A hybrid prediction method on luminous flux maintenance of high-power LED lamps

A hybrid method, applied thermal modeling and temperature measurement, is proposed to estimate the junction temperature (Tj) of high-power light-emitting diodes (LEDs) at system level, and further project long term lumen maintenance of LED lamps. First, 3D finite element modeling on commercial LED lamps is performed based on lamp structure and material information to find a steady and reliable relationship between the junction point and a referenced point on the heat sink. Then, accurate thermal measurement is conducted on the heat sink to calibrate and verify the finite element model. The predicted Tj of LEDs from modeling, in conjunction with LM-80 luminous maintenance data using TM-21, is applied to project the luminous flux depreciation at the system level. The proposed approach is validated by aging tests at room ambience. Results show the thermal resistance modeling can be simplified into a one-dimensional model when a LED lamp operates in a steady situation. Thus, the Tj of LED lamps operating at any specified ambient temperature can be achieved quickly. The estimation method for predicting luminous maintenance of LED lamps is efficient and fast. The proposed method is expected to be useful for the fast qualification of LED lamps. © 2015 Elsevier Ltd. All rights reserved.