|The development of light-emitting diode (LED) technology has resulted in widespread solid state lighting use in both consumer and industrial applications due to their high-energy efficiency, long expected life, and less toxic than compact fluorescent lighting.
Previous research shows that there are at least two predominate degradation mechanisms. One mechanism is the degradation of the LED chip due to the increase of non-radiative recombination which reduces the total number of emitted photons. The other predominate mechanism is the degradation of the accumulated optical parts from temperature or electric stresses. Both failure mechanisms could cause the reduction of luminous flux output. Also, degraded optics work as a light spectrum filter and will rescale the intensity of light at each wavelength, which will cause the color coordinates to shift. Besides, increase of non-radiative recombination of electron and holes will create more heat in the die region that will worsen the degradation of optical parts of the LED package. Both failure mechanisms could cause the reduction of luminous flux output and color shift of LED packages.
Currently, TM 80 is the established test method for LED packages to access the quality, reliability and durability before being introduced into to the customers. However, TM80 test only concerns the temperature stress form the operation environment. In LED-based luminaire applications, some are operated in the harsh environment. During the operation, the LED package not only experiences the high bias current and ambient temperature, but also the humidity around the package, especially for the road lighting and automobile lighting. There is literature gap for the accelerated test methods for high power LED packages that includes both the thermal and humidity effects from the ambient environment. Currently, TM21 long term lumen maintenance projecting method are used to extend the test data to the desired time to predict the lumen maintenance time. One principal issue of TM21 is that it is derived based on the Arrhenius equation and is lack of additional stresses to characterize the non-temperature dependent failure mechanism, such as humidity. Therefore, there exists a need of physics based method to estimate the life time of LED packages under certain operation condition which includes both ambient temperature and humidity.
The disadvantage and deficiencies in TM80 and LM21 validates the needs to develop new acceleration test method and estimation techniques to quantify the reliability of high power LED package under a variety of operation conditions. In this paper, experiments are set up to explore the different effects of the humidity and the thermal stress on various LED packages. Commercial available warm white (3000K) and cool white high power LED packages are used to study the degradations caused by temperature and humidity. The test results in this work increases the understating the reliability of LED packages through the investigation of failure mechanism under various acceleration test conditions. Besides, the prognostic health management method developed in this work increases the accuracy of estimation of remaining useful life of high power LED package. These methods will greatly reduce the time and effort needed to estimate the lumen maintenance time of LED package during the application.