## Friday, September 6, 2013

### Use of the Wedge Method for Emissivity and Reflection Independent Temperature Measurement

By: Ralph Rudolph
R. Rudolph Consulting LLC @
www.temperatureconsultant.com

A technique called the Wedge Method or Roll Nip method is finding increased use in measuring strip temperatures in the metal production/processing industries as it is touted as providing the dual advantages of appearing to be independent of the material emissivity and the presence of any ambient reflected radiation. Basically, the concept is quite simple: Picture a horizontal steel strip that contacts and at least partially wraps around a large roll, usually a deflector roll used to change strip direction or a bridle roll used to set strip tension. Aim a radiation thermometer almost parallel to the strip into the gap formed between the roll and strip tangent point, as deep as you can go. (Viewing at an angle from the side is fine). This gap, as the claims state, can be treated as a blackbody with an emissivity of 1.0 (see Figure 1). Hence, you don’t have to worry about ambient radiation as reflectivity is 0.0 and you don’t have to worry about changing material emissivity. This is partly true and partly wishful thinking.

Blackbody conditions exist for a cavity if and only if all sides of the cavity are at the same temperature. If the roll being used has a very low thermal mass (heats up easily) and there is a large wrap around the roll and sufficient strip tension to allow heat transfer to occur between the strip and the roll, then the roll will heat up to near strip temperature over a time period, but because the roll has natural convection, conduction and radiation losses, the roll can never quite reach the strip temperature. Emissivity never reaches 1.0. It should be obvious that if the strip abruptly changes temperature, which can happen with strip thickness or furnace temperature changes, it will take time for the roll to change temperatures. Heat transfer between the two can take quite a while during which time the temperature reading from the wedge system will be quite inaccurate.

Figure 1.

So, if a system is designed well, with a major roll wrap, low thermal mass roll, sufficient strip tension and steady long term operation (no major changes in strip temperature), this method can work as claimed (except that emissivity must be set somewhat lower than 1.0 to compensate for the roll being at a slightly lower temperature than the strip).

Given human nature, however, I’ve seen numerous instances where folks have not understood why the wedge method can work and who have misapplied it. Believe it or not, I’ve seen an instance where a so-called wedge method has been applied with zero roll wrap, with the strip simply passing over a support roll. And this system was (unfortunately) designed by the equipment provider who should have known better. I would guess that a majority of wedge method applications that I’ve seen have been poorly designed, with little attention paid to the amount of roll wrap or roll material and with little understanding of what occurs during changes in strip temperature.

There is a modification to the wedge method that can provide a significant improvement: Mount a second Radiation Thermometer to monitor roll temperature and compare this reading to that of the wedge RT. Using a PC with input and output cards (and most any older PC will work), abrupt deviations between the two readings which occur as strip temperature changes can be used to correct for errors. If accuracy is desired, it’s well worth the extra expense. You get what you pay for.

1. hi Ralph,

yes, I can confirm that this is a relatively slow measurement system, but it works well for a lot of aplications.
I wander how much emissivity deviates from unity.

TATA steel R&D, Netherlands

2. Hello Ralph,

Didn't realize that you were still active and not retired in AZ. As you know we used this very successfully on several product lines in the Cleveland and Indiana Harbor Annealing furnaces, even in the Aliquippa Tin Mill Anneal line, all parts of the now-defunct LTV Steel.

I think Acelor-Mittal runs all, if not most, of these lines now.

The original work on this technique was reported by the Japanese steel company research group at NKK Steel and is in their published article from the 1992 International Temperature Symposium in Toronto. If anyone would like a more detailed reference, contact me at grp@MrPyro.com.

And, Yes to Johan. You can monitor the emissivity or cleanliness of the strip. I did it many times at the continuous anneal line in Cleveland OH works at the preheater exit where we had a dual 1.6 micron pyrometer set up. The use of a "short waveband" RT was a formidable asset since the effect of slightly imperfect blackbody conditions was minimized to the point where we could use a constant emissivity setting of 1.0 and only had to neglect conditions when strip at a different temperature came along.

The roll -strip temperature equilibrium occurs relatively quickly, faster than the furnace control system could respond.

The emissivity can vary quite a bit. It is, of course, wavelength-dependent. The real emissivity changes are easily seen during a run an one thickness strip. However, when the gauge changes, there is delta-T, time dependent change that obscures the emissivity change due to the thermal mass variation. But one can see it "fade" away with time; again fairly quickly.

Ray Peacock
Temperatures.com, Inc.

3. This is very cool, thanks for sharing this! We are currently doing some temperature measurement research of our own, so we have been using temperature measurement devices to help assist in our current project. It is something that I have always been fascinated in so that is why I am currently on a project that allows me to use this technology.