Hi Bill,
One possible analytic model that you may want to look at is Classical
Lamina Theory. This is an analytic model which attempts to describe the
macromechanical behavior of composites such as PWBs. I've used the model
assuming no restraining forces or moments, and looked at global warpage as
a function of time. (See
http://eislab.gatech.edu/pubs/conferences/wam97-scholand/) If you were to
set the board curvatures to zero, you could calculate the applied loads to
maintain PWB flatness, and calculate from these loads the internal
stresses. However, how one would model the effect of components on a
populated board would require some trial and error (you would have to
treat the components as a 'lamina'- effectively smearing them out to a
homogeneous layer).
Although our results should be extrapolated to your situation with
caution, we have observed high degrees of 'unsmoothness' in the warpage
vs. time plots (see Figures 4.5-1 and 4.5-2 in aforementioned paper),
which would indicate high creep rate loadings (d/dt warpage) for boards
without stiffeners. The most severe slope changes occur at the
temperature where the FR-4 transitions into the rubberoid region,
*suggesting* that preventing severe warpage with stiffeners would not
introduce undue stresses in the board. But I encourage you to run some
analysis to verify these trends apply to your situation as well.
Good Luck!
Andy
----------
From: [log in to unmask] [SMTP:[log in to unmask]]
Sent: Friday, October 24, 1997 8:52 PM
To: [log in to unmask]
Subject: Re: [TN] BGA Joint Reliability Under "Static" Stress
Hi Bill,
In a message dated 10/23/97 14:42:41, you wrote:
>Greetings Mr. Engelmaier!
>My name is Bill Barthel. I work for a contract assembly company called
>Plexus. We are rather involved in area array packages and recently your
>name was mentioned as a possible source for resolution to a re-occurring
>question.
>Most PCBs will not be perfectly flat, especially after assembly processes
>and if they have high (10+) layer counts. We have numerous examples but
>always find the boards to be well within the IPC spec. of 7ppm.
> None-the-less, many designs include stiffener bars due to their size and
>weight.
>The question I am looking for help on is:
>When stiffeners are applied to the PCB, stress is induced on all joints.
>What is the effect on fatigue life (per PCB bend radius?), particularly
>for BGA?
>I have posed this to some well connected people and no one seems to know
>of any work published in this area. Could you direct me to a possible
>source? Or, could you provide you opinion and any level of
>quantification possible?
I can not recall any work published in this area either.
You are looking at two damage mechanisms that could lead to failure, but
neither is related to fatigue.
(1) Solder joint overstress failures can be caused,essentially immediately
after soldering, if the PCBs assume a warp geometry after the soldering
process that produces tensile stresses in the solder joints that are large
enough and rapid enough to cause overstress failures. The strain rate has
to
be high enough for the solder not to be able to accommodate the
displacement
strains by creep at the prevailing temperature, thus allowing the tensile
stresses in the solder joint to build-up the fracture magnitudes, also at
this prevailing temperature.
(2) Solder joint creep rupture failure can be caused, over time, if the
PCBs
assume a warp geometry after the soldering process that produces tensile
stresses in the solder joints that are large enough to cause creep rupture
in
the solder joint. That could be at total displacements strains in excess
of
~30%. The time to failure depends on the magnitude of the total
displacement
strains as well as on the prevailing temperature.
In both these conditions, component size (smaller is better) and solder
joint
height (taller is better) play important roles; as does PCB thickness
(thinner is better, i.e. single-sided PCMCIA). If failure does not occur
as a
result of these conditions (which are present to some degree in all
assemblies), not significant effect on the fatigue life of the solder
joints
need to be anticipated.
Werner Engelmaier
Engelmaier Associates, Inc.
Electronic Packaging, Interconnection and Reliability Consulting
23 Gunther Street
Mendham, NJ 07945 USA
Phone & Fax: 973-543-2747
E-mail: [log in to unmask]
##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV 1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with
following text in the body:
To subscribe: SUBSCRIBE TECHNET <your full name>
To unsubscribe: SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for
additional information.
For the technical support contact Dmitriy Sklyar at [log in to unmask] or
847-509-9700 ext.311
##############################################################
##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV 1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in the body:
To subscribe: SUBSCRIBE TECHNET <your full name>
To unsubscribe: SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional information.
For the technical support contact Dmitriy Sklyar at [log in to unmask] or 847-509-9700 ext.311
##############################################################
|
