Fatigue life evaluation of wire bonds in LED packages using numerical analysis

Reliability is essential for large-scale applications of high-power light-emitting diode (LED) devices, modules, and systems for general illumination. In this paper, the reliability of a novel application-specific LED package (ASLP) is investigated by thermal shock testing, failure analysis, and fluid-solid coupling thermo-mechanical simulation. The reliability of the ASLP modules was validated with a dual-bath liquid thermal shock testing from 233 to 398 K. The non-destructive failure analysis was conducted to the catastrophic failure ASLP samples by fluorescent penetrant inspection. The delaminations at the interfaces within the ASLP module were detected. The failure mechanisms were identified by digital optical microscopy and field emission scanning electron microscope inspection after the decapsulation. The experimental results show that fracture failure occurs at the wedge joint of the bonding wire, which leads to the catastrophic failure of the ASLP module. The stress and strain behaviors of the ASLP module, especially the bonding wire under thermal shock loading, were analyzed through thermo-mechanical modeling with the nonlinear time- and temperature-dependent material properties. Significant thermal gradient within the ASLP module during the thermal shock testing was taken into consideration by the fluid-solid coupling transient thermal transfer analysis. The effects of the delaminations detected by the fluorescent penetrant inspection on the reliability of the bonding wire were also examined. It is found the delaminations existing at the interfaces within the ASLP module induce significant plastic strain to the wedge joint and results in fracture failure. The results from the numerical simulation can make a good prediction of the failure mechanism of the ASLP modules under thermal shock loading.

Heat seal connectors (HSC) have been evaluated as a regular connection method for electronic components. In order to test the bonding reliability between HSC and the liquid crystal display (LCD) panel, several tests were carried out: thermal aging, thermal shock and vibration tests. The thermal aging test provided a constant condition of 120/spl deg/C for the HSC. The thermal shock test gave the HSC an alteration between extreme environments (-25/spl deg/C to 125/spl deg/C). The HSC peel strength after thermal aging decreased rapidly within the first 50 hours of aging, increased again and reached a peak at 150 aging hours. Afterwards, the peel strength decreased gradually. For the thermal shock test, the peel strength trend was that it decreased rapidly initially and became constant after 200 cycles of thermal shock. Peel strength after the vibration test fluctuated with the increase in the number of vibration cycles. Moreover, microstructural examination was also conducted on the treated HSC samples. A small reduction in the thickness of the copper tracks inside the HSC was found after thermal aging and thermal shock treatments. Also, some defects were detected which may be related to the decrease in peel strength.

This study investigated the microstructural evolution and evaluated the thermo-mechanical reliability of a Sn-3.0Ag-0.5Cu (in wt.%) flip chip package during a thermal shock test. The reliability of the flip chip bonded packages was evaluated by means of the thermal shock in the temperature range of 233 K to 398 K. After 250 thermal shock cycles, cracks finally occurred at the corner solder bump joint on the chip side interfacial regions. In that, the crack initiated at the solder region near the intermetallic compound (IMC) layers. For the interpretation of the failure mechanism of the package, finite element analyses were conducted. From the finite element analysis or computational simulation result, the distribution of the plastic work in the corner solder bump was found to be most considerable. This means that the corner solder bump has the highest potential for first crack initiation and growth, because the thermo-mechanical failure of the solder joints is mainly caused by the accumulation of the plastic work. This matched well with the experimental results.

Detection and analysis of defects in reactor pressure components by location and interpretation of acoustic emission sources

On the accumulation of irreversible plastic strain during compression loading of open-pore metallic foams

In this study, La2Zr2O7, Yb2O3 and Gd2O3-doped La[Formula: see text](Gd, Yb)[Formula: see text]Zr2O7 thermal barrier coatings have been manufactured from the crystalline and amorphous feedstock by plasma spraying. The thickness of the bond coat is 100±25 [Formula: see text]m and the thickness of the top coat is 250±25 [Formula: see text]m. Thermal shock tests were carried out in furnace temperature of 1200∘C. The thermal shock tests indicate that lifetime of the La[Formula: see text]Gd[Formula: see text]Zr2O7 coating has 55 cycles which is the highest thermal shock lifetime compared to other coatings in this study. The type of feedstock, amorphous or crystalline, strongly affects the thermal shock lifetime of La2Zr2O7 coatings. Coating damage after thermal shock tests was investigated using scanning electron microscope to observe the microstructure and X-ray diffraction techniques to analyze the phase composition. The failure of coatings is mainly in consequence of the accumulation of stress between bond coat–ceramic coat interface due to thermal expansion coefficient differences and thermally grown oxide layer formation during thermal shock cycling tests.

The phosphate system enamel containing no lead was produced and fund to be suitable to the plugging material of fluorence substance of the Electroluminescent.Aluminum plates were pretreated by the chemical and anodic oxidation method, to which the phosphate frit without lead was applied.The adhension of enameled plates were investigated by bending and thermal shock test, and the corrosion was also tested comparing lead system enamel with phosphate system enamel without lead. The results:1) In the bending and thermal shock test, aluminum and alloy plates of chemical treatment showed no good adhension for frit. Aluminum and alloy plates of anodic oxidation treatment showed the same good adhension as borosilicate and lead frit. In the thermal shock test, anodic oxidation treatment of alloy plates with sodium perborate, lithium borate, sodium phosphate and sodium perborate-sodium phosphate-potassium phosphate showed especially good adhension.2) In the corrosion test, the phosphate enamel without lead showed superior resistant to lead system enamel against inorganic acid.

Studies on aluminium enamelling carried out so far have revealed that the pretreatment of aluminium plate is the most important. As reported in the first paper, (Journal of the Metal Finishing Society of Japan, 8 [7] 11 1957), the samples coated with film obtained by anodic oxidation in dilute solution of sodium perborate gave the good results with respect to the adhesive properties and the luster of the finished products.The present paper contains the results of the investigations carried out in expectation of obtaining highly adhesive enamels by the formation of an intermediate layer having semiglass structure. It may be formed by the chemical reaction between the surface film, the coating of γ-alumina produced on metal surface by the anodic oxidation in aqueous solution of alkali borate, silicate and phosphate or their mixture, borosilicate, lead, and phosphate frits.The results may be summarized as follows:1. Good results in bending test of borosilicate enamel were obtained when the metal plates were pretreated by the solution of sodium borate, potassium borate, and water glass.In the binary solutions the mixtures of lithium borate and water glass gave the most satisfactory results.Among the ternary mixtures the solutions containing sodium borate and water glass as the main constituents gave especially good results.The samples treated with the mixed solutions gave the better results in thermal shock tests than those treated with single component solution.In all cases aqueous solutions containing potassium ion were found to be the most satisfactory.2. Lithium borate gave the best results in the bending test of lead enamel although there was no difference among sodium metasilicate, sodium perborate and lithium borate in the results of thermal shock and adhesion tests.The ternary solution of above three components gave the best results in bending test.3. In the bending test of phosphate enamel sodium phosphate solution was the best, while above three solutions gave the results of about the same level as those of lead enamel in thermal shock and adherence test. The ternary mixture gave the best result.

The reliability of QFP (Quad flat package) solder joint using Sn-8Zn-3Bi solder paste under the thermal shock test was investigated. Considering the environmental restriction such as ROHS, the QFP Cu LF (Lead-frame) was coated with lead-free materials (Sn, Sn-3Bi). To analyze the reliability under thermal shock treatment, the samples were placed in the thermal shock chamber (248K/423K, Dwell time: 30min). During the thermal shock test, the solder joint cross-sections were observed every 500 cycles. No crack initiation and propagation was observed through all type of plated Cu LF. The measured pull strength slightly decreased, as the number of cycles increased. After 1000 cycles, the pull strength of Sn and SnBi plated Cu LF reduced by 30% and 20%, respectively, compared with that of initial condition. Observing the fracture surface morphology by FE-SEM, the fracture mode changed and the IMC fractured area on the both fracture surface was increased. The IMC was identified as γ-Cu5Zn8 by chemical composition analysis using EDS. The reduced pull strength was affected by IMC fracture and fracture mode change. However, the pull strength does not decrease steeply but gradually. Consequently, the Sn-8Zn-3Bi solder joint shows the reliable solder joint strength, fracture mechanism, and compatibility with lead-free plated Cu LF during the thermal shock temperature of 248K to 423K.

This report compiles the work of BNL in support of Thermal Shock Resistant Cement evaluations (TSRC) by Schlumberger. BNL has worked on the formulation of thermal-shock resistant cement since 2011. All the materials on the blend including publications and presentations are in public domain and were sent to Schlumberger for review prior the beginning of the evaluations. Initial efforts focused on achieving long and controlled thickening time of TSRC at 85°C in thickening time tests in a pressurized consistometer. Since the results of BNL set control were not reproduced by Schlumberger technical feedback on Schlumberger testing was among the main efforts of the BNL team in the beginning of the collaboration. When it became clear that the main reason of results deviation was the difference in the mode of retarder addition (dry-blended at BNL and pre-dissolved at Schlumberger) the efforts shifted to the possible decrease in the cost of the formulation by changing concentrations and nature of the alkali activator as well as evaluating performance of more economical formulations with calcium-aluminate cement, Secar #51, Secar #71 and Ciment Fondu, suggested by Schlumberger based on communications with Kerneos (supplier of calcium aluminate cements). Another piece of information obtained by Schlumberger from Kerneos that prompted evaluations of different calcium-aluminate cements in the composition of TSRC was the fact that originally used Secar#80 contains some performance-improving additives that may interfere with the set control and rheological properties of the slurries. In summary, the report contains the following information: 1. Review of the setting behavior of TSRC slurries prepared by dry-blending TA powder or pre dissolving it in mix water in calorimetric tests at 85°C; 2. Results of evaluation of different alkali activators for improved compressive strength development and cost reduction; 3. Results of evaluation of Secar #51, 71 and cement Fondu as an alternative to Secar #80 in TSRC formulation – thermal shock tests and phases’ analyses of set cement; 4. A brief discussion of early-compressive-strength-development test conditions appropriate for geothermal wells.

A thick Cu column based double-bump flip-chip structure is one of the promising alternatives for fine pitch flip-chip applications. In this study, the thermal cycling (T/C) reliability of Cu/SnAg double-bump flip-chip assemblies was firstly investigated and the failure mechanism was analyzed through correlation of T/C test and the finite element analysis (FEA) results. After 1000 thermal cycles, the T/C failure site was the Cu column/Si chip interface, where was identified via a FEA as the location of the maximum stress concentration during thermal cycling. During thermal cycling, the Al pad and Ti layer between the Si chip and Cu column bumps were displaced due to thermo-mechanical stress. Based on the low cycle fatigue model, the accumulation of equivalent plastic strain resulted in thermal fatigue deformation of the Cu column bumps, and ultimately reduced the thermal cycling lifetime. In addition, the normal plastic strain of the y-direction, ε <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">22</inf> , was determined to be compressive and was a dominant component in relation to the plastic deformation of Cu/SnAg double-bumps. As the number of thermal cycles increased, normal plastic strains in the perpendicular direction to the Si chip were accumulated on the Cu column bumps at the chip edge in the low temperature region. Thus it was found that displacement failure of the Al pad and Ti layer, the main T/C failure mode of the Cu/SnAg flip-chip assembly, occurred at the Si chip/Cu column interface by compressive normal deformation during thermal cycling. Next, the effect of Cu column height was investigated for the enhancement T/C reliability. As results of T/C test for 60 um and 85 um Cu column heights, flip chip assemblies with thicker Cu column height showed better T/C reliability. In the real time moiré interferometry, shear strain and normal strain of the x-direction was almost same regardless of Cu column height. On the other hand, the normal strain of y-direction (perpendicular direction to the Si chip) at Si chip/Cu column interface for 85 um-thick Cu samples shows significantly reduced value compared with 60 um-thick Cu samples. This relaxation of the normal plastic strain of the y-direction is the origin that thicker Cu column height guarantees better T/C reliability

Due to price competition, low-cost design plays an important role in the power amplifier (PA) market. The cost of new single board design saves about 50% of direct material cost. Other advantages include decreasing direct labor cost, increasing maximum power output, increasing PA efficiency, and decreasing piece part count. For quality and reliability of solder joints, liquid-to-liquid thermal shock test would be considered during the design cycle. It was postulated that the failure is related to the coefficient of thermal expansion (CTE) mismatch between different materials, in the PCB, heat spreader, lead, cast heat sink, and device. The high power RF devices used in power amplifiers for base station applications have pushed the requirements of advanced materials and processes to provide new low-cost packaging solutions. A methodology is presented for design option evaluation and failure possibility prediction for solder joints on single board power amplifier. This method integrates the finite element analysis models and experimental data. The finite element model identifies the displacement at the solder joint locations. Several sets of experimental tests have been conducted and analyzed to evaluate the design improvements. Good correlation has been achieved between finite element analysis (FEA) predictions and experimental tests on design options.

Microstructure and thermal shock performance of Y2Hf2O7 coating deposited on SiC coated C/C composite

Based on the ultrasonic welding technology, this study uses three different welding widths to weld copper cables with different specifications. The influence of welding width on the mechanical properties and microstructure of each group of welded joints was systematically studied for the first time. The thermal shock test was carried out for each group of welded joints under optimum welding width to simulate the influence of severe temperature change environment on joint performance. It is found that the cross-sectional area of joint is 20 mm2 and optimal welding width of joint composed of two and three cables is 7 mm. The optimal welding temperature of the joint composed of four cables is 5 mm. Under the optimal welding width, the average shear strength of two-cable joint reaches 309.4 N. The four-cable joint is only 232.2 N. Moreover, the welding strength weakens significantly as the number of cables and the peak temperature decreases. The high temperature of bonding interface is the key factor to form a good weld. The peak temperature during welding is negatively correlated with the porosity of joint and positively correlated with peeling strength of joint. In addition, the morphology of ultrasonically welded joints has changed obviously after thermal shock test. With the participation of oxygen, the surface of welded joint is gray and bright brass, while the interior of joint is purple due to lack of oxygen. Moreover, the phenomenon of atomic diffusion and thermal expansion generates joints which were initially in a mechanically interlocked form and welding interface of the metallurgical bond under the action of high temperature. So the maximum joint peel strength is slightly improved.

Hermeticity and reliability of laminates bonded by Al-Al thermocompression bonding applying different bonding parameters have been investigated. Laminates of diameter 150 mm were realized by bonding a wafer containing membrane structures to wafers with patterned bonding frames. The laminates were bonded applying a bond force of 36 or 60 kN at temperatures ranging from 300 to 400 °C for 15, 30 or 60 minutes. The hermetic properties were estimated by optical measurement of membrane deflection after bonding. Reliability was tested by exposing the laminates to a steady-state life test, a thermal shock test, and a moisture resistance test. Laminates bonded applying a bond force of 60 kN at temperatures of 350 °C or higher resulted in hermetic and reliable bonds. All the dies on the laminates survived steady-state life test and thermal shock test. Leakage was observed in some dies after exposure to moisture resistance test. Measurements of membrane deflection estimated a maximum leak rate below 10-11 mbar×l×s-1 for all tested bonding parameters.