In the recent Maintenance Tips you sent, you include the following:

"As a 13 year level III I am not sure I agree with the September 18 Infrared tip at all.

Original Tip:
http://maintenancetalk.com/blog.php/tipsblog/infrared_inspection_tip5/

Emissivity and delta T is absolutely essential. Field settings are important but there are only two that cannot be adjusted in the software, focus and perspective. Depending on the type of instrument used (long wave/short wave) determines the other parameter adjustments.

Most units I have been used have paint somewhere around .85-.9 and depending on the temperature of the surface, the emissivity also changes. Shiny copper connections can be under .3 ö a situation where the temperature could read from 100 at .1 to nearly 400 at .3. I trust you see my point.

Field testing of E is essential to reliable infrared thermography.

Thanks for the good work you do.

Regards.
Randy Springer
Pepper Maintenance Systems"

First, I want to thank Randy for his comments. Second, he is correct in that no paying attention to an emissivity/background correction setting at the time the image is captured can lead to problems later.

This is especially true if you are trying to measure temperatures of very low emissivity surfaces, as he suggests, as the correction could push the radiometric temperature off scale or into a higher range. Since "range" cannot be adjusted later, that creates a problem.

I know the person who originally posted this tip would concur as he is a highly skilled thermographer.

I would point out, however, that measurements of low-emissivity surfaces (basically any unpainted metal or metal that is not HIGHLY oxidized) are most often highly unreliable, even if the correct settings are used. Copper is one of the worst "culprits" as it has a low-emissivity even when it appears "oxidized." Basically of it looks copper-colored at all, you can forget measuring temperatures!

Anyone who has doubts about this nasty little bit of physics can demonstrate it to themselves with a simple experiment:

1. Take the change out of your pocket and heat it in a frying pan to about 200F or greater.

2. Try to measure the temperature of any coin using any corrections settings desired.

3. Try more measurements from any other angle.

4. Note that you will have substantial, unacceptably large variations and errors among your measurements.

5. Try measuring temperature differences and you'll find the same thing to be true.

Interestingly, if you simply apply a high-emissivity "target" (such as electrical tape) to any of the coins, and life becomes good again. That should be a lesson to all of us!

Somehow, I deleted the balance of my post:

So, thanks again, Randy, for your clarification. Clearly you are aware that getting good results depends on training, qualifying experience and good practice. I hope many who now, as the price of cameras tumbles downward, are jumping into the field recognize that all is not as it first appears. They would be better to take a look at the coins as I've described rather than just tossing them in the proverbial "slot machine" called "Gee, this looks easy, probably anyone can use an IR camera!"