Author: Chien-Ming Huang
Title of dissertation: Durability distribution analysis of lead-free solder joints for printed circuit board applications
The date, time, & location: Friday, April 7th, 2023, at 01:00 pm in EGR-2164
List of all your committee members:
Dr. Jeffrey W. Herrmann, Chair and Advisor
Dr. Abhijit Dasgupta
Dr. Shapour Azarm
Dr. Katrina Groth
Dr. Aristos Christou, Dean’s Representative
Abstract:
Fatigue models for predicting the cycles to failure of solder joints under temperature cycling situations have been discussed and developed for decades. However, most models were developed according to specific solder materials, components, and printed circuit boards in each research. There is no study to cover and compare the fatigue models of solder joints through these different conditions. Therefore, the accuracy of the durability prediction of solder interconnections by using any of the available fatigue models can be unknown. On the other hand, current energy-based fatigue models for predicting the cycles to failure of the solder joint under thermo-mechanical loadings can only provide point estimates of the characteristic life or mean life. Nevertheless, the prediction of the fatigue life should be distributions with the uncertainties. Unfortunately, no study has been found that discusses the uncertainty of the cycles to failure, especially for the solder joints under temperature cycling. Therefore, the uncertainty propagation analysis of the cycles to failure is necessary to better estimate the distribution of the fatigue life of the solder joint.
The first part of this dissertation discusses nine existing energy-based fatigue models for different solder materials and components, and then analyzes major divergences between these studies. Moreover, the constants of the fatigue models are compared according to the divergences. Finite element simulation tool is applied to demonstrate the contributions of the factors on strain energy density and the variation on the predictions by applying these fatigue models. The results lead to conclusions with the advantages and limitations of using these available fatigue models for durability prediction of solder interconnections. These results provide insights that can help product designers understand and exploit the predictions of fatigue life while designing a printed circuit board and estimating its durability.
The second part of this dissertation identifies 11 uncertain input variables, which can propagate the uncertainties, via basic mechanics theory. The eigenvector dimension reduction method and FEA simulation tool are employed to determine the distribution of the system response, which is the strain energy density. Then, the distribution of strain energy density is converted to the distribution of characteristic life (in cycles) by choosing the appropriate fatigue model. Finally, the distribution of cumulative density function of the fatigue life of the solder joint is determined by taking the interval of characteristic life and specific shape parameters. In the end, the uncertainty of the fatigue life of the solder joint can be well estimated if the shape parameter information is available.