Junction Temperature Prediction for LED Luminaires Based on a Subsystem-Separated Thermal Modeling Method

Miao Cai*, Zhi Liang, Kun Miao Tian, Ming Hui Yun, Ping Zhang, Dao Guo Yang*, Guo Qi Zhang

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

9   Link opens in a new tab Citations (SciVal)
44 Downloads (Pure)

Abstract

The junction temperature (Tj) of LEDs in an LED luminaire is useful for projecting the luminous flux maintenance or lifetime of luminaires. Normally, a LED luminaire has a unique and complex geometric outline, and its LED package size is only a few millimeters. The thermal modeling procedure with one luminaire model is complex. Achieving a precise Tj of luminaires quickly in modeling is still a great challenge. In this work, aiming at providing a simple modeling method to predict a precise luminaire Tj, a subsystem-separated thermal modeling method for luminaires is proposed by following the existing form of multi-domain modeling. The detailed package model is linked with the luminaire model, while the average temperature at the bottom surface of the LED package is associated with an assumed equivalent convective heat transfer coefficient (ECHTC) at the package model. An adequately assumed ECHTC is obtained, while a unique average temperature is achieved from the luminaire model. Results show that a precise luminaire Tj is projected rapidly by modeling the package model with the adequate ECHTC and a proposed power law equation. The proposed modeling method not only provides a precise Tj prediction with a small error but also effectively simplifies the modeling of LED luminaires.

Original languageEnglish
Article number8809692
Pages (from-to)119755-119764
Number of pages10
JournalIEEE Access
Volume7
DOIs
Publication statusPublished - 2019

Keywords

  • junction temperature
  • LED luminaires
  • prediction
  • subsystem
  • thermal modeling

Fingerprint

Dive into the research topics of 'Junction Temperature Prediction for LED Luminaires Based on a Subsystem-Separated Thermal Modeling Method'. Together they form a unique fingerprint.

Cite this