Termal, elektriksel ve optik performans için yag:ce fosfor ile birleştirilmiş sıvı soğutmalı mavi LED'in deneysel ve nümerik olarak incelenmesi

At the new age of illumination, light emitting diodes (LEDs) have been proven to be the most efficient alternative to conventional light sources. Yet, in comparison to the other lighting systems, LEDs can withstand a lower maximum operating temperature while junction temperature (Tj) is among the main factors dictating their lifespan, reliability, and luminance performance. Thermal concerns in high power LEDs are important not only because of the dramatic failure of the chip, but also lowered performance in terms of light characteristics. Generated heat and increased device temperatures are direct indicators of poor performance. In fact, heat generation in LED packages occurs mainly on the chip and phosphor layers, both of which lead to a decrease in light output due to the reduction in internal and external quantum efficiencies. In addition to the heat losses over an LED die due to the losses in combination of carrier charges, optical losses are also significant. Moreover, with the increase in Tj during the operation of an LED, the transparency of optical components in the package is disrupted and leads to absorption losses. This eventually results in degradation of components in the package and performance losses. However, the highest absorption losses are observed in the LED die due to internal reflections of a portion of emitted light over the chip surface and eventually loss as heat. Although phosphor coating over LED die is a common practice, absorption losses and increased temperature of a phosphor coated layer over the die negatively affect the optical performance. Therefore, removing the phosphor layer from the die and coming up with a novel thermal package design will enhance the optical efficiency of an emission process. In addition to absorptive and reflective losses, certain losses are related to down conversion or Stokes loss of the phosphor layer in the package. During the conversion from blue to white light, a part of light emission is lost as heat, and it affects the amount of light emission and light characteristics. Since the optical performance of LEDs is thermally limited, novel cooling techniques that also favor optical behavior are continuously investigated. Air cooling is very effective in cooling low power LEDs due to availability, low cost, and reliability, although its performance is very limited due to the low thermal capacity of air. On the other hand, the efficiency of indirect liquid cooling is an order of magnitude higher compared to air cooling; however, the cooling performance is negatively affected by additional interfaces between LED to heat sink and heat sink to air. Direct immersion cooling of an LED chip with natural convection is found to be a promising and cost-effective solution for high power LEDs subjected to high heat fluxes. In addition, temperature uniformity of package components is achieved especially at higher driving currents that cause local temperature gradients. Moreover, in recent years, the interaction of unrestricted particles with the dispersed multi-phase flow has been linked to a number of important engineering applications. Among these applications, the novel idea of immersion cooled phosphor particles, which has the potential of significantly increasing the thermal limits of phosphor converted white LEDs (Pc-LEDs), has yet to be thoroughly investigated. Thus, considering the above-mentioned thermal, optical, and electrical concerns with current LED devices, this thesis focused on the accurate thermal characterization of LEDs through experimental and computational measurements and proposed a novel immersion cooling technique combined with YAG:Ce to improve the performance of Pc-LED while light systems. With this objective, first, an extensive literature review was conducted on experimental junction temperature (Tj) measurement of LEDs. For the purpose of attempting to address the governing phenomena, benefits, drawbacks, possibilities, and applications, a wide range of measurement techniques and systems are covered. The literature review section included a large number of published measurement approaches such as Temperature Sensitive Optical Parameters (TSOPs), Optical Temperature Probing, Temperature Sensitive Electrical Parameters (TSEPs), and a few other methods. Some of the corrections noted in non-ideal thermal calibration processes are discussed and presented. In addition, a summary of the experimental parameters employed in the literature is given as a reference. Secondly, based on the outcomes of the literature review, a more accurate and time efficient forward voltage method (FVM) was studied for experimental characterization of transient and opto-thermal behavior of bare and coated LED chips under operation. For pulse duration smaller than 50 ms, pulse currents ranging from 1 to 10 mA were found to be causing an acceptable temperature rise in all short pulse durations. On the contrary, pulse currents as high as 100 mA were suggested to be avoided unless they performed in the 1 ms range. Moreover, according to the bare chip thermal characterizations and reported interrelations between output and input powers, Tj and input currents, efficiency higher than 35% cannot be achieved in useful emission power losses due to the Auger losses that significantly reduce internal quantum efficiency and interfacial thermal resistance in thermal design. Correspondingly, an immersion cooling technique that directly targets heat generation zones was suggested to increase extraction efficiency and further reduce the junction temperature. In the experimental study, a preparation method for an immersion cooled LED package with dispersed phosphor particles was introduced. Details of the package design, step-by-step preparation, and material property specifications were explained in detail. Following that, an experimental setup that enables thermal characterization of immersion cooled LED package simultaneous to the particle velocity measurements was demonstrated. As a result, Tj and fluid temperature near the dome surface were measured to be 100 °C and 40 °C when LED has driven at 1 A current. The effectiveness of the immersion cooling technique was acknowledged by reaching lower junction temperature at higher driving currents. In addition, image processing of particle-based investigation of phosphor has revealed that in the presence of blue light, a 10-fold increase in particle velocity can be observed. This variation was mainly attributed to the photopheresis force acting on the particle when optical effects were incorporated. The origins of this force were discussed, and alterations in particle and fluid motion due to the intense photopheresis effect were reported. The findings presented in that part aim to shed light on future studies for particle trapping and the light attraction of photoluminescence particles. Thirdly, two-phase flow analyzes were conducted computationally by utilizing a discrete phase modeling approach in ANSYS Fluent environment. The impact of different heat generation rates of an LED package was investigated by considering natural convection currents and corresponding phosphor particle trajectories inside a fluid domain. Phosphor location and movement within a buoyancy-driven flow were further analyzed by including the self-heating effect. For that purpose, heat transfer, fluid flow, and energy paths of self-heating phosphor particles were extracted, and the influence of particle sizes was analyzed in detail. The results have shown that with immersion liquid cooling, the highest phosphor particle temperature was recorded to be under 147 °C, while larger size particles introduced higher heat transfer rates to the Pc-LED package for the same number of particles. Moreover, depending on the particle size and position, individual phosphor particles can follow a different trajectory that can affect the probability of obtaining white light emission.