Tungsten Lightbulb Filament Failure

Skills: SE & optical microscopy, EDS

This was a short introductory project for my Failure Analysis and Prevention course. My team investigated the failure mechanism of incandescent lightbulb filaments, specifically evaluating the theory that grain growth in tungsten is the primary cause of failure. We obtained a failed incandescent bulb and a new one to serve as a control. Filament samples from both were extracted and mounted together to ensure identical sample preparation. The sample was grinded and polished to expose clean cross-sections suitable for microscopy. A chemical etchant consisting of 30 mL HCl, 15 mL HNO₃, and 30 mL HF was used to reveal the underlying microstructure. The etched sample was then imaged using both optical microscopy and scanning electron microscopy (SEM) to compare the microstructures of the failed and new filaments and assess potential failure mechanisms.

Microscopy revealed that the failed filaments had noticeably fewer grains — and therefore larger grain sizes — compared to the new filaments. We counted grains across 32 cross-sections for each filament which yielded a result of:

Average of 3.125 grains per cross-section for the new filament.
Average of 1.906 grains per cross-section for the failed filament.

This results in 1.74x larger grains on the failed filaments than on the new filaments. This supports the hypothesis that grain growth is a primary contributor to filament failure over time.

New Filament

Failed Filament

The images above illustrate that the failed filaments generally have fewer grains per cross-section than the new filaments. In addition to grain growth, we observed significant faceting in the failed filaments, likely caused by condensation and evaporation processes within the bulb. These faceted, thinner regions create localized hotspots, further thinning the filament and accelerating failure. While our data clearly demonstrated grain growth in the failed samples, the presence of other damage mechanisms — particularly faceting — makes it difficult to conclude that grain growth is the primary cause of failure. Given the limited scope and timeframe of the project, we determined that multiple factors likely contributed to the filament failure. A brief 1.5-page (which was the max limit) summary of our findings is available below.