? about Measurement Uncertainty

[**DONOTDELETE**]

Moderators and Administrators please transfer this post accordingly if I have placed it in the wrong section.

I had a flashlight's (custom built by McGizmo) light product output measured by a reputable organization. Terms that are beyond my knowledge base were provided to me along with the measurement value. Of particlular interest are two terms: metrologist and measurement uncertainty. I have already researched metrologist and am fairly comfortable with the definition. The second term is much too complex for my comprehension given my VERY perfunctory understanding of the light product output of flashlights and the units of measure that have been thoroughly discussed in these pages.

I have an electrical engineer on my staff, but admits to not having in-depth knowledge of "measurement science." Are there any metrologists, metrological engineers, measurement scientists, or anyboby with a thorough understanding of measurement uncertainty on this BB willing to provide an ignoramus like me a working understanding.

I am tapping other sources as well.

Thanks in advance.

 

[SilverFox]

Hello KT,

I am not an expert, but have an observation.

I find it helpful to replace the term "measurement uncertainty" with "tolerance."

A flashlight may have an output of 70 lumens (plus or minus 5 lumens). The measurement uncertainty is the plus or minus 5 lumens.

Tom

 

[**DONOTDELETE**]

I was posting in another forum while you were likely responding in this one. My POC stated that "...while tolerance and MU are related, they are two separate entities." At least that is how I remember HEARING it. I caution about my hearing abilities though. /ubbthreads/images/graemlins/icon15.gif /ubbthreads/images/graemlins/tongue.gif /ubbthreads/images/graemlins/help.gif /ubbthreads/images/graemlins/grin.gif

This is an open invitation TO my POC to post.

 

[Roy]

I'll give "measurment uncertainty" a shot. I'd say it's the point at which you cannot say what a measurment actually is. Take a ruler for example....the best you can report a measurment that is half way between the smallest indicators...ie, if the ruler only measured in 1/8" marks, the best you could report would be 1/16". Another example is graduated glassware in a chemistry lab. On all graduated cylenders(sp) there is a accurcy measurmet of +/- some volume.

To sum up....ALL measuring devices have a margin of error that should be reported and also have a measuring limit. When I was in the business of calibrating radiation detection instruments, the radioactive sources we used had calibration certificates traceable to National Institute of Standards. Because of the all errors of measurment at each stage along the way, our instruments were only cailbrated to +/- 10%.

Hope this helps.

 

[SilverFox]

Hello Roy,

I stand corrected.

It looks like measurement uncertainty refers to the limits of the device doing the measuring.

Tolerance has to do with variability in manufacturing.

These are indeed two separate entities.

I just looked up uncertainty and under the statistic definition is listed
"The estimated amount or percentage by which an observed or calculated value may differ from the true value."

Tom