Research On Junction Temperature Characterization And Fast Life Predication Of LED Lamps Based On A Hybrid Method

LEDs are getting strong support by from the governments around the world due to its energy conservation and environmental protection, long life and other significant advantages. Europe, the United States and other developed countries treat LEDs as a newly strategic industry. The LED lighting is gradually replacing traditional lighting technology, which begins to dominate the development direction of lighting market. However, LEDs as a new technology, due to its’ complexity characteristics, the thermal management and reliability problems of LEDs products become an important factor that restricts the development of LED industry. In this paper, LED lighting lamps are chosen as the research object, and the computational fluid dynamics simulation method and step accelerated degradation test based on performance degradation are utilized. Firstly, the thermoelectric coupling model of LED lamps is established, and the junction temperature of LED lamps is obtained more accurately. Then, the reliability of the LED lamps is evaluated based on the thermal simulation and Monte Carlo method. At the same time, based on the basic theory of reliability test and related mathematical model, using Matlab software as development environment, a reliability assessment application software platform for the LED lamps is developed. The main research contents of this thesis are as follows:(1) Focus on the uncertainty problem of the boundary conditions in the current LED thermal analysis, in order to make the modeling closer to actual situation, the traditional thermoelectric coupling model for a LED lamps and the thermoelectric coupling model for its’ encapsulation is set up respectively. Then, with the aid of Icepak software and mathematical interpolation processing, the junction temperature simulation for LED lamps is conducted. The simulation result is verified through a validation experiment.(2) Based on the thermoelectric coupling stimulation, the junction temperature distribution of LED package inside the LED lamp is studied. A fast life prediction method of LED lamps is proposed based on the relationship between LED junction temperature with its lifetime, and the Monte Carlo simulation method. The reliability of light source and other components of LED lamps is discussed, and the system reliability of the LED lamps is studied. Then, the fast life prediction method of LED lamps is realized.(3) Life prediction method of LED lamps based on reliability test is established. The Matlab7.0 software being used as development platform, LED reliability assessment algorithm is programmed and, the system panel development is completed. Completed step acceleration test, the preliminary application of this system is successfully in the process of the evaluation for the reliability of LED lamps.The final results are as follow: 1) Two-step thermoelectric modeling method for junction temperature prediction of LED lamps not only effectively overcome the uncertainty problems of the boundary conditions in the traditional finite element analysis, but also can obtain more accurate results than the traditional CFD analysis method. At the same time, the process of simulation analysis is simplified. In the process of LED thermal analysis, the thermoelectric properties of LEDs are considered, which make analysis result closer to actual situation. 2) Based on the thermal analysis, the statistical analysis method is applied, the junction temperature distribution of LED device inside the LED lamps is obtained. The process of the LED production is randomness which effect on product quality lead to the error of the simulation and experiment is considered comprehensively. The randomness error is considered in the process of the simulation. Using the Monte Carlo simulation method to obtain the life distribution of the LED light source, and the LED lamps life prediction is completed. This method is more fast and efficient than test method and it can be used for the reliability design of the LED lamps. 3) In the end of the thesis, a reliability test of LED lamps is carried out based on the performance degradation step accelerated degradation, and tool software for LED reliability test data processing is developed. The software is simple and convenient to be used. It can satisfy the general requirements of reliability test, and improve the work efficiency of reliability evaluation of LED lamps. It shows an application value.