January_AMP_Digital

A D V A N C E D M A T E R I A L S & P R O C E S S E S | J A N U A R Y 2 0 2 0 1 8 the process is greatly improved through automation. Using the automation of- fered by Robo-Met.3D, acquisition times for 3D datasets can be reduced fromweeks [15] to days or hours with lim- ited supervision [10,16] . Without automa- tion, the time and cost associated with mechanical serial sectioning rapidly in- creases. Further, automation improves the repeatability and consistency of each slice, making data sets more uni- form and less challenging to post- process for analysis. Compared to nondestructive char- acterization methods such as µCT and ultrasound, mechanical serial section- ing can provide equivalent or better resolution and greater characteriza- tion detail. Microcomputed tomogra- phy relies on differences in density that appear as grayscale intensities in reconstructions, but serial sectioning can provide greater contrast between disparate materials with very similar densities. Additionally, serial section- ing is not limited to grayscale imaging modalities. It has been coupled with immersion etching to develop imag- ing contrast between metal phases [17] , electron backscatter diffraction for re- vealing grain orientations [11,12] , and a variety of optical imaging modes (e.g., bright field, dark field, polarized light) to take advantage of specific proper- ties within materials. The limitations of mechanical serial sectioning include its destructive nature and substantial data collection times. Attention to proper experimental setup and process con- trol, enabled by real-time experiment and data collection monitoring [9] , is vi- tal to limit unintentional damage to samples and unnecessarily prolonged experiments. With proper setup and care, me- chanical serial sectioning can provide data and characterizations not practi- cal to attain with nondestructive 3D or traditional 2D techniques. The follow- ing case studies present 3D investiga- tions of a faulty micro-inductor, cracks in glass-to-metal seals, and porosity in additively manufactured stainless steel. Each study showcases a differ- ent volumetric footprint obtained us- ing mechanical serial sectioning and the qualitative or quantitative determi- nation of a feature generally unobtain- able by other means. These examples highlight the use of mechanical serial sectioning for evaluating multimateri- al components and singular alloys not only for academic interests, but also for specific engineering investigations with implications regarding part acceptance and manufacturability. CASE STUDIES Case study #1 – Failure analysis of a micro-inductor: As service requirements for microelectronic components be- come more rigorous and size reductions continue, manufacturing parts to meet specifications can become difficult. In safety-critical applications, compromis- es in functionality due to manufactur- ing, handling, or assembly are unac- ceptable. As part of a root cause analysis for a failed micro-inductor smaller than 1 x 1 x 0.5 mm 3 , serial sectioning was used to determine if failure was due to a manufacturing defect or an isolated re- sult of downstream handling. In this study, electrical failure was diagnosed as resulting from a severed copper wire somewhere within the mi- cro-inductor. Identification of the break- age location would help determine the most likely cause of failure. If no rem- nants of the initial, as-designed solder Fig. 2 — Failure analysis of an inductor: (a) Secondary electron image of the inductor and detached wire with serial sectioning direction noted; (b) plot shows depth into the part as a function of slice acquired; (c) from left, slices at depths of 0, 235, 265, and 318 micrometers illustrate top and bottom of initial wire connection. (a) (b) (c)