Realistic non-destructive testing of integrated circuit bond wiring using 3-D X-ray tomography, reverse engineering, and finite element analysis

摘要

Within industries that manufacture and/or utilize semiconductor devices, integrated circuit (IC) bond wiring is tested for product assurance and counterfeit detection purposes through invasive and destructive probing. The examined unit is either partially damaged or fully destroyed during these tests and the uncertainties that existed prior to testing reappear when a new unit must replace the probed unit. Because packaged circuits serve such diverse roles in countless critical systems across many applications, there is a strong need for robust, quick, and non-destructive testing. As of now, methods to non-destructively test these components involve either simplified geometric modeling and finite element analyses, which make concessions to accuracy, or include more accurate forms of geometric acquisition but remain untested, unverified, and computationally expensive. The goals of this study are to test the validity of micro-CT as a tool to import accurate bond wire geometries to single- and multi physical finite element testing and to produce a practical methodology for the image acquisition, processing, and simulation of integrated circuit bond wires with a focus on practicality and industrial applicability. A reverse engineering technique is examined as a valid simplification to the geometries retrieved from micro-CT. The reverse engineered geometry from micro-CT is then tested within a finite element simulation with the loading data gathered from a traditional destructive bond wire pull-test to examine its similarity. The results show that the proposed methodology can closely mirror the destructive test by highlighting the correct location of probable failure with the corresponding stress values in excess of the material's strength limits. In addition, the methodology reduces the finite element computational expense by a factor of four and produces a CAD editable model for geometric alteration or other finite element testing environments; similar to the files created by part manufacturers prior to production. The differences being that the model can include production process-related variations and can be utilized by an end-user seeking validation for a given application. The broader implications of this methodology include its application to iterative product design and extension to multi-physical, dynamic, and/or inordinately expensive testing conditions.