Thanks; I found that quite interesting.

>Hilmar;
>
>If you are a bit interested in Metrology read the following otherwise skip it.
>
>When I was 17 I was very interested in the history and application of units of measurement. I ended up spending 15 years as a Metrology Engineer (Electronic and Mechanical). There are few absolutes when it comes to measurement standards. The source or definition changes with time as our abilities to understand and conduct measures are refined.
>
>By the way in case you are not familiar with the term Metrology, it is the Science of Measurement. Physical measurements for mechanical objects are done in a stable 20 C environment held to a very small number of deviation and is defined by the requirements of the object being measured. Electronic measurements are done in an environment of 23 C.
>
>In all cases objects to be measured are temperature soaked to insure they are stable within the test environment. The amount of air circulation is closely controlled. The environment is created by having a well-insulated room with an excellent heater and air conditioner, which run 24 hours a day and their outputs, are run through a common vent. The vent is controlled by a sensitive valve, which switches between heat and air conditioning. Sensors throughout the laboratory sense temperature variations such as a person walking into the room and cause the system to adjust accordingly.
>
>A need for metrology exists for any manufacturer to comply with International Standards of Measurement. To measure any physical phenomena in a manufacturing environment you use what is called a working standard.
>
>All objects made by man are imperfect and have what is called “tolerance”. You can call it slop or whatever seems suitable. A gear may have a tolerance of .001 inch between the spacing of its teeth and the device to measurement this value must be stable, repeatable and calibrated to a standard that is 4 to 10 times more accurate then the measurement specification. This is termed a “working standard”. A secondary standard is used to calibrate a working standard and a primary standard (typically kept at a National Laboratory) is used to calibrate a secondary standard. Primary standards are what Metrologist’s and Physicists work with to describe how a measurement quantity is defined and derived. This is important, as any one in the world with this definition will be able to produce the same result. That is the goal.
>
>
>Additional considerations when working with standards used for measurement include tolerances and repeatability and uncertainty. If you want to measure something you have to take into account factors when humans are involved. Take ten people and make a critical measurement and you will get ten different results. If even one person makes the same measurement the results will likely be different. Therefore Statistical Analysis is used to average the measurement errors and then measurement accuracy is given. So you might have an object, which is said to be 1 inch plus or minus .001 inches at 20 C.
>
>I hope the above was not too boring as when a person has a passion or a strong interest in something they are “different” then others not acquainted with that discipline. I guess there are all kinds of geeks! :)
>
>Tom