Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: [Phys-l] 50 year incandescent bulb

Jason,

"But wouldn't a beefier filament be an option?"
That will lower the resistance and raise the temperature of the filament.
That would shorten the life at a constant voltage.
"How about making some dirtier (presumably more resistive) W."
That would raise the resistance and lower the light output and temperature.
This also creates a bigger problem with the thermal affects. See below.

"perhaps some other metal with a lower evaporation at the temp of interest?"
As for an elemental metal, W is it. Carbon is a higher boiler though.

"I'm sure at some point, the inner surface of the bulb would be pretty
opaque, but that wouldn't seem to be a problem for awhile since I don't
recall any noticeable film on any bulb I've ever seen."
Yes for awhile. At lower temperatures, like running a higher voltage bulb
at a lower voltage, the evaporation rate is lower and the lifetime is
extended. Eventually tungsten is deposited on the envelope and blackens
it. That lowers the net light output you see. I did this as a kid by
using 4.5 volt bulbs in a 3.0 volt flashlight. The bulbs burned dimmer for
years but started to turn dark.
Your darkeneing statement is on the verge of thermal vacuum evaporator
processing just like evaporating aluminum or gold metal in one of these
machines.
***************

I am reluctant to post this but here goes. First, LEDs are going to kill
stinking shorter lifetime CFLs. CFLs are a scam, IMHO.

You can design a 50 year light bulb. Light bulbs are designed around
tungsten wire because it is the highest elemental boiler next to elemental
carbon. There are mainly two processes at work. The temperature and
thermal history of the tungsten filament determines life. So a lower
voltage than the rated voltage translates to a lower the temperature and
longer life. At higher temps, the grain weaker grain boundaries and
crystals of the tungsten evaporate faster. That thins the wire and raises
the resistance at that point. That starts a thermal runaway condition that
leads to filament failure. More on that later. Actually the tungsten
evaporates at much lower temps too but obviously not at room temperature.
So there is a lower temperature threshold required.
Tungsten can be vacuum evaporated in my personal vacuum evaporator in my
garage or any other vacuum evaporator. I study geometry affects of
tungsten in it but slanted towards electron guns and filaments.

Take a 100 watt bulb that runs at 120VAC. The static resistance is about
10 ohms. So the current flowing is 120V/10 ohms = 12 amps. 12 amps is
the surge current and then the filament starts to heat. That "instantly"
changes the filament resistance and it increases to about 144 ohms at
normal operating temps (~2800 °K). (Stefan-Boltzmann and all that.) So
the current at temp is 100W/120V = .83 amps. R at temp = 120V/.83 = 144
ohms. So you can run 18 bulbs.

In bulbs, they introduce a partial pressure gas to minimize the loss of
tungsten by evaporation, unlike in a vacuum. They say 25 watt bulbs are
under some vacuum.

The other process at work is erosion of the filament by gases like oxygen.
So they use argon or nitrogen as the unreactive padding gas. Even the
shape of the filament has an affect. A sharp bend in the wire increases
the resistance at that point and the diameter.
Conductivity of heat is also a factor in light emission. So bulb filament
supports are point contacting to keep the temp high. With a variac, you
can see how the temp is lower at the point of contact of a support in a
long filament. Try a 25 W bulb at around 30 VAC, as I recall.

To minimize convection heat losses, the filaments in a light bulb are
coiled or are double coiled. In an electron microscope (EM), the filament
is not coiled because it is in a vacuum and you can vary the supply voltage
at will. The sides of an EM filament are actually hotter than the bent tip
after a few hours of use. This means that the tip does not emit electrons
at the same rate as the hotter sides at saturation. <snip>

At the point of future failure in a filament, the resistance rises from the
evaporative thinning. That raises the resistance there and the total
resistance overall. That lowers the total current. So less total heat is
developed in such a filament at constant voltage. The problem is that more
and more of that lower amount of heat is being shifted to the point of
failure with the higher resistance. That means it gets hotter there and a
thermal runaway process is starting. This process leads to filament
failure.
Away from the point of failure, the heat generated is actually less. That
is why bulb manufacturers only slightly bend filaments in a light bulb to
make sure NO DEFECT SITES exist in the filament or sharp bends. Defect
sites cause hot spots and are potential points of premature failure. Since
the filament in an EM is almost at the melting point, any increase in heat
(or lowered self-bias resistor) causes a faster burn out. The steep slope
of the thermal runaway curve reaches a peak dissipation wattage of almost
THREE TIMES the normal heat dumped or dissipated at the point of failure
when the filament is new. The operating point at the top of the wattage
peak can never be reached. That would take a "tungsten" filament that had
an infinite dissipation wattage rating and much higher melting point than
W, C, or a ceramic.

I did not plot the model of a light bulb filament versus wattage. I only
did that for higher operating temperature EM filaments. But the values
should be close and you can see how the failure happens. Eventually the
higher resistance from thinning from evaporation or erosion, shifts the
heat load onto the point of failure. That makes the tungsten much hotter
and it melts at that point of higher resistance and thinning.

So any defect site can cause the failure, even if the bulb is new. Any
nonuniformity of the tungsten filament during manufacturing changes the
resistance and that can lower the temperature (higher diameter) or increase
the resistance (lower cross section, dents, kinks, cuts, etc.) at a
constant voltage.

If you run the bulb at a much lower temp, tungsten will still evaporate.
Black tungsten will form on the envelope and cut down the light one sees.
So lower voltages come at a price. Higher operating costs (per lumen),
darkened bulbs over time, less light, and a more yellowish-orange light.

Some of this is on a site at General Atomic, as I recall.

A lot of this electron gun material and heating affects except for false
peaks are covered in Michael Haine's book in 1961-62 and a joint journal
article in 1952 by Einstein and Haine. He briefly discusses the heating
affect factors and flattening of a bent tungsten wire tip.
There is an excellent US patent by Baltzer on some of these geometry
affects in the early 1990s. See www.uspto.gov. Click on patents, search,
quick search.

There is a light bulb that has been burning for almost a century. When you
see the glowing carbon filament in the bulb, you will see orange light.
You won't need dark glasses or sunglasses.

Paul
Industrial Scientist
Material Science
US patent holder



At 08:59 PM 5/1/07 -0700, you wrote:
I was wondering if that was the case...

But wouldn't a beefier filament be an option? Or is that also a
resistivity compromise for ohmic heating? If so, how bout makign some
dirtier (presumably more resistive) W or perhaps some other metal with a
lower evaporation at the temp of interest? (perhaps not, as W seems to be
the metal of choice)

I'm sure at some point, the inner surface of the bulb would be pretty
opaque, but that wouldn't seem to be a probelm for awhile since I don't
recall any noticable film on any bulb I've ever seen... but maybe I'm just
not lookign close enough...


At 08:35 PM 5/1/2007, you wrote:
At 10:21 AM -0700 on 5/1/07, Jason Alferness wrote
But then, I'm still a little confused why they can't design a 50 year
incandescent bulb...

Chuck Britton replied:
My (limited) understanding is that a 50 incandescent bulb could be
designed - but it would not give off as much light as you would
expect.
I notice that (some?) 'long life' incandescent bulbs are 'rated' at
130 volts rather than the more common 120 volts.
I.e. the filament is cooler when it is used at 120 volts and is
putting out less than the rated wattage (wattage rated at 130 volts
=> longer life due to 120 volt use)
============================================

Bulb lifetimes are limited by evaporation of the tungsten filament. At
lower filament temperatures, the vaporization rate of the tungsten
filament is reduced, but so is the light output. So the filament
temperature is a compromise between light output and lifetime.

Larry Woolf
General Atomics
_______________________________________________
Forum for Physics Educators
Phys-l@carnot.physics.buffalo.edu
https://carnot.physics.buffalo.edu/mailman/listinfo/phys-l



---------------------------------------------------------------------------
----------------

"A person, who is nice to you, but rude to the waiter/waitress, is not a
nice person. (This is very important. Pay attention. It never fails.) "

Dave Barry from "FIFTEEN THINGS THAT IT TOOK ME OVER 50 YEARS TO LEARN "


Jason Alferness
University of Washington
Room B256B, Physics & Astronomy Building
Campus Box 351560
Seattle, Washington. 98195-1560

Phone: (206) 221-2974
FAX : (206) 685-0635
email: alf@phys.washington.edu
_______________________________________________
Forum for Physics Educators
Phys-l@carnot.physics.buffalo.edu
https://carnot.physics.buffalo.edu/mailman/listinfo/phys-l