Thermal effects can affect your results.
Its consistency of temperature, not the actual temperature, that is important.
Thermal errors can stack up.
For measurement uncertainty purposes, you want to assure that linear expansion dimensional errors attributable to temperature variation are minimized- less than 10% of your intended accuracy.
Thermal Expansion Coefficient – The thermal expansion coefficient (CTE) of tool steel is added to the measurement uncertainty calculation where relevant. The Testing Laboratory considers consistency in temperature most important. This policy was derived from MIL-STD-120 which states: “Whenever precision measurements are to be made, the temperature should constantly be kept as near to 68 degrees as possible. Since most gages and measuring instruments are usually made of steel…..the requirement that the temperature remain constant is more important than the actual temperature.”
Based on the above statement in bold, our laboratory tracked the temperature with its computerized temperature control system over a period of a month in order to determine the amount of deviation from 68 degrees. The amount of this deviation is used to calculate the Linear Expansion per unit length per degree Fahrenheit. This amount is used in the calculation of relevant measurement uncertainties.
For steel, the coefficient we used was 0.000006″ per degree of temperature change. (That’s six millionths of an inch per degree F)
For copper and copper alloys we used 0.000009″ per degree of temperature change. (That’s nine millionths of an inch per degree F.)
For aluminum, the figure we used was 0.000013 ” per degree of temperature change. (That’s thirteen millionths of an inch per degree F.)
While room airconditioning is important don’t forget that handling gages can affect your measurement system too.
Note that gaging can pick up operators body heat and that temperature errors can thus stack up…