Quantitative Mechanism of Significant Benefits of Underfill in Flip-Chip Assemblies

2003 ◽  
Vol 125 (1) ◽  
pp. 84-92 ◽  
Author(s):  
Xia Cai ◽  
Liu Chen ◽  
Qun Zhang ◽  
Bulu Xu ◽  
Weidong Huang ◽  
...  
Keyword(s):  
Plastic Strain ◽  
Cross Sections ◽  
Flip Chip ◽  
Solder Joints ◽  
Axial Strain ◽  
Strain Range ◽  
Interface Stress ◽  
Acoustic Microscopy ◽  
Plastic Strain Range ◽  
Element Simulation

The thermal fatigue failure of SnPb solder joints of flip chip on board with and without underfill for two types of flip-chip packages was investigated by conducting thermal cycling test, scanning acoustic microscopy observation, and cross section inspection. The corresponding 3-D finite element simulation was performed to analyze the effects of underfill on thermomechanical behavior. The viscoelasticity of underfill and the viscoplasticity of solder were considered in the 3-D simulations. The Coffin-Manson equation with material constants C=5.54,β=−1.38 was fitted from the combination of the lifetime measured and the shear plastic strain range simulated by 3-D model. In the case with underfill, the plastic strain of every solder joint becomes very similar and little dependent on the position of solder joints. The modeled axial strain distribution coincided well with the distribution of microstructure coarsening visible in cross sections. The mismatch of thermal expansion resulted in an overall warpage of the assembly for the case with underfill, which decreased the shear deformation of the solder joints and increased the interface stress on the chip. The interface stress distribution from the 3-D simulation agreed very well with the experimental observations.

Fatigue Properties and Microstructure of SnAgCu Bi-Based Solder Joint

2020 ◽  
Vol 143 (1) ◽  
Author(s):  
Minghong Jian ◽  
Sinan Su ◽  
Sa'd Hamasha ◽  
Mohammad M. Hamasha ◽  
Atif Alkhazali
Keyword(s):  
Fatigue Life ◽  
Plastic Strain ◽  
Solder Joint ◽  
Solder Joints ◽  
Hysteresis Loops ◽  
Strain Range ◽  
Fatigue Performance ◽  
Fatigue Properties ◽  
Solder Alloys ◽  
Plastic Strain Range

Abstract The reliability of solder joints plays a critical role in electronic assemblies. SnAgCu solder alloys with doped elements such as Bi and Sb is one of the candidates for high reliability applications. However, the mechanical and fatigue properties of the actual solder joint structure have not been studied for these new alloys. In this paper, a cyclic fatigue test was conducted on individual real solder joints of different alloys, including SnAgCu, SnCu–Bi, SnAgCu–Bi, and SnAgCu–BiSb. The fatigue property of those solder joints was analyzed based on the characteristic fatigue life and stress–strain, hysteresis, loops. The results show that solder joints with both Ag and Bi content have a better fatigue resistance than the solder joints with Ag or Bi content only. The results of SnAgCu and SnCu–Bi solder alloys show similar fatigue performance. Also, the fatigue performance of SnAgCu–Bi is close to SnAgCu–BiSb in the accelerated test. But the SnAgCu–Bi alloy is estimated to have a longer characteristic life under low-stress amplitude cycling. The microstructure analysis shows a bismuth-rich phase formed around the Ag3Sn precipitates. Adding bismuth in the solder alloy can significantly improve the fatigue properties through solid solution hardenings. On another hand, the plastic strain range and work dissipation were measured from the hysteresis loops for all tests. The Morrow Energy and the Coffin–Manson models were developed from the fitted data to predict the fatigue life as a function of work dissipation and plastic strain range.

Low-Cycle Fatigue of AISI 1010 Steel at Temperatures up to 1200°F (649°C)

1977 ◽  
Vol 99 (3) ◽  
pp. 432-443 ◽  
Author(s):  
C. E. Jaske
Keyword(s):  
Plastic Strain ◽  
Fatigue Resistance ◽  
Stress Amplitude ◽  
Low Cycle Fatigue ◽  
Strain Range ◽  
Cyclic Stress ◽  
Cyclic Life ◽  
Stress Strain ◽  
Plastic Strain Range ◽  
Tensile Hold

This program was undertaken to develop isothermal low-cycle fatigue information for AISI 1010 steel in air. Such information is needed to help predict acceptable conditions for equipment and structures operating at elevated temperatures. Tensile properties and cyclic stress-strain behavior were also developed. For lives between 103 and 106 cycles to failure, fatigue curves were developed at 70, 400, 600, 800, 1000, and 1200°F (21, 204, 316, 427,538, and 649°C). Data for these curves were obtained from constant-amplitude, fully reversed strain-cycling tests of axially loaded specimens. Results from the same experiments were used to define cyclic stress-strain curves at each of the above temperatures. Dynamic strain aging caused a maximum amount of cyclic hardening at 600°F (316°C). In terms of stress amplitude, the maximum fatigue strength was at 600°F (316°C). In terms of either total strain range or plastic strain range, the maximum fatigue resistance was at 400°F (204°C). At temperaures above 600°F (316°C), fatigue resistance decreased as temperature increased. Tensile hold periods caused a significant reduction in cyclic life at 800 and 1000°F (427 and 538°C) but had no noticeable effect on cyclic life at 600°F (316°C). Fatigue resistance was quantified in terms of power functions relating fatigue life to both plastic strain range and stress amplitude, and cyclic stress-strain response was quantified in terms of a power function relating stress amplitude to plastic strain amplitude. The method of strain-range partitioning provided good cyclic life predictions for the limited number of tensile hold-time experiments, although other types of hold periods were not evaluated.

scholarly journals Isothermal Fatigue Behavior of Sn-Pb Solder Joints

1990 ◽  
Vol 112 (2) ◽  
pp. 94-99 ◽  
Author(s):  
T. S. E. Summers ◽  
J. W. Morris
Keyword(s):  
Fatigue Life ◽  
Plastic Strain ◽  
Strain Rate ◽  
Fatigue Behavior ◽  
Strain Range ◽  
Load Drop ◽  
Plastic Strain Range ◽  
Double Shear ◽  
Isothermal Fatigue ◽  
Fatigue Data

Isothermal fatigue data were collected for the compositions 5Sn-95Pb, 20Sn-80Pb, 40Sn-60Pb, 50Sn-50Pb and 63Sn-37Pb within the binary Sn-Pb system. All of these compositions are commercially available and include those most commonly used. Because Sn-rich solders are rarely used, they were not investigated here. The fatigue life was defined by a 30 percent load drop. The solders were tested in a double shear configuration joined to copper at 75° C. The displacement rate chosen was 0.01 mm/min, which corresponds to a strain rate of 1.5 × 10−4s−1 for our specimen configuration, over a 10 percent plastic strain range. Additionally, the microstructural changes during fatigue are presented. The various solder compositions studied exhibit strikingly different as-solidified microstructures. These differences are discussed in terms of their effect on the isothermal joint failure mechanism and joint isothermal fatigue life.

Failure Analysis of Flip-Chip Bumps After Thermal Stressing

2000 ◽  
Author(s):  
Tejpal K. Hooghan ◽  
Kultaransingh Hooghan ◽  
Sho Nakahara ◽  
Robert K. Wolf ◽  
Robert W. Privette ◽  
...  
Keyword(s):  
Cross Sections ◽  
Flip Chip ◽  
Focused Ion Beam ◽  
Ion Beam ◽  
Diagnostic Technique ◽  
Acoustic Microscopy ◽  
Under Bump Metallization ◽  
Thermal Tests ◽  
Transmission Electron ◽  
Electrical Bias

Abstract This paper describes a new diagnostic technique for analyzing microstructural changes occurring to flip chip joints after accelerated thermal tests. Flip chip reliability was assessed at high temperatures, with and without the application of electrical bias. A combination of standard metallurgical polishing techniques and the use of a focused ion beam (FIB) lift out technique was employed to make site-specific samples for transmission electron microscopy (TEM) cross-sections. We studied evaporated 95Pb/5Sn bumps, on sputtered Cr/CrCu/Cu/Au as the under bump metallization (UBM). Thermally stressed samples were tested for electrical continuity and evaluated using 50 MHz C-mode scanning acoustic microscopy (C-SAM). Failed samples were crosssectioned and large voids at the UBM were observed optically. TEM specimens taken from the predefined UBM region of degraded flip chip devices provided critical microstructural information, which led to a better understanding of a cause of degradation occurring in the flip chip joints.

A Two Surface Model for Transient Nonproportional Cyclic Plasticity, Part 1: Development of Appropriate Equations

1985 ◽  
Vol 52 (2) ◽  
pp. 298-302 ◽  
Author(s):  
D. L. McDowell
Keyword(s):  
Plastic Strain ◽  
Kinematic Hardening ◽  
Strain Range ◽  
Cyclic Plasticity ◽  
Plastic Strain Rate ◽  
Internal Variables ◽  
Surface Model ◽  
Fading Memory ◽  
Plastic Strain Range ◽  
Hardening Modulus

A two surface stress space model is introduced with internal state variable repositories for fading memory of maximum plastic strain range and non-proportionality of loading. Evolution equations for isotropic hardening variables are prescribed as a function of these internal variables and accumulated plastic strain, and reflect dislocation interactions that occur in real materials. The hardening modulus is made a function of prior plastic deformation and the distance of the current stress point from the limit surface. The kinematic hardening rules of Mroz and Prager are used for the yield and limit surfaces, respectively. The structure of the model is capable of representing essential aspects of complex nonproportional deformation behavior, including direction of the plastic strain rate vector, memory of plastic strain range, cross-hardening effects, variation of hardening modulus, cyclic hardening or softening, cyclic racheting, and mean stress relaxation.

scholarly journals Thermal-Stress and Low-Cycle Fatigue Data on Typical Materials

1965 ◽  
Author(s):  
K. E. Horton ◽  
J. M. Hallander ◽  
D. D. Foley
Keyword(s):  
Plastic Strain ◽  
Low Cycle Fatigue ◽  
Mechanical Strain ◽  
Thermal Strain ◽  
Strain Range ◽  
Fatigue Tests ◽  
Cycle Fatigue ◽  
Ferrous Alloys ◽  
Plastic Strain Range ◽  
Analysis Of Results

This paper presents the results of low-cycle-fatigue tests wherein either thermal strain or mechanical strain was the independent variable. The materials investigated were primarily ferrous alloys for use in nuclear reactors. The analysis of results was based on plastic-strain-range measurements which could be made reproducibly in the 2 × 10−5 range. Graphs of plastic strain range versus cycles to failure were often found to be independent of large variations in temperature and cycle time. The results from thermal-fatigue and constant-temperature-fatigue tests were usually indistinguishable on these graphs, suggesting that identical metallurgical phenomena occurred in each type of test.

Sign in / Sign up

Export Citation Format

Share Document