checking the cal of a NIR moisture meter

[omby]

We have 'inherited' a NIR moisture meter that measures the moisture of granular material on a conveyor belt. The belt itself, whilst covered over above and to the sides is basically outside, and subject to wind gusting along its length.

The setup in the manual is simple enough. Place a test sample of known moisture under the measuring head (sample from our technical lab) enter the error between the sample and the measured value and the meter corrects itself.

Ok in theory, but as soon at the sample was put on the belt, its moisture reading started to change and got worse with time - I assume due to the breeze affecting the samples moisture content.

Is there some material I could enclose the test sample in that won't affect the IR measurement but will protect it from environmental conditions??? or does anyone know if sealed test samples are commercially available, or any other ideas? The instrument has a seperate measuring head and control unit, so bringing back to the workshop is impractical.

 

[m777182]

NIR moisture meter measures almost exclusively surface moisture. If your bulk material is hygroscopic your NIR instrument will drift -on a static sample. When material starts moving on a belt you have at each instant a "fresh " sample with unknown moisture content. Because the setting of the instrument takes some time there is maybe a solution that we used many years ago: we put a sample of bulk powder onto a large pan on the record player so that the linear velocity of material was the same as on a belt but we were able to make chemical analysis of water content in few minutes intervals. The rate of moisture absorption was then estimated and we were able to calibrate the instrument.
Some instruments provide compensation of the air humidity in a cone of the air that the light beam transverse from the light source to the sample so that surrounding air would not influence the measurement. Check yours!
Another point we observed was that the instrument was sensitive to the type of the bulk surface: if it was concave towards the light beam it concentrated the reflected beam and we got amplified signal.In turn, by convex shape the beam was weaker and all those things were not written in the manual.
Good luck
m777182

 

[omby]

Thanks.

Thats a good point about the profile of the surface of the material. If and when I get satisfactory readings I will have to experiment....
I was thinking of having a 'scraper' on the belt before the measuring head anyway, as the instrument is quite particular about the distance between itself and the sample. This would provide both the right measurement distance and a flat surface.