
Publica
Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten. Lifetime modeling for JEDEC drop tests
| Institute of Electrical and Electronics Engineers -IEEE-: 10th International Conference on Thermal, Mechanical & Multi-Physics Simulation and Experiments in Microelectronics and Microsystems, EuroSimE 2009 : Delft, Netherlands, 26 - 29 April 2009 New York, NY: IEEE, 2009 ISBN: 978-1-4244-4160-0 ISBN: 978-1-4244-4159-4 pp.309-317 |
| International Conference on Thermal, Mechanical & Multi-Physics Simulation and Experiments in Microelectronics and Microsystems (EuroSimE) <10, 2009, Delft> |
|
| English |
| Conference Paper |
| Fraunhofer ISE () Fraunhofer IWM () |
Abstract
Since lead was banned from the solder joints, mechanical tests of microelectronic components and modules have gained much importance not just in mobile electronics but quite in general. The standardized JEDEC drop test is commonly used. It delivers repeatable results for a wide range of loads in the components mounted oil the well specified test PCB. As yet the JEDEC drop test is time consuming and thus expensive, the goal of this study has been to develop a lifetime model for virtual assessments of the drop test performance by means of FEM simulation. The development started with assuring that the dynamic PCB deformation is captured precisely. Hereby, the constraint conditions were found to he most influential for minimizing the differences between the acceleration curves measured and the FEM simulation results (LS-Dyna). Subsequently, a number of known result criteria were evaluated by means of their capability to estimate the sequence of component failures within theexperiments. It turned out that none of the criteria was able to predict the failure sequence accurately. Therefore, a new criterion was composed. Combining plastic strain rate and resulting force integral, the experimental failure distribution was matched perfectly for all three package types investigated. The lifetime model based on the combined criterion was able to forecast the experimental number of cycles to failure with less than +/- 25% inaccuracy.