Simulation and Interpretation of Wetting Balance Tests Using the Surface Evolver

Abstract

The wettability of PCB pads and component terminations, and the variation of wetta-bility with time, are important factors in the successful formation of a solder joint. However, reliable techniques for determining these properties, under conditions representative of the soldering process, do not as yet exist. The development of wetting balance tests has provided a technique for the qualitative comparison of surface wettability, but the test has poor repeatability and it is difficult to relate the test results to the actual surface properties. The small size of the terminations in surface mount technology (S.M.T.) prevents successful use of the conventional solder bath based wetting balance test. The globule wetting balance goes some way towards removing this limitation and is now often used as a means of comparing the solderability of S.M.T. devices, but is even worse in terms of repeatability and interpretation. This paper presents results from an evaluation of conventional and globule wetting balance tests using both computational modelling and experimental techniques. The results from the experimental tests show that reasonable consistency in the measured force can be obtained, but that the insensitivity of this force to the wetting angle of the component, particularly in the globule block test, precludes their ability to accurately establish the wetting angle. It is therefore concluded that this test is unsuitable for providing direct quantitative measurements of the wetting angle of the surface of interest. The computational models, through allowing an analysis of the sensitivity of the test to the uncontrolled test variables also provide an explanation for the significant scatter obtained in wetting balance test results. It is also seen, from comparison with test results, that models which incorporate commonly quoted text book values for the solder surface tension do not adequately predict the forces empirically observed. A technique is therefore also described whereby the solder surface tension under the actual test conditions may be deduced using the results from a scanning mode wetting balance test, thereby greatly improving the capability of the models to predict the surface tension forces, although significant errors in the predicted meniscus shape remain.