My company has a Flir E45 and I am currently trying to measure a block of aluminum with different surface finishes to see how well this camera will work for us.  One of these surfaces is polished and when I try to measure the temperature, the reflection of the camera, or my reflection shows up as different temperatures.  Once I move, or move the camera the temperature changes.  I would like to be able to measure the temperature of a surface without have to paint or put tape on it.  I was wondering, what can be done to about this, or is there no way to us an IR camera for this application? 

You can never get rid of surface reflection, although in some cases you can control it this can be very difficult to put into practise.

 I believe its normally taken that if you have an emissivity of less then around 0.6 you won't get the accuracy to ensure repeatability of results.

 Unfortunately unpainted aluminium is one of the most reflective surfaces out there, and I imagen polished aluminium has too low an emissivity to work with.

I am very new to IR but I have had this same problem and I have turned off the lights and took an image.  Can someone tell me if this is a good practice?

Lee.

CHCC 

---

Idaho Thermal Imaging LLC
Votum Thermography LLC,
Meridian, Idaho
208-941-2545 Cell

Emissivty and reflection isn't something that can be explained well on a message board.

This is the sort of learning thats covered in great depth in level 1 training courses, and I reccomend booking yourself on a course asap if you plan on doing any extent of work with thermography.
 

CHCC:

I am very new to IR but I have had this same problem and I have turned off the lights and took an image.  Can someone tell me if this is a good practice?

Lee.

CHCC 

 

Unfortunately—and you are in good company among many new and old thermographers in believing otherwise—turning out the lights makes no difference in how reflective a surface is. It may change what is being reflected (cool lights rather than hot lights) but it is typically not very helpful and is definitely NOT the solution to measuring temperatures of low emissivity/high reflectivity surfaces. We have many excellent webinars that should help you understand these concepts. They are not impossible but do require a clear, simple explanation, some observation and, occasionally, some time to sink in!

---

Thermally Yours,

John
ASNT NDT Level III #48166
The Snell Group
www.thesnellgroup.com
www.thermalsolutions.org
800-636-9820

Isaac:

My company has a Flir E45 and I am currently trying to measure a block of aluminum with different surface finishes to see how well this camera will work for us.  One of these surfaces is polished and when I try to measure the temperature, the reflection of the camera, or my reflection shows up as different temperatures.  Once I move, or move the camera the temperature changes.  I would like to be able to measure the temperature of a surface without have to paint or put tape on it.  I was wondering, what can be done to about this, or is there no way to us an IR camera for this application? 

 I don't know what your needs for precision are nor do I know the environment in which you are working. It is, however, doubtful you will be able to get reliable measurements unless you have complete control over the thermal surroundings. Even then, errors may be unacceptably large.

 The problem is really two fold. The high reflection is the most obvious. The more insidious is the low emission of radiation. An analogy is that low emission is the thermal truth being whispered while high reflection is lies being screamed. All we hear is noise and it is nearly impossible to know what is really true. I suspect all the surfaces of the block, given that it is unfinished aluminum, are "lying" to you but that the polished side is "screaming the loudest."

If this is an important project and you really cannot increase emissivity, give us a call and we can discuss possible solutions ranging from some you might try to having someone come in to work with you on a consulting basis. 

---

Thermally Yours,

John
ASNT NDT Level III #48166
The Snell Group
www.thesnellgroup.com
www.thermalsolutions.org
800-636-9820

I talked to a guy at Flir and he said that placing the block in a container, like a cardboard box, and rotating it so that the aluminum reflects the temperature of the box might do the trick.  I then have to measure the temperature of the cardboard and set that as the reflected temperature in the camera and take my reading of the aluminum.  I was wondering what you all thought of this?  I tried offsetting the aluminum 30 degrees from the camera face and I was getting numbers way too low.  Increasing the angle seems to raise the temp, but what would the optimum angle be?  Also, do you know what the emissions for cardboard and electrial tape are?

I've tried something similar Isaac, going as far to suspend copper and aluminium plates from wire inside a sealed cardboard box.

I never managed to get the repeatability needed for the experiment to be considered a success; measured emissivity would vary too greatly as the metal cooled down. I put this down to the systematic error involved in putting a very hot object inside a small-enclosed space.

The card absorbs the little energy that gets radiated out from the aluminium and a proportion gets emitted back. Card is a much better emitter then aluminium, so as the metal cools down the card surface gets slightly warmer and the reflected component in the aluminium becomes a variable rather then the constant we need to input into the camera.

There is also the ‘grey-body' assumption to consider depending on how much detail you want to get into, but I've never come across an explanation of how the assumption holds up for low-e surfaces.

The way to get good results is going to be to increasing the local emissivity to a more suitable level, or use a high-emitting target. Electrical tape normally has an emissivity of 0.95.

 If you explained your application a bit someone might know a better approach.

Throwing this out there: how about drilling holes into the block?  The cavity effect allows for an increase in E and may give you suitable results.  You'd have to do some testing to ensure the hole width and depth are adequate to provide those suitable results, considering your camera's spot size ratio, typical measuring distance/angle, etc.  Also assumes you can drill into the block, of course.

I've seen a sample aluminum block with a series of holes ranging from about a quarter of an inch to about and inch and a half deep that seemed to do the trick.  The block was at room temp and the holes appeared to get warmer as they got deeper.  I'd post a sample image for you if I had a small block of aluminum, but it is a pretty easy test to set up.

Comments?
 

How about using a polished mirror to reflect off the block. Take your readings not from the block but from the mirror. Since the mirror will perfectly reflect only the block, then you may only be getting the temperature of the block.

But then again, the block might replect the mirror from itself and the mirror the block and so on.

I think the fact is that the block is so low e that it itself does not radiate it's heat. It may be an actual temperature but because it is shiny it is not radiating its temperature well enough to be measured using IR which is a technology that only measures radiated heat energy. Its just not sending out the information strong enough to be picked up. There is nothing you can do except to change the way it radiates or sends out it's heat energy by painting it for example. Or if you need the temperature of just one single spot, keep everything constant and correct when it comes to reflection and the temperature spot. Any movement of the spot will drift the temperature and any change of the reflection will change the temperature.

 I think the best thing is to stick or fuse a thermocouple to give constant temperature readings.

 

Good evening all.  This is an interesting topic for me as most of what I work with is aluminum, polished, painted, wet, dry, snow and ice covered.  While I consider myself a rank amateur compared to the senior folks here, I'll share what I've seen with the reflection issues so far.

First off, before trying to eliminate the error of reflection, consider whether or not there is data there.  In my case, I can tell if polished plate is frosted or not, because, at night for example, a polished plate facing upward will appear offscale cold, reflecting the sky temp, whereas a frost patch will appear closer to ambient. If we apply heat (say, a substrate heating mat) to remove frost, you can watch the temp rise, then peak, then abruptly drop off scale as the plate reaches a dry, reflective state.  That's good data for us as we know exactly when the cycle is complete.

If reflections are of undesirable things, one thing I do is measure the temperature of the reflection source and learn everything I can about the environment.  Since I work often in high humidity, the source temperature of the reflection, especially at distance, may not be what I want to correct for. 

 I tried to attach a shot taken in a cold research chamber.  The image is grainy due to the fog and cold, and also the fact that the imager itself was still about 25C warmer than the conditions.  The plate on the left is wet (just about to frost) and giving a good reading (based on an RTD contact check).  The one on the right is a control.  But have a look at the reflection on the plate on the right.  The plate is reflecting the wall, which is reflecting a source across the room.  The point here is not to get too fancy when trying to control reflections, as you might actually introduce errors.

 Something else that helped was setting up a spreadsheet to work with the emissivity, reflectivity and, in my case, transmissity of the air, sky temp and distant sources.  I've had pretty good luck finding the correct emissivity and temp corrections for the state of the surface as its condition changes.

 I hope this helps (and the pic made it!)

Cheers

-John

A good reflector is a bad emitter. End of story.

 If you want to image it successfully you MUST change the emissivity. There may be patterns that will not be visible if you dont. All of the other suggestions, drilling holes, creating a wedge, painting, coating with water, are simple changes in emissivity. I would suggest a thin coat of oil.

---

Bob Berry
BINDT, IRT Level 3 EL, Level 3 CV (ISO 18436)
Thermal Vision
8 Old Fair Green
Dunboyne
Co Meath
Ireland
bob"at"thermalvision"dot"ie
www.thermalvision.ie

Good thread! When possible, increasing emissivity is often the simplest way to improve accuracy of measurements. Tape and paint can work in some instances. I've seen jets painted with Crayola black water paint! Oil can work as well but it is challenging to get a uniform coating. Drilling holes only works if the entire volume of the block is uniform in temperature.

 A "reflective wedge" can also work very well; this involves creating a temporary cavity with a sheet of polished metal (copper is best but aluminum can work fine) positioned at an angle right at the surface to be measured. In the "wedge" space created between the two is a cavity that can often provide a high level of accuracy. 

 In essence this technique is similar to to one FLIR suggested you use of suspending the block inside a box. A couple suggestions: (1) use a large box to reduce internal influences, (2) use a thermally stable box (1-2" rigid insulation rather than cardboard, (3) work with a significant difference in temperature (and thus radiation) between the block and the inside of the box, (4) realize the hole through which you view the block can easily affect your readings.

 In the case of the wedge and the box you are trying to make the background uniform so that any difference in emittance shows up. 

 Honestly, and I don't mean to seem presumptuous, but it sounds like you are in pretty deep! If this is an important project, I would urge you to get some well informed, hands-on consulting advice as you will save yourself a great deal of time and effort. If you are just experimenting, have at it and enjoy learning about this amazing Thermal world we can see. 

---

Thermally Yours,

John
ASNT NDT Level III #48166
The Snell Group
www.thesnellgroup.com
www.thermalsolutions.org
800-636-9820

Here is one more example of reflections, offered not to confuse, but to show that this is NOT obvious. The cold cup is reflecting the hot cup, due mainly to the angle at which we are viewing the curved cold cup surface. The spoon, polished silver, has a very high reflectivity, and is reflecting both the hot cup and the cold cup. It is also reflecting the hot cup at the concave bowl of the spoon which "sees" the hot cup. The spoon also has a low  emissivity but we can't tell much about that as it is at room temperature. 

 If we could remember what the visual world looked like through our eyes as a one or two-year old child, we'd have a greater appreciation for the confusion we sometimes encounter in the thermal world with our IR imagers!

---

Thermally Yours,

John
ASNT NDT Level III #48166
The Snell Group
www.thesnellgroup.com
www.thermalsolutions.org
800-636-9820

Well, thank you all for the advice.  It seems my little project turned out a little more complicated than what was intended for my gauge R and R.  If aluminum is this difficult to measure, what about silicon.  The idea was to determine the gauge capability of our IR camera, and then measure the temperature of thin (.6 mm deep and 3.2 mm wide) silicon plates, which have a mirror like finish, more so than that of the aluminum I am trying to measure.  Also, these silicon plates cannot be covered or even touch with a human hand because of microscopic holes in them.  Is this going to be impossible?  If one cannot measure the temperature of shiny aluminum in a controlled environment, what hope does one have in measuring silicon when it cannot be put in a box? 

Are there any other remote temperature measuring techniques out there besides the IR camera?