## How to measure the temperature of tungsten wire?

Every fusor and fusion system seems to need a vacuum. This area is for detailed discussion of vacuum systems, materials, gauging, etc. related to fusor or fusion research.
Jeroen Vriesman
Posts: 276
Joined: Mon Feb 07, 2011 9:08 pm
Real name: Jeroen Vriesman
Location: Netherlands
Contact:

### How to measure the temperature of tungsten wire?

Any ideas on how to measure the temperature of a glowing Th-tungsten wire without buying expensive equipment?
For activation I need to know when certain fixed temperatures are reached ( 2800, 2200 and 2000 K ).

How (un)reliable is the resistance as a temperature indication?

Rich Feldman
Posts: 1335
Joined: Mon Dec 21, 2009 11:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

### Re: How to measure the temperature of tungsten wire?

The ratio of hot resistance to cold resistance can do the job for you,
if lead resistance is taken into account where necessary.
Tungsten wire doesn't have a super high temperature coefficient,
it just has an exceptionally wide range of useful temperatures.

That's subject to finding resistivity vs temperature tables for your specific W-Th alloy.
I do know that W-Re alloy wire is made in several standard, well-characterized %Re values, for use in ultrahigh temperature thermocouples.

If the wire at unknown temperature is not conducting electricity,
you can match its luminance and/or color with a reference
such as an electrically or optically dimmed lamp filament.
That's the principle behind classical optical pyrometers.

References such as CRC handbook teach that incandescent tungsten filaments
have a brightness temperature slightly lower than the true temperature, because emissivity is less than 1.
They have a color temperature slightly higher than the true temperature, because that emissivity is spectrally tilted in the blue direction WRT a blackbody spectrum.

Here's a hypothetical experiment based on that conjecture. Suppose we use a microscope (and neutral density optical filters) to observe a thick glowing filament that was made with a tiny window into an internal cavity. The latter approximates a black body better than an ordinary surface. I think the port would appear brighter, but redder, than the surrounding metal.
All models are wrong; some models are useful. -- George Box

Roberto Ferrari
Posts: 349
Joined: Tue Mar 02, 2004 5:21 pm
Real name: Roberto Ferrari
Location: Argentina
Contact:

### Re: How to measure the temperature of tungsten wire?

Hi Jeroen

I dont´t know why my earlier post didn't gone through.
At Books and References posted an old article.
I did carburization of Th-W filaments many years ago. You need to measure very precisely the resistence.

Frank Sanns
Posts: 1722
Joined: Fri Jun 14, 2002 6:26 pm
Real name: Frank Sanns
Location: Pittsburgh, PA USA

### Re: How to measure the temperature of tungsten wire?

Back in the day, K temperature was measured against a heated filament. You would look through the device and turn a wheel that heated the internal filament until it matched the filtered image you were trying to measure the color temperature of. It worked really well.

I guess you could calibrate a standard incandescent bulb at various voltages so you knew the response curve. Then all you would do is dial in the voltage to make the filament glow like the one you are testing. Then look on the chart of what temperature that voltage gives. That way you only need find a color temperature meter for a quick calibration then you will have your own setup.

prestonbarrows
Posts: 211
Joined: Sun Jun 24, 2012 5:27 am
Real name:

### Re: How to measure the temperature of tungsten wire?

The resistivity values for tungsten are widely available in the literature thanks to the incandescent bulb and vacuum tube industries. The limiting factor here is the specific alloy any any doping will change this value to some extent. Also, the exact length and diameter of the wire may be hard to measure if you are using pre-made coiled filaments.

If you are using pre-made filaments and you know who made them, the manufacture should be able to provide you some basic current vs temperature curves for ideal cases.

Exactly how long the filament is and how it is mounted is also important to consider. In vacuum, conduction through the wire itself is often the dominate cooling mechanism and there will always be a temperature gradient with the end terminals being hotter than the center. Similarly, the geometry and temperature of the walls surrounding the filament as well as how shiny they are will significantly affect the radiative cooling losses.

Anything above 1200C or so will be a blindingly bright white so that is an easy first indicator. You won't be able to directly measure those temperatures by contact probes. Optical measurements are possible with the right expensive gear if you can get your hands in it, most optical pyrometers are tuned for lower temperatures. The trouble is most vacuum viewports are opaque in the IR band, but at these temperatures there is significant emission in the visible bands if you have a proper spectrometer. Another crude semi-destructive method I've used for non-emitters is to place a small material sample in contact with the filament. Molybdenum melts around 2800K Chrome is about 2200K and iron is about 1800K. Again, this is muddied by the gradients and losses and the reality of how to mount the sample, but if you can melt a sample you know for sure that at least part of the filament is hotter than that temperature.

Put simply, it is very difficult/impossible to calculate exactly what temperature you will get at a given current a priori. You can work out some bounding values but it usually comes down to empirical testing on your particular device in situ. Luckily, one typically does not care about the filament temperature directly, it is simply a means to an end, so you can skate around the issue.

You mention activating the filaments so I assume you are using them as a thermionic electron emitter? If so, you basically want to just keep stepping up the filament current while you monitor the electron emission through your bias circuit. When the filament activates, there will be a pronounced increase in emitted electron current. Keep ramping the filament current up until you get the performance you need. Try to keep good notes so you can replicate the setup (especially if you are winding your own filaments) when the filament inevitably fails as filaments do.

Measuring the resistance with your power supply is very repeatable, but difficult to pin down to an exact temperature in kelvin. Usually you just forget about the exact temperature, find a current (or resistance, same thing) that works for the application and just shoot for that.

Jeroen Vriesman
Posts: 276
Joined: Mon Feb 07, 2011 9:08 pm
Real name: Jeroen Vriesman
Location: Netherlands
Contact:

### Re: How to measure the temperature of tungsten wire?

Hi Preston,

activation by measuring the emission at high temperature is not working, when the wire gets hotter, there is a huge increase in emission, but that drops within about 20 seconds because the Th evaporates.
Ideally the operation temperature is where the diffusion of the Th is enough to compensate the evaporation.

When cooling down after activation to the operational temperature, first the emission drops to almost nothing: 5uA (no Th monolayer), the it slowly rises to a reasonable value, about 300uA.
But I would expect about 5mA, so activation succeeded a little bit...

This is just "flashing" for 60 seconds (during this the emission rises to about 40mA in the beginning of the flashing and then drops to a few mA ).
No carburization yet.

Roberto: I'm very interested in any info about the carburization process.

I do have a manual monochromator, but the flashing time is too short to operate it in a reliable manner and determine the temperature, but maybe I can automate it with a stepper motor to find the peak, maybe combine it with resistance measurements.

Frank Sanns
Posts: 1722
Joined: Fri Jun 14, 2002 6:26 pm
Real name: Frank Sanns
Location: Pittsburgh, PA USA

### Re: How to measure the temperature of tungsten wire?

Here is what Rich was talking about and I was trying to explain. They work really well.

http://www.instrumentationtoday.com/opt ... r/2011/08/

prestonbarrows
Posts: 211
Joined: Sun Jun 24, 2012 5:27 am
Real name:

### Re: How to measure the temperature of tungsten wire?

Here is a good paper on tungsten filaments and in particular activating thoriated emitters and the carburization process.
EDIT: Looks like you already have this posted elsewhere. Cheers

Roberto Ferrari
Posts: 349
Joined: Tue Mar 02, 2004 5:21 pm
Real name: Roberto Ferrari
Location: Argentina
Contact:

### Re: How to measure the temperature of tungsten wire?

Hi Jeroen

Give me some time to dig for old notes...

Roberto Ferrari
Posts: 349
Joined: Tue Mar 02, 2004 5:21 pm
Real name: Roberto Ferrari
Location: Argentina
Contact:

### Re: How to measure the temperature of tungsten wire?

Hi Jeroen
The purpose of the procedure is to cover a Th-W filament with a W carbide layer. As the carbide layer grows, also builds up the fragility of the carburized filament, so it is better to have the filament already mounted in order to not manipulate the wire after the procedure.
The carburization is achieved by pyrolysis of a hydrocarbon with the filament heated.
In case of preparing a small piece of Th-W, the best is to run some tests.
Design a chamber as per fig 3 of Horsting. Optimum concentration of hydrocarbon is 0.30 g C6H6 per 10 liters H2.
This design uses a small belljar open at low side to atmosphere but an alternative procedure with a mechanical pump would be helpful. You can flash the filament in vacuum and at carburization process drag the vapors.
Using info at Kamal_Jeeth, characterize your filament. This will require precision measurement of V and I.
You will be able to correlate a temperature with a resistivity.
Apply a V adequate to bring the filament to 2800 K for approx. 10 seconds. This flash will reduce some of the Th oxide to Th metal.
Then reduce the current to bring the filament to 2200 K, let it stabilize for 1 min.
Then admit a flow of benzene, reading carefully the resistance. A raise in the resistance would be sign of carburization. Probably the carburization time will be around 1 min max.
As soon as it is raised by a 10%, close the vapor flow.
Let it cool down.
Operate it at 1930 K.
Attachments
Kumar_Jeeth.pdf