Ask A Scientist

Physics Archive



Index Key:  PHY065
Author:     amj
Subject:    Light bulbs
Text:       What happens during the last 10 seconds before a light bulb burns 
out?

Response #:  1 of 2
Author:      A. Smith
Text:        From my experience, light bulbs generally seem to go out right 
after you turn the light on - actually within a second, so there is not 
question of waiting 10 seconds there.  Light bulbs contain a narrow wire (the 
filament) and as this wire heats up, it becomes more resistive.  Since current 
is the ratio of voltage to resistance, more current than normal flows in the 
light bulb during the first fraction of a second, until the filament has 
completely heated up.  This is the time of greatest stress on the wire, and 
the power (which is proportional to the square of the current) can be 
considerably higher than what the bulb uses under normal conditions.  Over 
time (even if the bulb is never turned off) the metal atoms on the filament 
evaporate and eventually weaken the wire sufficiently that it will break if 
subjected to stress.  If left on long enough, the wire can break even without 
any stress, but then it does not do anything spectacular, I do not think.  
Basically what happens is the metal gradually evaporates and at some point the 
wire breaks, and the bulb is burnt out.  If you look inside a burnt out 
unfrosted bulb you can usually see bits of the filament dangling from the two 
supports in there.

Response #:  2 of 2
Author:      Obiwan
Text:        Generally, I would agree with the previous response, but there is 
another factor which is significant.  As the metal evaporates at a point, or 
as a flaw develops (we all know bulbs often fail after shocks, especially when 
the shock happens while the bulb is lit) the resistance at that point will go 
up.  This is because the resistance of a wire is inversely related to its 
thickness.  As the resistance increases, the temperature at that point also 
increases due to I^2R heating.  As the temperature goes up, the resistance at 
that point goes up even more (resistance is inversely proportional to 
temperature).  This is a run-away situation.  The increasing resistance drives 
temperature up, the increasing temperature drives resistance up, and so on.  
Eventually, the filament melts, and the bulb burns out.  This whole process, 
once initiated, takes place very fast - it is very difficult to predict in 
advance when a bulb will burn out.




Back to Physics

This note was taken from a posting on the Electronic Bulletin Board System (BBS)
for Science, Math, and Computer Science K-12 Educators called NEWTON.
NEWTON can be reached at:
http://www.newton.dep.anl.gov or telnet://newton.dep.anl.gov