A major concern is the reliability of the solder joints, which has traditionally been predicted using modeling techniques with established SnPb solder materials data. Although the mechanical behavior of lead-free solders is known to be different from that of their traditional counterparts, there is a dearth of credible measurement methods, in order to evaluate the likely lead-free solder performance. However, currently available test instruments do not readily meet these constraints. This paper describes a new approach developed at NPL to overcome these constraints, resulting in the design of an instrument IPTM (Interconnect Properties Test Machine), with a sample geometry that permits small solder volumes to be studied in pure shear [Ref 1 pp.468]. Example data are presented and illustrate how the low cycle fatigue resistance of solders can be characterized. Advantages of the new approach include: (i) it can accommodate various solder alloys and surface coatings, (ii) the solder joint volumes mirror those in modern assemblies, (iii) the solder is under shear, (iv) the samples are relatively easy to manufacture, (v) the construction of the specimen allows direct microscopic examination during the test at room temperature. A 4-point measurement system for resistance monitoring has also been evaluated and found to correlate well with load decreases recorded during fatigue testing of solders. The results of this study reveal that lead-free solders at high temperatures have increased lifetimes when creep constitutes major part of mechanical cycle. A route to predict a fundamental reliability variable of solder has been identified.
Discussion 1 MODEL SOLDER TEST SPECIMENCreep, relaxation and fatigue are the key parameters that need to be measured at various temperatures, and ideally test instruments need to be able to accommodate measurements under dynamic conditions. Since small samples are to be studied the displacement measurement accuracy required needs to be better than 0.1 µm. These requirements are quite demanding and as a consequence a number of novel solutions have been generated [Ref 2, pp. 67].A solder sample undergoing repetitive cyclic strain fatigue will fail gradually. The resistive force to the applied strain will decrease as a crack grows and microstructural changes occur. In each repeating strain cycle the maximum and minimum force will reduce as the crack grows. The rate of degradation (the difference between the maximum and minimum forces per cycle) is a measure of the solder performance. The point used to define the number of cycles to failure does not necessarily have to coincide with a 100% crack length. Hence a load drop parameter is defined as the reduction in the ratio of the difference between the maximum and minimum load at n cycles, compared to the difference between the maximum and minimum load at zero cycles. Whilst a proposed Japanese standard for tensile solder specimens recommends a 20% drop of load to define life-time [Ref 3], a similar US standard [Ref 4, pp 57] recommends a 50% lo...
