| Overview
This section summarizes our review and analysis of isothermal
tensile creep data for bulk lead-free solders of composition
close to that of the NEMI-selected Sn-3.9Ag-0.6Cu alloy. As
with the review of Sn3.5Ag properties, the ultimate goal of
this work is to develop preliminary constitutive models for
SAC and other NEMI recommended lead-free alloys for use in
stress / strain analysis, e.g. Finite Element Analysis (FEA),
and solder joint life prediction models. All reviewed data,
such as the tabulated creep data in Appendix B, will also
be included in a web-based material database.
Literature data was acquired for comparative analysis and
to identify gaps and differences among existing datasets:
- Isothermal bulk solder creep data from publications by
Kariya et al., 2001, Schubert / Wiese et al., 2001 and Neu
et al., 2001 was tabulated and plotted. Specimen geometry
and preparation, as well as a description of the microstructures
are summarized. The data covers two orders of magnitude
on the stress axis and eight orders of magnitude on the
strain rate axis. Test temperature is in the range -55°C
to 150°C.
- The raw data is presented in a tabular format for future
inclusion in a materials property database and possible
further analysis by others.
The intent of the analysis of the creep data is to develop
simple, first-order creep models in an attempt to bridge data
from several sources. Sub-datasets that do not fit in are
also highlighted and possible reasons are given for differences
in trends. It is also hoped that the gathered data will be
of use for others to develop more sophisticated creep models.
Main findings of this review and analysis are:
- Although there appears to be a few outliers, and the
SAC type alloys have slightly different Ag and Cu weight
percentages (including the Sn-2.5Ag-0.8Cu-0.5Sb CastinTM
alloy), the creep datasets from three independent sources
show first order consistency.
- Most of the data fit a simple hyperbolic sine creep equation
with a stress exponent of 6 and an activation energy of
67.9 kJ/mole (= 0.70 eV).
- Additional data by Kim et al., 2001, for specimens that
were rapidly cooled also fit the first order SAC creep model
while data for slowly cooled specimens is offset by two
orders of magnitude in terms of creep rates. As expected,
cooling rates have a significant effect on mechanical properties
(because of differences in microstructures). Those effects
are larger than the effect of alloy composition for the
alloys that were investigated and will have to be considered
in the development of more advanced creep models.
- The review of SAC data also points to lesser data available
at strain rates less then 10-6/sec, stresses less than 10
MPa and temperatures above 75°C or combinations thereof.
Future work should consider gathering creep data under those
conditions since they are representative of stress conditions
experienced by solder joints of electronic assemblies in
use.
The results of the creep data analysis are for bulk solder
in tension and do not include creep data for small size solder
joints as found in actual electronic assemblies or solder
joints in shear, although discrepancies between tensile and
shear creep models are discussed briefly.
This review of SAC properties is not exhaustive, additional
data will be added in when available. |
