No abstract

It is necessary to improve the lifetime prediction based on FE-Methods of different electronic packages in order to reduce the time and costs of new developments. This paper purposes a method describing the crack propagation of chip on copper substrate solder joints. The Chips that were studied are power transistors. They were soldered on Copper Substrate (NiAu metallization) with two different solder alloys (SnPb eutectic and SAC 305). The chip dimensions and the solder joint thickness have an influence on the lifetime, so that two different chips with two different solder thicknesses were used as test specimens. These were thermally loaded, and the state of the solder joints was regularly checked. Three different kind of methods were used to characterize the damage of solder joints: the Scanning Acoustic Microscope (SAM) detects the crack initiation and propagation; the cross section analysis can validate the results of the scanning acoustic microscope and can show the microstructure changes; and the thermal resistance which is influenced by the damage of the solder joint was measured and correlated to damage. After presenting the results of these investigations, a general FE-method predicting the crack initiation and propagation using the Paris Laws is presented.

The time and temperature dependant creep deformation of solder alloys has to be studied to describe material characteristics and failure behaviour in order to use it for lifetime evaluation by FE-simulations. It is often found in literature that material behaviour of eutectic SnPb solder is described considering only secondary creep. As this paper will show, primary creep may not be neglected because of its appearance in the performed characterisation tests and application examples. To study the creep behaviour, the grooved lap geometry described by Reinikainen was used and additionally modified to see crack and cavities in the solder joint using SAM (Scanning Acoustic Microscope). The test specimen was cyclically loaded for different stress levels and temperatures. Primary and secondary creep was observable under conditions of load reversal. A combined creep law using primary and secondary creep will be presented and an example is shown to highlight the relevance of the primary creep