. 2007. SFS6000 -7-
714 Part 7-low voltage electrical device
714: Requirements for special facilities or locations, external lighting facilities.
Lock the door standardoimilito SFS Ry. 4p.
Still M, Venzke H, Durst F, Melling A. 1998.
Influence of humidity on convection heat transfer of small cylinders. Exp Fluids. 24:141-150. [USNO]
US Naval Observatory
A table where the sun or moon rises/falls for a year: the position of the world [Internet]. usno. navy. mil; [
Reference 2010 12/27.
It can be obtained from: Wang Kai, Mao Zhangming, Liu Zongyuan, Chen Fei, Chen Hao, Luo Xiaobing, Liu Sheng and other places. 2010.
An application-specific LED Package module integrating optical, cooling, and power supply functions.
Proc, 2010 conference on electronic components and technologyECTC 2010. 686 p.
Luo Xiaobing, Xiong Wei, Liu Sheng. 2008.
Simplified thermal resistance network model for high-power LED street lamps.
International Conference on Electronic packaging technology and high density packaging Proc 2008. ICEPT-HDP 2008. 1 p.
Luo Xiaobing, Xiong Wei, Cheng Ting, Liu Sheng. 2009.
Level design and optimization
Positioning plate wing radiator for high-power LED street lamps.
Proc, 2009 conference on electronic components and technology. ECTC 2009. 854 p.
Jaana Jahkonen (1)
*, Marjukka Puolakka DSc (1)
And Liisa Halonen DSc (1)(1. )
University of Alto, Australia, electrical engineering, lighting, otakari B . 7, Espoo 02150, Finland.
* Newsletter author: Jaana Jahkonen, E-mail: jaana. jahkonen@aalto.