As a general acceleration transform you can see "Soldering in Electronics" by R.
J. Klein Wassink, 2nd Edition, page 216.  He gives an equation for acceleration 
in temperature/humidity environment for corrosion of 

    A = exp( (B*(1/T(L)-1/T(T))) + C*(RH(T)^2 -RH(L)^2)))

     Where A = Acceleration Factor
           B = ~7000K, equivalent to an activation energy of 0.6 eV
        T(L) = Temperature or life (usage) environment
        T(T) = Temperature of test environment 
           C = 4.4*10^-4, with RH expressed in percent 
       RH(T) = Relative Humidity (in percent) for test environment
       RH(L) = Relative Humidity (in percent) for life (usage) environment

     (NOTE:  I think I did the parenthesis correct but I suggest you look at a  
             copy of Klein Wassink or call me and I'll fax you this page.)

Caveats:  
   1.  This assumes that the effects of temperature and humidity are independent 
   effects (not true in many cases/environments)
   
   2.  Assumes that you don't use a test environment that introduces new failure 
   mechanisms (e.g. testing in condensing humidity when you don't see condensing 
   in life environment or testing at temperature beyond Tg of PWB when such 
   temperatures are seen in life environment.
   
   3.  Doesn't account for affects associated with electromigration (ie. voltage 
   driven effects)

Overall, our corrosion people think of this as a ROM for comparing environments 
but it is a case where there seem to be more exceptions to the rule than cases 
that follow it. 

Jim Maguire
Principal Engineer, E/E M&P
Boeing Defense & Space Group
(206)657-9063
[log in to unmask]



______________________________ Reply Separator _________________________________
Subject: Time Reduction of SIR Testing
Author:  [log in to unmask] at esdigate
Date:    8/21/95 12:10 PM


     Address,

     Here at Hitachi Computer Products Division (Norman, OK) we introduce  
     bareboards to extreme environmental tests.  These tests are 
     representative of 10 years of fault free life.  However, these tests 
     are extremely long and cut into our overall product life cycle.  
     Without sacrificing bareboard quality, I am needing a time to 
     temperature ratio for Humidity testing.

     i.e..

     Our Humidity tests parameters are:

     a. >=800hrs 
     b. 80?C (176?F)
     c. 90%RH
     d. R>10E9ohms @ 250VDC
     e. test times are every 50 hrs until 250, then at 500 and 800

     purpose:  we are evaluating the absorption and diffusion of moisture 
     and moisture vapor.  We are also checking metal migration, metal 
     through migration and metal surface migration (SIR).

     Can anyone provide a time:temperature ratio for SIR testing?

     i.e..  for every degree increased or humidity, time can be increased 
     or decreased by such.

     Please call and e:mail asap.  Thank you.

     Thomas John Gulley (PCB Project Engineer) 
     e:mail [log in to unmask]
     or
     405-360-5500 x634









     For several years I have been testing our customer bareboards