Search

Low‐cycle fatigue testing of a lead‐free solder (InnoLot) based on Sn‐3.8Ag‐0.7Cu (SAC387) with three simultaneous additions of bismuth, nickel and antimony was conducted using miniature‐sized fatigue specimens at different temperatures and strain amplitudes. The experiments show a decline of the load capacity of the solder alloy with the number of loading cycles. The fatigue life of the solder is also decreased by the level of imposed temperature. The temperature‐modified Coffin‐Manson and Morrow models were used to analyse the behaviour under fatigue and predict lifetime. The parameters in the two fatigue models were determined by considering experiments at different temperatures and total strain amplitudes. In comparison with other reference lead‐free solders, the InnoLot solder shows much better fatigue strength. The better fatigue strength is found to result from the effect of BiNiSb elements. Also, lifetime predictions were made with both models for the solder alloy under different conditions. [ABSTRACT FROM AUTHOR]

In this paper, we investigate the effects of isothermal aging on the microstructure and creep properties of a novel lead-free SAC387-3Bi-1.5Sb-0.15Ni (InnoLot)/Cu solder joints. Single-lap shear solder joints of the InnoLot solder were aged up to 1 year at 100 °C, and then loaded to failure in shear creep tests under different applied shear stresses and testing temperatures. The findings show that both aging time and temperature lead to the global increase of the interfacial IMC thickness significantly, but the growth of the Sn-rich IMC layer during aging remains limited in comparison with other solder materials due to the Ni content. In addition, shear creep tests reveal that the shear creep strength of the aged solder joint decreases initially before remaining dramatically stable in the long term. In terms of failure modes, shear creep tests at room and elevated temperatures show distinct rupture features. While rupture within the bulk solder happens at higher temperatures, rupture through the interface occurs at lower temperatures. When the specimens are exposed to isothermal aging, creep deformation becomes more predominant. The aged microstructure exhibits a brittle-to-ductile transition with occurrence of intergranular fracture due to the dynamic recrystallization inside the solder matrix. It is observed that the mechanical factor is of first-order importance in the formation of the new grain structure of the solder. Image 1 • Single-lap shear solder joints are tested after isothermal aging. • After initial drop with aging time, creep resistance remains stable up to one year. • Ni content slows down IMC growth compared to conventional solders. • Recrystallization is observed under load with brittle-to-ductile transition. [ABSTRACT FROM AUTHOR]

Characterization of solder alloys is of primary importance for manufacturers due to the wide use of these alloys as interconnects for electronic packaging. Numerous investigations have so far been dedicated to determination of their mechanical properties like their creep response under thermo-mechanical loading, in particular. There is a broad variety of experimental devices and techniques that the researchers employed to study the different types of solders. We have recently designed a micro-tensile tester to characterize miniature solder specimens under various conditions of temperature and extension rate. In the present work, this experimental apparatus is complemented with imaging equipment, giving the ability to perform Digital Image Correlation (DIC) for full-field measurements on solder materials. The suitability and accuracy of both the optical system and the DIC technique for tin-based alloys characterization are first evaluated, based on specific assessment criteria. It is found that the system resolution is sufficient to obtain reliable results. A full-field strain measurement is then done by testing a standard specimen. Materials parameters, such as the elastic constants, are extracted from DIC data and compared with the measures obtained from the conventional sensors equipping the micro-tester, showing the ability of the newly designed optical system to be used in the small strain range. The study of necking has also been proven to be accurate with this system, allowing the precise evaluation of the highly localized strain in the specimen. [ABSTRACT FROM AUTHOR]

Due to the RoHS and WEEE legislations for restricting the use of six hazardous materials in the manufacture of various types of electronic and electrical equipment, developing novel Pb-free solders becomes a real challenge for many industrials in recent years. In addition, mechanical properties of the lead-free alloys are very important factors in the design and reliability evaluation of the soldered joints. This paper reports the findings of an investigation into the tensile properties of a new lead-free solder alloy which contains Ni, Bi and Sb additives (SAC387-3Bi-1.5Sb-0.15Ni). The fabrication procedure of the bulk samples is described in this study, as well as the mechanical testing of the obtained specimens. The tensile tests are conducted at temperatures between room temperature and 125 °C and under strain rates between 2.0 × 10 −5 and 2.0 × 10 −2 s −1 . The results show that both temperature and strain rate may have great effects on the mechanical behavior of the solder alloy. The strength was found to decrease with increasing test temperature and decreasing strain rate. It is noted that the mechanical properties of SAC387-3Bi-1.5Sb-0.15Ni solder alloy are strongly dependent on temperature and significantly sensitive to strain rate. Compared with the reference material without additives, additions of Sb, Ni and Bi elements result globally in an increase in strength of the solder alloy. Also, the stress-strain curves from tensile tests are used to identify nine Anand material parameters by using non-linear least square fitting. The identified parameters prove to be in good agreement with those found in literature on other usual solder alloys. Based on the Anand model, a FEM analysis of a multilayered IGBT packaging module under cyclic thermal loading is presented to predict the fatigue life of the solder joints. In addition, SEM and EPMA microanalyses of both as-cast bulk and fracture specimens are carried out to assess effects of testing conditions on microstructure changes in the SAC387-3Bi-1.5Sb-0.15Ni solder material. The improved strength is found to result from the solid solution hardening effect of Sb and Bi in the Sn matrix, together with the formation of (Cu,Ni) 6 Sn 5 and Ag 3 (Sn,Sb) intermetallic compounds (IMCs). [ABSTRACT FROM AUTHOR]

Nowadays, one of the strategies to improve the reliability of lead-free solder joints is to add minor alloying elements to solders. Therefore, the aims of this study are to investigate the effect of Ni and Sb, as well as that of the testing conditions (temperature and strain rate), on the mechanical properties of two new lead-free solder joints, Sn-3.8Ag-0.7Cu-3Bi/Cu and Sn-3.8Ag-0.7Cu-3Bi0.2Ni1.54Sb/Cu. The changes, after testing, in their respective microstructure will also be investigated. The procedure to fabricate appropriate joint samples is described in this research. The shear tests were conducted by micro-tensile machine at temperatures between room temperature and 125 °C and strain rates between 2.0×10 −4 s −1 and 2.0×10 −2 s −1 . The experimental results indicate that, with additions of Ni and Sb in Sn-3.8Ag-0.7Cu-3Bi, the ultimate tensile strength (UTS) and the yield stress ( σ y ) are improved. Moreover, both temperature and strain rate may have substantial effects on the mechanical behavior and the microstructural features of the solder alloys. During the shear tests, the strength decreases with the increase of the test temperature and/or the decrease of the strain rate. The enhanced strength is attributed to the solid solution hardening effects of Sb in the Sn matrix and the refinement of the microstructure with the addition of Ni. Ni and Sb additions lead to the formation of new (Cu, Ni) 6 Sn 5 and Ag 3 (Sn, Sb) IMCs. Needle-like (Cu, Ni) 6 Sn 5 appeared in the Sn-3.8Ag-0.7Cu-3Bi0.2Ni1.54Sb/Cu, instead of the globular-like Cu 6 Sn 5 inSn-3.8Ag-0.7Cu-3Bi/Cu. In addition, the nine Anand material parameters are identified using the shear stress-strain data and a non-linear least square fitting and its validity is checked by means of the experimental data. The findings found that these Anand parameters are also in good agreement with data given in literature for other solder alloys. The obtained material parameters of the Anand constitutive model were utilized to analyze the stress-strain response of an Insulated Gate Bipolar Transistor (IGBT) under thermal cycling. Furthermore, the SEM analysis of as-fabricated joint specimens as well as fractured specimens at different testing conditions was done to observe the effects of these factors on the microstructure of the solder alloys. [ABSTRACT FROM AUTHOR]

Failure at grain boundaries has a critical effect on the overall fracture behaviour of polycrystalline aggregates, as is the case in many metals. In the case of dynamic embrittlement, segregation of impurities occurs at grain boundaries, lowering their cohesive strength. The material response is then dominantly determined by grain boundary properties. The self-consistent scheme is extended to account for grain boundary decohesion by using a cohesive law to represent crack initiation and propagation. After introducing the imperfect interface conditions into the Eshelby's equivalent inclusion solution, the effective tensile response and brittle intergranular fracture of a Cu-Ni-Si alloy is predicted. The proposed analytical model allows for the identification of parameters for both crystal plasticity and cohesive constitutive laws, from a single macroscopic tensile curve. Afterwards, multiscale computations of artificial microstructures are done using the analytically calibrated values of material parameters. Comparison of the results with experimental data shows a satisfactory agreement. [ABSTRACT FROM AUTHOR]

Highlights: [•] Grain boundary decohesion is included in a new transition rule for polycrystals. [•] Homogenized behaviour of a copper alloy under intergranular cracking is predicted. [•] A single tensile fracture test can be used to identify grain boundary parameters. [ABSTRACT FROM AUTHOR]

Highlights: [•] Thermal–mechanical fatigue model is developed, based on cohesive zone method. [•] Mixed-mode loadings and effects of temperature and of strain rate are accounted for. [•] FE life prediction of solder joint in power electronics under passive cycling is given. [Copyright &y& Elsevier]

Abstract: A Cohesive zone model (CZM) is developed for simulating the thermo-mechanical fatigue of solders in electronic chip packages subjected to active power cycling. Repeated on–off switching of the device is the source of thermal strains, resulting in the initiation and the propagation of a fatigue crack in the solder joint. The fatigue failure in the solder material may cause local intensification of the temperature gradient due to partial or total degradation of the thermal conductance in the cracked region. In this paper, the coupling between fatigue cracking and crack heat transfer is accounted for by using a thermo-mechanically coupled cohesive zone element. Thus, the quantity of heat flux transferred across the interface will be dependent on the degradation level of the interface thermal conductance. Typically, the cohesive zone conductance is taken as a function of the separation between the cohesive surfaces, meaning that the thermal conductance can be restored across the interface during crack closure. It may alternately be chosen as being dependent on the damage’s level of the cohesive zone element, describing in this way an irreversible thermal degradation in the solder joint. Simulations are carried out in order to analyze and predict the solder fatigue under active power cyclic loading. The numerical results reproduce qualitatively the desired behavior for the cracked interface and demonstrate how the temperature field and the heat transfer in the system are radically changed. Then, the evolution of the average damage level in the cohesive layer, reflecting the crack propagation level during power cycling, is used to estimate the number of cycles to failure at which the device becomes faulty. [Copyright &y& Elsevier]

Abstract: Finite element analysis of kink band formation in wood is carried out using an anisotropic failure criterion. The criterion is capable of describing the mixed mode I/mode II crack development observed in the kinked region. The evolution of cracking is simulated with the help of the so-called smeared crack approach. As for the finite rotation of the fibers in the kinked material, it is implemented in the finite element code through a hypo-elastic law characterized by an objective derivate using the rotation of the fibers. This formulation enables to follow strictly the matter under finite strain; which means that the evolution of the orthotropic directions is correctly described. Numerical results, such as the predicted load carrying capacity of wood under compression and its post-cracking behavior, prove to have a good agreement with the experimental observations. Further, the applicability of softening orthotropic plasticity under finite strain conditions for simulating compressive failure modes in wood is established. [Copyright &y& Elsevier]

Abstract: In power electronic packages, one of the main limiting factors for the module reliability stems from failure of the electrical interconnection which ensures the contact between the chip and the lead frame. The aim of this work is to model, using FEM and some analytical developments, the interconnection heel crack mechanism appearing in service. The forming process impact is particularly evaluated and it is established that the initial residual stresses contribute to limit the wire/ribbon life time. [Copyright &y& Elsevier]

Abstract: A modeling approach for the kink band formation in wood is developed in order to predict the compressive strength and the band inclination for some wood species (spruce and beech). The model accounts for the large fiber misalignment observed in wood and for its behavior in shear and in the direction transverse to the fibers. The behavior of wood under this last regime proves to play an important role in giving the kink band its orientation. For this purpose, wood behavior, which is size dependent, is experimentally examined at the kink band scale in order to properly determine all the needed material parameters for the model. Then, combined with a criterion well suited for the description of the anisotropic failure of wood and using the experimental measurements, the resulting formulation for the kink band mechanism gives estimations of both the compressive stress at kink formation and the band inclination, in very good accordance with experimental observations. [Copyright &y& Elsevier]

Abstract: OSB is a construction panel made with layers of precision manufactured wood strands that are aligned, formed into panels and pressed with an exterior grade adhesive resin. The aim of the paper is to provide a theoretical analysis of the hygromechanical behaviour of this material by using a two-step method. First, the mechanical and morphological parameters characterizing the wood particles, as well as the information concerning their orientations, are used in a multi-scale approach for the prediction of the behaviour of a single layer. Then, the classical theory for the analysis of laminates is used to obtain the stiffness constants of the panel. The flexural Young moduli of the panel as well as its in-plane hygroelastic constants are studied in relation with the orientation distribution followed by the particles. The study permits to establish that the orientation distribution function is a key factor in obtaining high structural properties for OSB. It is shown that an improved strand alignment may help to gain significant performance with respect to stiffness and behaviour under moisture conditions. [Copyright &y& Elsevier]

Abstract: In this work, the energy release rate is computed for a curved crack in a two-phase particulate composite Al/SiC. The crack lies in a flexible interface between the spherical particle and the surrounding matrix. Applying classical approaches to calculate the energy release rate of that kind of crack is not an easy task, especially as the problem is of three-dimensional nature and the material is heterogeneous. Thus, the principles of the so-called θ-method are used to deal with these difficulties. Using the variational formulation, the method is rewritten for the case of cracks lying in compliant interfaces, which permits to obtain the expression of the energy release rate in the form of a domain integral. The expression is eventually discretized in the framework of the finite element method in order to have a usable form for the simulations. The validation of the method is made in the first place by considering the simpler case of a crack around a cylindrical inclusion, for which an analytical solution exists. The energy release rate is then evaluated for the cracked Al/SiC composite with compliant interfaces while making some micromechanical parameters vary, like the mechanical properties of the constituents. The evolution of the energy release rate is also represented with respect to growing crack lengths. [Copyright &y& Elsevier]

The assumption of imperfect interface using the multi-scale approach enables a more realistic description of composite behavior. In a heterogeneous material, the junctions between the constituents do not always play their full role because defects may appear in these highly constrained regions. In this work, the interface is considered as an elastic medium with vanishing thickness. It can lead to a jump in the displacement but continuity of the traction across the interface is possible. The interface is modelled analytically and is given two stiffness parameters which characterise its normal and tangential extension. A simple polynomial form of the displacement jump is used to implement this condition together with the approach of Eshelby. The overall elastic moduli of a two-phase composite is computed by taking into account the existence of imperfect interfaces. The same interfacial model is modelled using the finite element analysis for predicting the macroscopic response of the composite. [Copyright &y& Elsevier]

In this study, the effective behavior of a metal–matrix composite reinforced with particles is numerically predicted by taking into account the existence of imperfect interfaces between the constituents. The bonding conditions between the particles and the matrix are characterized by the existence of a jump in the displacement which is proportional to the traction transmitted across the interface. In order to describe the degradation of this flexible interface, two approaches are implemented in a finite element analysis. One of these approaches consists in using a damage law deriving from the thermodynamical framework and giving an evolution of the interfacial properties during the solicitation. The other approach is based on a failure criterion written as a Coulomb law and combining the effects of the normal and tangential tractions at a given point of the interface. [Copyright &y& Elsevier]

• DIC-based evaluation of fracture parameters of lead-free solder. • Identification of appropriate constitutive laws for monotonic and creep behaviours. • Fitting procedure using measured displacement field for evaluation of SIF. • Domain integrals based on DIC data points used for evaluations of J and C(t) • Methodology capabilities and shortcomings are discussed in relation to FEM results. New lead-free solder alloys for electronic packaging contain doping elements conferring upon them improved properties in terms of ease of processibility and strength under high temperature operation conditions. Studies on their resistance to quasi-static rupture remain however relatively limited. In this work, the SAC-based solder alloy, known as InnoLot, is considered and its rupture under monotonic and creep loadings is investigated experimentally and numerically. The Digital Image Correlation (DIC) technique is employed to evaluate the stress intensity factor, as well as the J- and C(t) -integrals, from full-field measurements on notched specimens. The numerical formulations of the theoretical concepts for computation of the physical fracture quantities are provided in the framework of the DIC algorithm. Results derived from the DIC experiments are compared with predictions given by finite element models, and good agreement is found in general. The observed discrepancies can be explained by lack of resolution of the optical setup measurements in the infinitesimal strain regime or by uncertainty associated with the identified material parameters used in the evaluation of stresses from the acquired strains. [ABSTRACT FROM AUTHOR]

Abstract: In automotive power electronics, low cost and reliability requirements both demand to optimize each part of the package in order to increase its life time during use. The wire bonding breakage accounts for one of the most common failures observed on power modules. In particular, the heel crack mechanism, mainly due to Joule self-heating, can occur in the module, leading to its death. Thus, the wire geometry design has to be set up to maximize its endurance. This paper presents an analytical method able to predict the optimal wire length regarding a wire displacement criterion. Finally, comparisons between experimental results and predictions of the analytical model are given. [Copyright &y& Elsevier]