MODEL BASED APPROACHES FOR SELECTING RELIABLE UNDERFILLFLUX COMBINATIONS FOR FLIP CHIP PACKAGESAuthor: Satyanarayan Iyer et al.
Company: SMART Modular Technologies
Date Published: 9/25/2005 Conference: SMTA International
The current research focuses on developing models that relate the assembly reliability with the properties of underfills and fluxes. The important properties of underfills and flux and their significance on the failures were studied using statistical techniques. Factors that were found to be significant were used for subsequent modeling. These models were developed using experimental data from JEDEC Level 3 testing of flip chip packages that used different underfill-flux combinations. Results for 95 different underfill-flux combinations were available for modeling. All other parameters except for the underfill-flux combinations were the same across all the data.
Regression and backpropagation neural network models were developed and validated using two approaches. The first approach was the commonly used approach of dividing the available data points into sets for model building and testing (validation) and second was the cross-validation approach. The results indicated that the neural network models performed better than the corresponding regression models in relating the failures with the selected properties of the underfill and flux. The neural network approach of modeling was recommended based on the results obtained as a part of this research effort and also due to the ability of neural networks to better capture non-linearaties.
Keywords: Flip chip, underfill, flux, regression, neural networks.
Members download articles for free:
Not a member yet?
What else do you get when you join SMTA? Read about all of the benefits that go along with membership.
Notice: Sharing of articles is restricted to just your immediate work group. Downloaded papers should not be stored on an external network or shared on the internet.