Analysis of cooling system of LED road lighting fixtures

Preface

In China, lighting accounts for about 13% of the total electricity consumption. Urban public lighting, especially road lighting, accounts for a huge share of the entire lighting field. In recent years, with the national strategy of building a conservation-oriented society, as a new type of energy-saving light source, LED has become an inevitable trend to enter the road lighting field based on the continuous improvement of its production technology and performance.

1. The relationship between LED junction temperature and performance

There are two main bottlenecks that limit the development of LED road lighting fixtures: one is cost. At present, the price of LED light sources is still relatively high compared to traditional light sources. However, with the advancement of production technology and the extensive participation of domestic enterprises, the cost of LED will gradually decrease. Another bottleneck is heat dissipation. Since LEDs are semiconductor light -emitting devices , the characteristics of semiconductor devices will change significantly as their own temperature changes. For LEDs, the increase in junction temperature will cause changes and attenuation in various aspects of the device's performance. This change is mainly reflected in three aspects:

(1) Reduce the external quantum efficiency of LED;

(2) Shorten the life of LED;

(3) The main wavelength of the light emitted by the LED is shifted, which causes the color of the light source to shift.

Among them, the external quantum efficiency of the device is a quantity directly related to the LED light efficiency . The decrease in external quantum efficiency will directly lead to a decrease in the LED light efficiency. As the temperature of the device PN junction increases, the main wavelength of the white light LED chip will move toward the long-wave direction . Statistics show that at a temperature of 100°C, the wavelength can be red-shifted by 4 to 9nm, resulting in a decrease in the absorption rate of the YAG phosphor and a decrease in the LED output luminous flux . Figure 1 shows the relative light output curves of the X2Lamps series products (a) produced by CREE and the LUXEONK2 series products (b) produced by LUMILEDS as a function of junction temperature. Both products are high- power white light LEDs with relatively mature technology. From Figure 1, it can be seen that when the junction temperature of the device rises to 65°C, the light output of the two will decrease by about 10% relative to room temperature.

Figure 1: LED relative light output versus junction temperature (click on image to enlarge).

For a single LED, if the heat is concentrated in a small chip and cannot be effectively dissipated, the chip temperature will rise, causing uneven distribution of thermal stress and an increase in the chip failure rate. Studies have shown that when the temperature exceeds a certain value, the failure rate of the LED will rise exponentially, and the reliability will decrease by 10% for every 2°C increase in the LED temperature.

Figure 2 shows the relationship between junction temperature and life of the LUXEONK2 series products produced by LUMILEDS. As can be seen from the figure, when the junction temperature rises, the light decay of the LED will be significantly accelerated and the life will be significantly reduced. Parameters such as light efficiency, color temperature , and life are all very important indicators in lighting applications. Therefore, how to control the temperature rise of LED lamps through reasonable heat dissipation design is an urgent problem that needs to be solved when LED enters the field of road lighting.

Figure 2: LED junction temperature and lifespan curve (click on the image to enlarge).

2. Thermal analysis model of street lamps

To analyze the thermal conduction model of street lamps, we must first select the temperature benchmark, which is the temperature index we want to analyze and control in the end. Most of the temperature rise indicators given by many LED lamp products on the market are based on the difference between the temperature of the lamp housing and the ambient temperature. There are certain limitations in using this method to measure the heat dissipation performance of lamps. Because the performance of LED devices is directly related to the junction temperature of its PN junction, the final indicator we care about is the junction temperature. Different lamps are very different in terms of LED light source selection, lamp material use, production process, and heat dissipation design, resulting in a large difference in thermal resistance from the LED PN junction to the lamp housing. In this case, it is unscientific to use the temperature rise of the lamp housing relative to the ambient temperature to judge the heat dissipation performance of the lamp. We must use certain methods to measure the junction temperature of the chip, and then measure the quality of the lamp heat dissipation design by the change in junction temperature.

Taking into account the various requirements of road lighting, as well as the requirements of street lamps in terms of light distribution, mechanics, protection level, etc., there are currently two main heat dissipation methods used in LED road lighting fixtures: the first is passive heat dissipation, that is, heat dissipation by installing heat sinks. This method has a simple structure, but the heat dissipation efficiency is relatively low; the second is active heat dissipation, that is, heat dissipation by adding external fans or water cooling. This method has a relatively high heat dissipation efficiency, but requires additional power consumption, which will reduce system efficiency and is very difficult to design.

These two heat dissipation methods have their own advantages and disadvantages, and the structure of the lamp products produced is very different. However, considering all aspects of lamp design and the characteristics of LED itself, the core structure of LED lamps is basically the same. It uses a 1W LED to be welded on an aluminum PCB board, and then the PCB board is fixed on the lamp housing through a suitable distribution method, and then the heat is dissipated by the housing. Therefore, the flow of heat can be simply summarized as the following process: first, the heat is transferred to the aluminum substrate that fixes the LED through the welding layer, and then the heat conductive adhesive of the aluminum substrate transfers the heat to the lamp housing, and then the heat is transferred to each heat sink through the lamp housing, and finally the heat is dissipated by convection between the heat sink and the air. The whole process can be represented by the equivalent thermal resistance model shown in Figure 3.

Figure 3: LED street light Schematic diagram of the equivalent thermal resistance model of the lamp (click on the image to enlarge).

There are three main ways of heat transfer: heat conduction, heat convection and heat radiation. From the diagram analysis, we can see that LED street lamps mainly use conduction and convection to dissipate heat during the heat dissipation process. Therefore, the thermal resistance of the lamp thermal system can be mainly divided into two types: conduction thermal resistance and convection thermal resistance. For conduction thermal resistance, its resistance value is R=ΔxPKA; for convection thermal resistance, its resistance value is R=1PhA. Among them: K is the thermal conductivity of the material, h is the convection coefficient, and A is the thermal conductivity area.

As shown in Figure 3, if you want to quickly transfer the heat generated by the LED to the atmosphere, the conduction thermal resistance that needs to be passed includes the soldering thermal resistance between the LED device and the soldering layer of the aluminum-based PCB, the thermal resistance of the aluminum-based PCB, the thermal resistance of the thermal conductive silicone layer between the aluminum-based PCB and the lamp housing, and the thermal resistance of the lamp housing. Since the welding and silicone coating processes may produce bubbles, introducing air thermal resistance, and such bubbles are surrounded by the soldering layer and the silicone, it is difficult to communicate with the outside gas, so this part of the thermal resistance can be regarded as the conduction thermal resistance of air conduction heat, rather than the convection thermal resistance.

From the previous formula, we can see that to reduce the thermal resistance of each part, the thickness of the material should be reduced as much as possible when the material is selected. The thermal conductivity of aluminum is K=202WPm& mid dot;℃, the thermal conductivity of nickel is 93WPm·℃, and the thermal conductivity of air is 01024WPm·℃. Therefore, in order to reduce the thermal resistance of the connection layer, it is necessary to reduce the bubbles in the welding layer and the silicone layer as much as possible. The following combines two specific heat dissipation methods to analyze the heat dissipation design of the lamp.

3. Passive cooling

Passive heat dissipation relies on the natural convection between the outer surface of the lamp and the air to dissipate the heat generated by the LED. This heat dissipation method is simple in design and can be easily combined with the mechanical structure design of the lamp, which makes it easier to meet the protection level requirements of the lamp and has a low cost. Therefore, it is currently the most widely used heat dissipation method. However, this heat dissipation method also has disadvantages, that is, the heat dissipation efficiency is not high, and the designed lamp is too heavy because of the large number of heat sinks. At the same time, due to the presence of the heat sink, the lamp shell is more likely to accumulate dust, which will reduce the maintenance factor of the lamp.

In order to scientifically analyze the heat dissipation efficiency of this heat dissipation method, we selected several LED road lighting lamps that adopt this heat dissipation method. After each sample was ignited for 3 hours (to ensure that the lamp has reached thermal equilibrium) in the same closed room (to ensure the same air convection coefficient), its junction temperature was measured. In order to eliminate the influence of room temperature fluctuations, we used relative temperature in the analysis, that is, the difference between the steady-state junction temperature and the room temperature to measure the heat dissipation effect of the lamp. The specific data is shown in Table 1.

Table 1: Comparison of heat dissipation effects of lamps using the same heat dissipation method

To measure the heat dissipation efficiency of the lamp, the input power and mechanical parameters of the lamp are also listed in Table 1. The heat sink area is the effective outer surface area of ​​the entire lamp (excluding the area of ​​the front glass). The heat dissipation area corresponding to unit power consumption is obtained by dividing the heat sink area by the input power of the lamp. This indicator is used to measure the heat dissipation efficiency of the lamp. In order to more intuitively reflect the relationship between the heat dissipation area corresponding to unit power consumption and the change in the junction temperature of the lamp, that is, the heat dissipation effect of the lamp, we make a scatter plot of these two indicators, as shown in Figure 4.

Figure 4: Relationship between heat dissipation area per unit power consumption and heat dissipation effect of lamps

By analyzing the data in Table 1 and combining it with Figure 4, it can be found that the larger the heat dissipation area, the smaller the junction temperature rise of the device will be, but the two are not in a simple proportional relationship, which shows that the efficiency of the heat sinks of each sample is not the same. In particular, the heat dissipation area corresponding to the unit power consumption of sample No. 3 is similar to that of sample No. 4, but its LED junction temperature rise is about 5°C higher than that of sample No. 4, which proves that its heat sink efficiency is much worse than that of sample No. 4. Analysis of the heat dissipation structure of sample No. 3 shows that the height of its heat sink is 15mm, but the spacing between its heat sinks is only 4mm, that is, relative to the height of its heat sink, the distance between the two heat sinks is very small, forming a very narrow groove, and the air in this groove is difficult to circulate with the outside air, which will affect the convection coefficient of the outer surface of the lamp, thereby affecting the overall heat dissipation efficiency of the lamp.

4. Active cooling

Active heat dissipation mainly increases the air flow speed on the surface of the radiator through water cooling, fans, etc., so as to quickly take away the heat on the heat sink, thereby improving the heat dissipation efficiency. Specifically, in the equivalent model of this article, the convection coefficient h is increased to reduce the thermal convection internal resistance between the heat sink and the atmosphere, thereby reducing the PN junction temperature rise of the LED chip . For LED road lighting fixtures, considering the use environment and protection requirements of road lighting fixtures, fans are generally used as a means to increase convection.

In order to prove the heat dissipation capacity of the fan, the following test was conducted: two lamp samples with relatively high energy consumption density were selected, and various temperature parameters including the junction temperature of the LED chip and the temperature of the heat sink were measured with and without fans. Considering that road lighting lamps are usually IP65 protective lamps, and fans require a higher waterproof and dustproof level, it is expected that the fans in the LED lighting lamp products that will be marketed in the future should be enclosed in the lamps. Therefore, when measuring the samples, we placed the lamps equipped with fans in a closed metal cavity. When collecting experimental data, in addition to collecting the junction temperature of the LED chip in the lamp, we also used two temperature probes to read the temperature of the lamp radiator and the metal cavity shell for reference. The specific test data is shown in Table 2.

Table 2: Temperature of various parts of the sample with and without fan

By comparing the test data listed in Table 2, it can be found that after adding the fan, the temperature of the LED chips of the two samples dropped by 3217℃ and 3613℃ respectively. Compared with the case without the fan, the junction temperature has dropped significantly, which shows that adding the fan has indeed improved the heat dissipation performance of the lamp to a great extent. From Table 2, it can be found that the temperature has dropped significantly mainly in the junction temperature of the LED chip and the temperature of the lamp heat sink. The temperature of the metal cavity surrounding the lamp has not changed significantly after adding the fan. This is because after adding the fan, it only increases the air flow speed on the surface of the radiator, increases the convection coefficient, and thus reduces the convection thermal resistance. However, the heat generated by the LED chip must be dissipated through the closed metal cavity shell, and the heat generated by the lamp is certain, so the heat dissipated through the outer surface of the metal cavity is also certain. No matter how we change the fan speed inside, the temperature change on the surface of the metal cavity should not be large.

5. Conclusion

This article selects some samples for the two common heat dissipation modes of LED road lighting fixtures and conducts systematic and scientific tests on the temperature of each part. Through the analysis of test data and combined with the thermal analysis model of LED lamps, various factors affecting the heat dissipation efficiency of lamps are analyzed in detail, thus providing guidance for the design of LED lamps with high heat dissipation efficiency, which is of great significance to promoting the application of LED in road lighting.