Continuing, I�ve optioned about the uselessness of shear testing of solder connections on this forum previously. So, the points that you make about the elusiveness of developing a standard for measuring solder connection strength is well taken. I�ll add a couple of kickers. The value of strength also varies with: * Method used. * Speed of shearing the solder. * Type of solder. * Age / graininess of the solder. * Temperature of testing.
Additionally, when shearing a component, you not only apply shear stress, but also roll the component over. This means that both a shear stress and a normal stress are present, again making the results very dependent on where you apply the load.
Moving on, a properly formed solder connection is stronger than the underlying copper to laminate bond. Measuring pad peal strength: IPC-TM-650, Method 2.4.21 for multiple solderings, use Method 2.4.8 for copper peal using a Instron machine. Typical requirement for Cu is: 6 lb (force)/in [106 kg (force) /m].
Consider that periodically micro-sectioning solder connections on a coupon is probably the best approach to process control of this nature. [Realistically, it's not like a shear test represents any condition that a board is likely to encounter. Yer just tryin' to see if you soldered the little square thing to the flat thing!!!]
Alternately, get some vise grips and just torque the component from the board and then look at the terminations. A truly interfacial separation shows essentially only base metal on one fracture surface and solder on the other termination. You see intermetallic compound [IMC] rarely, because in most of these cases no metallurgical bonding (wetting = formation of IMC) has taken place. If you break the component, pull the pad, or see IMC; push on to something else [like rework].
Finally, responding directly to your question [Pfew, FINALLY!!!! ]. Look in EP&P 4/01 p 56. Average shear strengths using a 5kg load cell and a shear rate of 0.01 in/sec are: 0201/734 g; 0402/1750g; 0603/2200g.