There are various methods to estimate these temperatures theoretically but in the absence of experimental data one cannot check its correctness.
Filament is of tungsten so possibly around 2200 C. May be Wiens law lambda T=0.29cmK will help. Common wavelengths are infra red and visible.
You can assess the filament temperature perfectly from the spectrum emitted (the shortest wavelength emitted equals the filament temperature). It will certainly not exceed the melting point of tungsten (about 3422 °C), usual temperatures are within 2300 .. 2700 °C - typically increasing with bulb wattage. Halogen-filled tungsten lamps typically work at higher temperatures than ordinary bulbs.
A good soure: https://de.wikipedia.org/wiki/Glühlampe (sorry but the german page is better than the english one)
You can measure electrical resistance of the filament (measure current) at low voltage applied (room temperature of the filament) , and then at nominal voltage. Then find temperature from tungsten reference data for R (T).
Dear Aleksandr A. Tarasov,
the tungsten data R(T) should depend on the type(s) of the:
1) filament(s) and the type(s) of the
2) devices (lamps).
So, where could we find some useful analytical parametric definitions or expressions for the distinct R(T) ?
Absolutely right ! We need to know only relative value of R, which is independent on filament geometry (net resistance of tungsten).
As filaments can be different, their temperature at given applied voltage will be different, but you always can determine temperature from R measurements.
There's some problem with the calculation of the temperature: the temperature distribution is inhomogenous! Cooler vs. the fixture points, usually the hottest in the middle. Inserting some rule of thumb value (Rhot = 10 * Rcold) I got a temperature in the 1800 K range. Which is certainly too low. Emission spectrum won't "lie".
Dear Prof. Agrawal,
this[1] R(T) is valid for the sag or the Non-sag Tungsten Wires ?
1. R(T)=R(293 K)*(T/293)^1.215
This is under the assumption that the temperature is uniform throughout whole wire.
So, the R(T) is the same (???).
However, the color (temperatures[1]) of the old sag (near reddish) and the "new" NON-sag (near yellow to daylight), Tungsten Wires are quite different (!).
1. https://en.wikipedia.org/wiki/Color_temperature
Dear John, of course, wire resistance R(T) changes because wire length and diameter changes under long time operation. But net resistance r(T) , which is independent on geometry, remains the same.
The both methods give us some averaged values of temperature. Advantage of R-method is simplicity. Of course, temperature is different along wire and there are "hot" areas, where wire should be broken first. So, If you want to determine local wire temperature from emission spectrum, you must collect light only from this local point at the wire.
If the wire resistance R(T) is valid, then how it could be explained that IR-pyrometers[1,2] present, for so many years (>60y), with so many remaining issues of : 1)high systematic errors[3], 2)aging issues and other 3)instabilities[3-6] ?
1. Infrared thermometer https://en.wikipedia.org/wiki/Infrared_thermometer
2. Multicolor optical pyrometer WO 1997004292 A1 http://www.google.com/patents/WO1997004292A1?cl=en
3. Infrared temperature measurement uncertainty for unchopped thermopile in presence of case thermal gradients http://www.sciencedirect.com/science/article/pii/S1350449510001398
4. Issues in high-speed pyrometry http://iopscience.iop.org/article/10.1088/0026-1394/33/4/3/pdf
5. Non-contact multiband method for emissivity measurement http://www.sciencedirect.com/science/article/pii/S1350449502001822
6. Comparison between multiwavelength infrared and visible pyrometry: Application to metals http://www.sciencedirect.com/science/article/pii/S1350449507001375
- Badges
- Science topic