Solder joint reliability prediction for chip components, MELFs, TSOPs, SOTs, etc.
Figure 2 shows a schematic of a
Solder Model Parameters
solder joint cross-section for any of the
components in Figure 1, together with
ε
f
'
c
0
c
1
c
2
t
0
cross-sections of failed solder joints of two
Engelmaier-Wild Creep-Fatigue Model for SnPb Solders
components. These pictures illustrate the
dilemma faced in geometrically defining
SnPb 0.325 0.442 6.00e-04 -1.74e-02 360
the parameter h.
Engelmaier Creep-Fatigue Model for Pb-Free Solders
It is clear that there is no geometric
from property data & accelerated test results
certainty as to a measurable solder
joint thickness, and that the possible
SAC405/305 0.240 0.390 9.30e-04 -1.92e-02 100
h-measurements could range from zero to SAC205 <0.240 ? ? ? 100
20 mils (0.8 mm).
SAC105 <0.240 ? ? ? 100
While the solder joint in Figure 3 is the
result of long-term temperature exposure,
SnAg <0.240 ? ? ? ~60
it illustrates the problem with very thin
Table 1. Model parameters for solder creep-fatigue model.
solder joints—they consist almost entirely
of intermetallic compound (IMC) layers
and Pb-rich remains of the solder alloy.
problems. Setting the mesh too loose compliancy resulting from the solder joint
The IMCs are stronger than the solder,
makes it unrepresentative and too tight thickness was not as significant. That
but brittle, and they do not creep; the lead-
can lead to near-singularities. And in meant that typically by this definition h
rich phase has much higher Solidus and
order to make this considerable effort ≡ 2.5 mils (63 µm). That definition does
Liquidus temperatures and thus creeps very
worthwhile, the correct material behavior not work for components with leads,
little if at all and is stronger but less ductile
has to be inputed, including the stress-, particularly with stiff leads (TSOPs,
than solder. Thus, while there is a joint,
strain-rate- and temperature-dependent SOTs, etc.), which have to be analyzed as
it really cannot be termed a ‘solder’ joint
creep behavior of the solder. Much of ‘leadless.’
because it is no longer made of solder.
this information just is not fully available, Clearly, using the minimum ‘h’ is
It is clear that the problem of
and incorrect properties lead to ‘GIGO’
8
. overly conservative; it may be a decent
geometrically defining the parameter h also
Furthermore, one is still faced with the approximation to use as the value for h
exists with solder joints for components
problem of finding a connection between the height on the fillet at 45° from the
with attachment leads.
the determined stresses and strains, and termination corner as an upper limit and
Creep-fatigue models, like the one
reliability. half that as a lower limit.
shown in Equations 1 and 2, are predicated
Accelerated testing (ACT) is also Alternatively, I have suggested a value
on the creep-fatigue behavior of solder;
problematic. Since the thin solder joint of h = 2. 0 to 2.5 mils (50-63 µm), if less is
thus the question arises to what extent
section is essentially strong, brittle IMC, measured.
these models can be used to estimate the
in many applications the IMCs are not But the results need to be treated with
fatigue life of the solder joints illustrated in
overstressed and solder joint failures might caution.
Figures 1 through 5.
never occur. However, the stress levels
Of course, the fillets shown prolong
in ACT, let alone HALT and HASS, are
the life of solder joints by the time it takes
much higher, and thus may show failures
for the fracture to propagate to completion;
that simply would not occur in product
crack initiation, however, occurs in the
operation.
thinnest portion of the solder joints.
Clearly, the solder joint fillets make a
So, how does one go about estimating
contribution in terms of the time necessary
the creep-fatigue life of these solder joints.
for the fatigue crack to propagate through
There is no easy answer to this situation.
them, but how do you quantify that?
h (?)
The problem is how to quantify the
For leaded solder attachments the
strain in a representative manner; one is
definition of “half the solder paste
stuck with some ‘averaging’ of the h-value
stencil thickness” was found practical
unless finite element analysis (FEA) is
and close enough in the work done at
attempted.
Bell Telephone Laboratories, because for
Of course, FEA has its own set of
attachments with compliant leads the
h (?)
h (?)
h (?)
Figure 1. Depictions of the components, like ceramic chip resistors and capacitors,
Figure 2. Schematic of a solder joint cross-section for any of the components in
SM diodes and MELFs with leadless end cap solder attachments with a problem-
Figure 1 together with the micrographic cross-sections of a ceramic leadless chip
atic definition of solder joint height.
carrier (top) and a chip capacitor (bottom).
www.globalsmt.net Global SMT & Packaging – June 2009 – 31
Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46  |  Page 47  |  Page 48  |  Page 49  |  Page 50  |  Page 51  |  Page 52