Fluid–solid coupling thermo-mechanical analysis of high power LED package during thermal shock testing

https://doi.org/10.1016/j.microrel.2012.03.021Get rights and content

Abstract

The virtual design by numerical simulation to model various accelerated reliability testing conditions is adopted to validate and improve the reliability of the high power LED package. In this study, the reliability of the high power LED package during thermal shock testing is investigated by fluid–solid coupling thermo-mechanical modeling by considering nonlinear time and temperature dependent material properties. Through fluid–solid coupling transient thermal transfer analysis, it is found that the maximum thermal gradient exceeds 75 K during the rapid cooling process and 91 K during the rapid heating process of the thermal shock testing which is ignored in the traditional isothermal assumption. The calculation results indicate that the equivalent plastic strain range of the bonding wire within the LED package with consideration of the temperature gradient is much higher than that with the isothermal assumption. The assumption of the isothermal condition is not appropriate which will lead to overestimation of the predicted lifetime. The viscoelastic behaviors of the silicone have significant influences on the lifetime prediction of the bonding wire and silicone with low elastic modulus and coefficient of thermal expansion (CTE) can significantly enhance the reliability of the bonding wire under the thermal shock loading. The results in this study could provide a guideline on design for reliability in the high power LED packaging.

Introduction

Solid state lighting in terms of high power LEDs will become the fourth generation illumination source due to its superior characteristics such as high efficiency, low power consumption, high reliability and long life [1], [2]. Currently, the high power LEDs have found the applications in general illumination, automotive front lighting, backlighting for LCD displays and city improvement engineering [3], [4], [5]. The theoretical viewpoint is that LEDs can work as long as 100,000 h under the perfect operating conditions. However, the material degradation and structural defects due to the electrical, thermal, mechanical and chemical stresses will lead to the lumen degradation, color variation or even early catastrophic failure of the LED modules [6], [7], [8], [9]. Therefore, the reliability of the high power LED package is becoming an urgent issue for the emerging illumination applications.

Due to large mismatch of CTE between different materials used in the LED package, significant thermo-mechanical stresses will be induced under thermal loading which might lead to failures such as cracks in the materials, delaminations at the different interfaces, and facture failures of the boding wire [8], [9], [10]. Therefore, thermal shock testing from 233 K to 398 K was adopted to evaluate the capability of the LED package to withstand the thermal loading by many vendors.

The LED package structure, in particular the bonding wire, suffers significant thermo-mechanical stresses and strains during the thermal shock testing. The deep insights of the potential failure modes of the bonding wire may be obtained from the accurate nonlinear finite element simulation. An approach which combines temperature cycle failure data and finite element analysis results to estimate bonding wire fatigue parameters and mean time to failure was developed by Chidambaram [11]. Thermo-mechanical modeling was conducted to simulate the thermal cycling induced stresses and strains in the bonding wire of the chip on broad (COB) package by Pang et al. [12]. Different finite element techniques including contact algorithm, 2D approach, and 3D global local modeling were employed to study the thermo-mechanical behavior of the bonding wire embedded in the electronic package by van Driel et al. [13].

In addition, considering that significant thermal gradient will be induced to the LED package during thermal shock testing, the uniform temperature assumption may not be appropriate. The thermal gradient will induce additional stresses and strains which can significantly shorten the fatigue lifetime of the bonding wire. The influences of thermal gradient on the reliability performance of electronic package have been considered by some studies. The thermo-mechanical responses of a ceramic ball grid array (CBGA) package under convective condition with different assumptions such as isothermal, constant heat transfer coefficient and local heat transfer coefficient were studied by Hong and Yuan [14]. An integrated CFD thermo-mechanical analysis was performed to evaluate the transient effects of thermal shock on the reliability of the PBGA package by Mercado et al. [15]. In their study, the sequentially coupled heat transfer and thermal stress analysis was performed by applying the heat transfer coefficients boundary condition which was obtained from the CFD calculation. And it was demonstrated that the assumption of uniform temperature distribution was not sufficient for determining the critical stresses leading to die cracking. However, few works have been reported on the fluid–solid coupling thermo-mechanical analysis of the electronic package reliability during thermal shock testing which can avoid conducting heat transfer simulation by setting local time dependent heat transfer coefficients.

In this paper, a fluid–solid coupling thermo-mechanical modeling by using nonlinear time and temperature dependent material properties was established to investigate the reliability of the high power LED package during thermal shock testing. The viscoelastic behaviors of the silicone were characterized by dynamical mechanical analysis (DMA) method and expressed by Prony series based on the Maxwell model. The temperature dependent CTE of the silicone were determined by the thermal mechanical analysis (TMA) testing. The sharply temperature change and the significant thermal gradient were obtained by transient thermal transfer analysis with the fluid–solid coupling method. The stress and strain behaviors of the LED package, especially the bonding wire, were analyzed through thermo-mechanical calculation results. The effects of thermal gradient within the LED package during the thermal shock testing were evaluated comparing to the results from the thermo-mechanical analysis with isothermal assumption. The effects of silicone viscoelastic behaviors on the stress and strain responses of the bonding wire were also estimated by comparing to the results from the thermo-mechanical analysis with only temperature dependent elastic material property. The effects of the elastic modulus and CTE of the silicone on the reliability performance of boding wire were also analyzed.

Section snippets

Model and boundary conditions

The geometry of the high power LED package which consists of die paddle, die attach, LED die, lead frame, molding compound (MC), silicone, polycarbonate (PC) lens and the bonding wire is shown in Fig. 1. The dimensions of the LED package are also given in Fig. 1. The dimension the LED die is 1 mm × 1 mm × 150 μm and the thickness of the die attach is 50 μm. The gold bonding wire which the diameter is 1 mil connected between the LED die and the lead frame can also be seen in Fig. 1.

The thermal shock

Thermal gradient during thermal shock testing

Fig. 6, Fig. 7 present the local temperature evolutions at different positions of the LED package during the rapid cooling and heating processes of the thermal shock testing respectively. It can be found that the temperature within the LED package changes sharply at the beginning of the thermal shock. The thermal gradient decreases with time and the temperature distribution within the LED package becomes isothermal after 100 s. Fig. 8 shows the temperature distributions of the LED package at the

Conclusions

In this study, a fluid–solid coupling thermo-mechanical modeling with nonlinear time and temperature dependent material properties was established to investigate the reliability of the high power LED package during thermal shock testing. Summary of some significant results is listed as follows:

  • (1)

    The accurate time dependent temperature distributions and significant thermal gradient within the LED package were determined by the fluid–solid coupling transient thermal transfer simulation of the

Acknowledgments

This work is supported in part by Key Project of National Natural Science Foundation of China (No. 50835005), and in part by National Key Technology R&D Program of China (No. 2011BAE01B14). The technical discussions from Mao Zhangming are gratefully appreciated.

References (21)

  • S. Liu et al.

    LED packaging for lighting application: design, manufacturing and testing

    (2011)
  • Z.Y. Liu et al.

    Status and prospects for phosphor based white LED packaging

    Front Optoelectron China

    (2009)
  • K. Wang et al.

    Freeform LED lens for rectangularly prescribed illumination

    J Opt A: Pure Appl Opt

    (2009)
  • K. Wang et al.

    Design of compact freeform lens for application specific light emitting diode packaging

    Opt Express

    (2010)
  • K. Wang et al.

    Optical analysis of an 80 W light emitting diode street lamp

    Opt Eng

    (2008)
  • L. Trevisanello et al.

    Accelerated life test of high brightness light emitting diodes

    IEEE Trans Dev Mater Reliab

    (2008)
  • Z.H. Chen et al.

    Comparison of LED package reliability under thermal cycling and thermal shock conditions by experimental testing and finite element simulation

    Electron Compon Technol Conf

    (2011)
  • Z.H. Chen et al.

    Reliability test and failure analysis of high power LED packages

    J Semicond

    (2011)
  • L.X. Tan et al.

    Effects of defects on the thermal and optical performance of high brightness light emitting diodes

    IEEE Transac Electron Packaging Manuf

    (2009)
  • S. Liu et al.

    Modeling and simulation for microelectronics package assembly: manufacturing

    (2011)
There are more references available in the full text version of this article.

Cited by (16)

  • A new algorithm for solving an inverse transient heat conduction problem by dividing a complex domain into parts

    2019, International Journal of Heat and Mass Transfer
    Citation Excerpt :

    The thermomechanical analysis plays an important role in the design and operation of equipment operating at high temperature differences. Such analyses are carried out in microelectronics [1], construction [2], power industry [3] and in research on nuclear fusion [4]. The fundamental problem in this type of investigations is the difficulty in determining some of the thermal boundary conditions.

  • Experimental investigation of a PCM-HP heat sink on its thermal performance and anti-thermal-shock capacity for high-power LEDs

    2016, Applied Thermal Engineering
    Citation Excerpt :

    It will then be quickly cooled by a cooling module with the temperature rapidly dropping under the nonworking or low-brightness condition. This phenomenon causes LED devices to be constantly attacked by periodic thermal stress, inducing package material degradation and structural defects [8,9,21,22]. Zhang and Lee [9] mentioned that in the LED package, gold bonding wires were surrounded by a silicone encapsulant, which had a higher thermal coefficient of expansion than the gold wire.

  • Influence of phosphor amount on microstructure and damage evolution of silicone/phosphor composite in light-emitting diodes packaging

    2015, Composites Science and Technology
    Citation Excerpt :

    This stress may result in damage of the composite and delamination of material interfaces. As cracks and voids nucleate within silicone/phosphor layer and delamination initiates around material interfaces, air and moisture could penetrate the silicone encapsulant, which subsequently forms silicone-air-phosphor and silicone-air-chip interfacial region [22–24]. The generation of these regions can lead to lumen degradation and material fracture, which eventually limits the performance and reliability of LED packaging [25,26].

  • Modeling and simulation of power electronic modules with microchannel coolers for thermo-mechanical performance

    2014, Microelectronics Reliability
    Citation Excerpt :

    As heat transfer path is mainly from the top chips to the bottom baseplate, assuming that except for the convection boundary conditions inside the microchannel walls and the bottom surface of baseplate exposed to the air, and as the thermal conductivity of silicone gel covered on the chips and wire was only 0.8 W m−1 K−1, in order to simply the calculation, it was assumed to be insulating. We would use the calculated convection coefficients to express heat transfer of flowing fluid to take into account the fluid coupling effect [15]. Material properties and thickness of each layer of the model were listed in Table 1.

View all citing articles on Scopus
View full text