The replacements for Opticals are very capable devices. Some are just as accurate, if not more so. Plus they are a heck of a lot easier to learn and to use. Even more they have all the newest technologies from detectors to digital outputs.
Today Opticals are but a fond memory in the minds of some who used them extensively. Progress rests on technology and today’s high-temperature portable noncontact temperature sensors are nothing short of amazing. Some even have Bluetooth wireless capability!
When the Minolta-Land Cyclops 52 first came on the market in the early 1980s, Optical Pyrometers, especially those made by Leeds & Northrup Company (L&N) were the most popular and widely respected, high temperature thermal radiation measurement devices in the world. (L&N evaporated as an organization a few years later). I was in the middle of it, working for Land at the time and saw the dramatic changes at the very beginning.
The Cyclops was succeeded by several more advanced models and now the latest models are made by Land alone. Minolta sold the manufacturing rights to them as a result of a re-organization within Minolta Camera Company.
Back then, even a cursory comparison of the respective Manufacturer’s Specifications for the two devices revealed a startling fact: the makers of them claimed about equal temperature measurement uncertainty under laboratory conditions for the two devices.
If one added in the extra signal handling capability of the Cyclops, it was clear that a real revolution was underway in the market.
For your information here’s the current specification of the Land Cyclops 100:
(Note: The Cyclops 100/100B replaces the following models: Cyclops 52, Cyclops 152, Cyclops 152A, TR-630 and TR-630A)
Specification
Temperature range: 550 to 3000°C/1022 to 5432°F
Indication: 4 digit LCD in viewfinder, external backlit LCD display
Measuring Modes: Continuous, Average, Peak and Valley
Datalogging: To iPAQ or laptop/PC running DL-1000v2 software. Wired or wireless Bluetooth connection (C100B only)
Optical System: 9° field of view with 1/3° (180:1 to 98% energy) measurement area. Eyepiece adjustable -3.75 to +2.5 diopters
Focusing range: 1m/39.3in to infinity
450 to 620mm/17.7 to 24.5in - with optional close-up lenses
215mm/8.5in fixed focus - with optional close-up lenses
Target size: 4.8mm at 1m//0.19in at /39.3in
1.8mm/0.07in - with optional close-up lenses
0.4mm/0.016in - with optional close-up lenses
Spectral response: 1µm with advance spectral filtering
Emissivity adjustment: 0.10 to 1.20 in 0.01 step graduations
Response time: 30ms
Display update time: 0.5s
Accuracy: <0.25%(K) of reading
Repeatability: <0.1%(K) of reading
Operating temp. range 0 to 50°C/32 to 122°F
Power requirement: One MN1604/6LR61/PP3 battery
Output: RS232, Bluetooth (C100B only)
Weight: 0.83kg/1.8lb
Sealing: IP54/NEMA3
Standard accessories: Lens cap, protection window/filter, battery, wrist strap
Optional accessories: Close-up lenses, Datalogger DL-1000, HP iPAQ, rugged waterproof carry case
You would think that a highly accurate temperature measurement device like the Optical Pyrometer would excel at measurements looking into high temperature furnaces, wouldn’t you? Lots of people believed that, too not so very many years ago, like in the 1970s.
Then the really serious workers, those involved in trying to optimize the trade-off between lifetime of the special alloy metal tubing used in petrochemical thermal reformer furnaces and production throughput began to look a little closer at the measurement errors involved.
The workhorse instruments, Optical Pyrometer (”Opticals”), suffered from a difficulty that all non-contact temperature sensors share: it is invariably quite difficult the measure the temperature of the object by any other means.
The fact is, people were getting by with the assumption that “Optical” were infallible.
The first person, that I am aware of, to put some good measurements and theory together on the subject was Dr. Tohoru Iuchi, then the temperature measurement specialist for the Nippon Steel Company in Japan. His topic was a slightly lower temperature but equally knotty measurement problem: moving steel strip in a reducing atmosphere furnace.
(The last time I met Dr. Iuchi was nearly two years ago when he gave a paper at the 2006 ThermoSense Conference in Orlando Florida. He gave an excellent paper on Radiation Thermometry of Semitransparent Silicon Wafers. He was then a staff member of the Sensor Photonics Research center at Toyo University in Saitama, Japan.)
In fact, most of the measurement advances made under this particular set of measuring conditions were made by those addressing the measurement problems in both reducing atmosphere annealing furnaces and ordinary steel reheating furnaces in the late 1970s and 1980s.
Many of the solutions for steel furnace measurements are described along with appropriate references and equations in the book, “Applications of Radiation Thermometry”, Eds. J.C. Richmond & D.P. DeWitt, that may be purchased even today from the Digital Bookstore at the ASTM International website (Click on the link below).
It includes a Chapter by Dr. Iuchi describing many of the devices developed by his group at the Nippon Steel Company’s Research Center.
The principle measurement problem correctly identified by Dr. Iuchi, was the effect of reflected thermal radiation coming from from sources hotter than the object being measured. Both the emitted radiation and the reflected radiation are received by the Pyrometer.
The poor Pyrometer can’t tell which is which… so, it adds them up.
What’s worse, the highly nonlinear properties of thermal radiation mean that just a little from the higher source is very much more than that emitted by the lower temperature object.
Even without an emissivity correction, in most cases, an object in a hotter environment will appear to be at a higher temperature than it actually is to an Optical Pyrometer.
Under ideal furnace conditions, when the furnace walls and roof are at about the same temperature as the object, the reflected radiation from the furnace components makes up for the imperfect emission properties of the object and enables one to make a temperature measurement without need of an emissivity correction.
So, the next worse thing about the nature of reflected thermal radiation, is that unskilled operators often attempt to make an emissivity correction to the reading they get when they have a little knowledge of the measurement physics.
When reflected thermal radiation, even from an equal temperature source, is present, the emissivity correction skews the reading higher than true. When the reflection source is significantly hotter than the object, an emissivity makes an already higher than true into a much higher than true one.
In 1986 a special symposium was held in Houston Texas, sponsored by the American Institute of Chemical Engineering (AIChE), entitled “Measurement of High Temperatures in Furnaces and Processes”. It was chaired by Drs D.P DeWitt & L. F. Albright of Purdue University.
The proceedings were published in The AIChE Symposium Series, No 249, Volume 82, 1986.
In the meeting, Dr, Hoyt Hottel of MIT gave an overview of the measurement problems and T.R. R. Beynon & R. Barber of Land Infrared described the choices of optimum measurement wavebands, describing in some depth the experimental work done by Land both in Europe and the USA of Steel Mill reheating furnaces and simulators. The published Proceedings captured all the talks and a very interesting panel discussion.
The problem was recognised by many organizations in the Oil Industry and a unique instrument was developed and patented by Exxon Research to help solve it and get around the limitations of the Optical Pyrometer.
A special unit now manufactured by the Pyrometer Instrument Company under license from Exxon called The PyroLaser is used effectively to correct for emissivity and attempts to solve the measurement problem of reflected radiation in such Petrochemical furnaces.
The Quantum I Portable Laser IR Thermometer, now made by Mikron Infrared, is a similar device and was designed by the original inventor of the patented Exxon Research device.
A competing Land Instruments device, the portable Cyclops 390B Furnace Pro uses a different approach by using an optimum measuring waveband, where estimated emissivity errors can be more readily tolerated.
The Optical Pyrometer has too short an effective wavelength (See ASTM STP 895 and the AIChE Proceedings, described above for an excellent article by E. A. Nutter that describes the term ‘Effective Wavelength’ and choice of optimum wavebands by Beynon & Barber).
The errors made by Opticals when attempting to measure tube surface temperature in thermal reformer furnaces were higher than with any other type of noncontact temperature measuring device then on the market! Now there are other ways to do the job better.
Also, at last, someone has done a very thorough job on pulling together all the loose ends involved in solving the measurement problem and written a book about it. In it both the near IR and Mid IR measurement approaches are described in detail and the problems of other verification means are described.
The person is an objective worker in the field, Dr. Peter Saunders, a well-known physicist who works at The Measurement Standards Laboratory (MSL), New Zealand’s national metrology institute. His book, Radiation Thermometry: Fundamentals and Applications in the Petrochemical Industry (SPIE Press Book) by Peter Saunders, was just released 3 August 2007 It is available from the SPIE Press, ISBN: 9780819467836, price for SPIE Members: $39.00(USD), and Non-member: $47.00 (USD). It’s the latest, but possibly not the last word on this measurement problem area. But after skimming through the chapters, I think it will be a while before anyone digs as deeply into this area of measurement.
Bottom Line:
Don’t try to use an Optical Pyrometer to measure temperatures in Thermal Reformer or Thermal Pyrolysis Furnaces, Steel Mill Reheat and Anneal furnaces and possibly other high temperature furnaces unless you have a very good idea of what to expect!
I have only got a little more to say about the shortcomings of Optical Pyrometers and it is better left for another time when I can add a little balance by telling what’s right with them, too.
Even More Limitations of Optical Pyrometers (Chapter 3)
This problem is one that affects a smaller number of users, but to them it is significant because it limits their ability to use the “power” of Optical Pyrometry to gain precise AND repeatable temperature measurements of their products. The limitation is , of course, the subject of this brief note, speed of response.
How quickly does an optical pyrometer respond?
Sounds simple, but it is not and does not have a simple answer. Like a lot of other parameters affecting Optical Pyrometers, the answer depends on the conditions prevailing at the time of measurement AND the properties of the actual Optical being used.
Suffice it to say, the response time for following a step change in apparent temperature is of the order of 1 to 10 seconds, possibly longer, for an Optical Pyrometer in the hands of a skilled user, regardless of how you define the response time (e.g. exponential response time, 90% response, or 95%, 98% or 99% response)!
That sort of time domain is a very long one for someone interested in monitoring temperatures that change in fractions of a second, let alone trying to control the temperature of a product in a process where the product temperatures are varying with random time swings in tens of milliseconds to tenths of seconds.
Fast temperature changes will add averaging errors of random amounts to any reported temperature values made with a “slugged” or much slower-responding sensor, and can greatly increase the overall measurement uncertainty of any reported, nominal average temperature.
Those sorts of errors occur in manufacturing and research cases where temperatures are driven by the object’s heating or cooling rate, rapid variations in the spectral emissivity of the object and the occurrence of unpredictable obstructions that randomly attenuate the passage of thermal radiation into the instrument.
Such events occur regularly in high temperature processes such as refining and casting of ceramics, liquid metals like copper, brass, gold, iron, nickel, silver and steel.
Note that glass is not covered here, simple because the melting and refining of liquid glass usually takes place in relatively slow processes in furnaces which are at temperatures close to that of the molten glass. (Some glass forming processes do run much faster, but they are also special cases for optical temperature measurement, most best treated one at a time)
Glass temperature measurement in melting and refining sections of industrial furnaces is one of the many furnace application areas for Optical Pyrometers with a long history of successful use. Even in those cases, closer examination of conditions can show other limitations of Optical Pyrometers. I’ll get to them a little later.
High speed temperature events can and do occur in the many processes used to shape and form products made from such materials where temperatures high enough to warrant use of Optical Pyrometry. These include, among others, heat treating, reheating, hot rolling, forging, extruding and welding.
One final comment on speed of response has to do with the few cases of “Automatic Optical Pyrometers”.
There have been a few on the market over the past thirty or so years. Pyrometer Instrument Company used to make one and some may still dwell in storerooms that somehow these days seem to connect to Ebay! The basic design was one by E.A. Nutter and is described in design detail in a paper he gave at the Fifth International Temperature Symposium in Washington DC in 1971!.
Spectrodyne in Pennsylvania makes one even today (not to Gene Nutter’s design, but similar to the one developed at L&N in their waning days of existence).
The problem is that even these automatic devices either have a time response that is no faster than a manual instrument or the specification in not provided.
On the other hand, Optical Pyrometer users should not feel put upon by reading these pages. I strive very hard to be as objective as possible based on my knowledge. The fact is Opticals have been bypassed by technology!
Superior and inferior alternatives to Optical Pyrometers exist on the market today, but since there are no standards for any of these noncontact temperature measuring devices, it can be a swampy morass to explore. I am aware of at least two companies that makes an equivalent wavelength response model with through the lens sighting, built-in spectral emissivity control and shorter response time than traditional units. They all use absolute, calibrated photo detectors and not the comparison and feedback method of Opticals.
The Model M90V from Mikron Infrared and Chino Instruments Model IR-AHU both are modern, automatic radiation thermometers that operate at the 0.65 micrometer waveband region. The Chino unit will measure as low as 900 Deg C, while the Mikron goes about 100 degrees C lower to 800.
To add even more fuel to the competitive spirit, Land Instruments offers a very unique handheld model, that they call the Meltmaster model. It operates at even a shorter effective wavelength of 0.55 micrometer. Since one of the biggest claims to accuracy in the Optical Pyrometer’s history is a very short wavelength, one would think that a 0.55 micron instrument would be more accurate that a 0.65 micron one.
Unfortunately, that is an over simplified assumption. One thing for sure is the fact that all these devices have a 0.5 second response time, not a heck of a lot faster than the manual Optical, but none the less, faster and automatic.
I’ll get into over simplifications next time and hope to make the issue a little clearer.
Optical Pyrometers were used very much in the development, design and manufacturing of electronic vacuum tubes. Very few are made today, but even in those uses, there were severe limitations to “Opticals”. Not the least was the fact that most of the measurements were made on heated components within a vacuum.
That meant the user had to correct for the spectral emissivity of the heated component as well as correct for the loss in apparent emitted radiation due to the transit of the thermal radiation through the glass envelope of the tube walls. Some smart scientists and engineers knew how to do the corrections and since they were almost always the same materials in any given tube design, the correction was relatively easy to apply.
Back in the electronic vacuum tube days, of course, we didn’t have PDAs and pocket calculators, so the corrections were often set out in a chart or table. Even simpler, was to build the correction into the process parameters and target the “Effective Spectral Radiance Temperature” rather than the true temperature.
Using spectral radiance temperature was a great simplification because the principle calibration sources for Opticals were flat filament calibration lamps that were composed of a flat tungsten filament lamp with a certified spectral radiance temperature versus filament current property. There was a bit more to it than that, but that was the essence of calibration traceability for the Pyrometer Calibration Lamps or Standard Lamps, as they were often called.
Some of the problems implied by this calibration and application shortcut were the spectral emissivity correction questions, that arose often, such as:
What is the spectral radiance temperature? How does it vary with wavelength?
How does it depend on the materials of the tube especially if they are not the same tungsten as the flat filament calibration lamp?
What if they are coated like with an electron-emission enhancement material like thorium oxide?
Why can’t we just reference everything to the true temperature and forget about the complexities of spectral radiance temperature?
NOTE: I am reasonable careful to use the term, spectral emissivity, not just emissivity. There is a big difference between the two and if you do not understand them, visit the E-missivity Trail at About Temperature Sensors for more complete details. Note, too, that because spectral radiance temperature depends upon the wavelength (or effective wavelength) of the measuring device that it can be thought of as a lot like beauty-its value depends on the “eye” of the measurement device.
The true temperature read by any radiation thermometer (and an Optical Pyrometer is an example of one kind that operates in the visible region of the Optical Spectrum) can be related to the spectral radiance temperature by a relatively simple equation, if the Pyrometer effectively has a single waveband and there is an insignificant amount of reflected radiation from the object within the measurement waveband, (or problems like absorptions or emissions within the sight path between the object and the instrument).
(Turns out that an effective single waveband can be applied to just about any radiation thermometer if one is careful and has an error tolerance of about 1/2 to 1% of absolute temperature see ASTM STP 895 “Applications of Radiation Thermometry” or any modern book dealing with Radiation Thermometry ).
Anyhow, the radiance temperature equation for any radiation thermometer that works under Wein’s Approximation [(C2/lambda x T)<<1], where there are no significant reflected radiation effect: is:
Trad = (spectral emissivity) xK x e^-(c2/[lambda x T]),
By a little algebra manipulation, it can be shown that the Trad and T can be shown related as:
(1/T -1/Trad)= (lambda/c2) x Ln {(spectral emissivity)},
Where:
Trad is the radiance temperature in Kelvin,
T is the true temperature in Kelvin,
c2 is the second Planck radiation constant approximately 14388 micrometer K,
lambda is the effective single wavelength, in micrometers (microns) and,
and e is the base of the natural logarithm, 2.718281… and,
Ln is the natural logarithm, i.e. If a= e^(b) then Ln a = b.
Bottom Line:
You need and equation or graph PLUS a knowledge of the spectral emissivity of an object to get to the true temperature (and then of course you need to convert from Kelvin to Celsius or Fahrenheit when using an Optical Pyrometer.
Back in the 1970s and before, people were willing to do that to get repeatable, accurate temperatures. Not any more! Beginning in the 1980s, detector-based, portable IR Thermometers came onto the market and changed the measurement areas forever. But, I am getting ahead.
Optical Pyrometers had a problem with spectral emissivity and people had a problem knowing which values to use. It wasn’t easy.
Technology has usurped the capabilities of Optical Pyrometers in almost every measurement situation. First, let me back up and explain about them, how they work and why I think they are obsolete.
For decades, the by-word in industrial temperature measurement in many processes was the Optical Pyrometer (aka: DFP, Disappearing Filament Pyrometer, Optical, etc.). A unit well-maintained and calibrated, in the hands of a trained and skilled user could be expected to deliver accurate, non-contact temperature measurements in many process lines under specific conditions and with appropriate corrections.
Thermal processes like glass melting, special semiconductor and ceramic heating, steel annealing & reheating, vacuum alloy melting & heating were among the applications for these measurement devices that came to depend on their regular use. One of the regular uses was to check the validity of online or fixed mount radiation pyrometers (thermometers) sighted into closed end target tubes in large furnaces, such as steel mill slab & bar reheating furnaces.
In glass & glass product manufacturing, the Optical was nearly a universal standard device for monitoring internal melter, refiner and forehearth temperatures and profiles, especially in very large, regenerative/recuperative furnaces.
Another outstanding use was the measurement of liquid iron temperatures in production of cast iron parts. In fact, this unique measurement was one of the very few for which the former Leeds & Northrup Company (L&N) manufactured units with special, emissivity-corrected temperature scales for the spectral emissivity of liquid iron at the wavelength of 0.65 micrometer, the operating wavelength of most devices; typically the scale was preset for spectral emissivity of about 0.39.
A skilled user could correct for those measurements made outside furnace environments, where nearly blackbody radiation conditions prevailed, and a different spectral emissivity correction was needed. The use of a very narrow waveband in the Optical meant that a single wavelength approximation to Plank’s Law could be applied.
[Wein's approximation to Plank's law was automatically ensured due to the temperature ranges measured in most applications, since c2/(lambda x T) >>1 for all temperatures below about 5,000 K, the temperature where the peak thermal emission would be at a wavelength of about 0.65 micrometer. That means, in all measurements below about 5,000 K, that the the instrument's response to temperature is very non-linear while its response to emissivity is linear. The net result is that an optical pyrometer can be quite accurate in reading temperature despite emissivity estimates being far less accurate.]
HOW THEY WORK
They are called disappearing filament optical pyrometers because their use in temperature measurement requires the pyrometer operator to perform a visual match between the brightness of the object of measurement and a heated lamp filament while viewing the filament, superimposed on the field of view of the object,in the optical viewing system.
The optical viewing system includes filers that restrict the waveband response to a relatively narrow wavelength range near 0.65 micrometer. A wavelength region of about 0.65 micrometer is in the red, so that the images appear red.
Some examples of the visual variations seen in a measurement are shown on the web pages at About Temperature Sensors and Spectrodyne, a manufacturer of their own design and repairer of used L&N optical pyrometers (there are no new ones since L&N stopped making them more than 10 years ago). Another optical pyrometer design and an example of the visual effects in the Pyro brand is on the Pyrometer Instrument Company website.
The operator’s goal with this measurement technique is to perform a match in the brightness of the object and the filament by varying the filament’s brightness. In many instances the match can be difficult to precisely achieve, especially near the lower end of the measurement range. When a match is made, the filament appears to vanish. Thus, operator practice and experience are very necessary.
OK, so two of the limitations inherent in optical pyrometers are emissivity sensitivity and operator training and experience required.
There are other limitations. I’ll get them to them in the next post and beyond; there’s a lot to go through and I’d like to take them one bit at a time.
Meanwhile, how quickly do you think you could make a brightness match between two objects in a field of your ’scope if you had to use a manual brightness control, like a rotary control?
For news of new devices or applications visit MeasurementDevices.com where you also may post PR & New Product news at no cost, without need to register!
Entries (RSS)