Naked Science Forum
Non Life Sciences => Chemistry => Topic started by: Mini on 10/06/2016 14:03:09
-
Hi,
I was wondering if it could be possible that there is a positive correlation between the lifetime of a lightbulb and the amount of precipitation in that lifetime. I have some data on the lifetimes of lightbulbs and the amount of precipitation within that lifetime and it seems to be that the more precipitation there has been, the longer the lifetime of the lightbulbs there is.
Since I have no knowledge of halogen lightbulbs, I was wondering if this could be a possiblity? What do you think is the impact of precipitation on lightbulb life?
-
Does your data consider the proportion of precipitation relative to the mass of the filament, or just the absolute precipitation?
Might it not be that filaments that were thicker to begin with had more tungsten to lose in the first place, so you see more precipitation because the filament has more to give away before it fails...?
-
Precipitation of what, and where?
-
Precipitation of what, and where?
precipiation of rain, snow etc. It concerns lightbulbs that are placed in message signs above highways.
-
it seems to be that the more precipitation there has been, the longer the lifetime of the lightbulbs there is.
It is true that many equipment failure modes increase with ambient temperature.
Precipitation tends to happen on cloudy/cooler days and precipitation directly cools the exposed part of the bulb. Could the slightly lower temperature due to the increased precipitation extend the bulb lifetime?
The filament is sealed in a pretty hermetic environment.
- I expect more accidents in icy/snowy/rainy weather. I expect more warning signs will be displayed when there are accidents ahead. And I expect that the lifetime of the bulb is shortest when it is illuminated the most (and even shorter for those scrolling warnings, where the light is flashing on and off repeatedly = thermal cycling = thermal stress failures)
- ...But maybe your road operator puts up general advice notices when there are no specific warnings? This would mean similar bulb usage in wet or dry conditions.
The electrical contacts are somewhat exposed to the environment.
- Corrosion of the terminals would contribute to the failure rate, and this could be affected by precipitation. But this is the opposite trend from what you find.
- Exposure to salt spray would significantly accelerate corrosion of the terminals.
it seems to be that...
So you need to perform a statistical test. Normally for psychology, they look for a p-value of 0.05, ie there is a 5% chance that this finding could happen by accident (or equivalently, if you performed the same experiment 20 times, you would expect to find the opposite result once).
If your organization doesn't already have an expected confidence level, p<0.05 is a good start.
See: https://en.wikipedia.org/wiki/Student%27s_t-test