Universal Testing Machine - Calibration and Measurement Uncertainty

T

tututu

TuTuTu
hi, all
We have univeral testing machine and we have it calibrated according to ISO 7500
Some information given in Calibration certificate as follows
Class of machine : 1,0 (accuracy class)
Decleration of Conformity: Machine is classified without taking notice of measurement uncertainty
Measurement Uncertainty: (some values were omitted)
200 kN 0,42%
1000 kN 0,66%
2000 kN 0,70%

My questions are
a) What is the relation ship between class of machine and measurement uncertainty?
b) Can measurement uncertainty be higher than class of machine?
c) If so How can I do the conformity evaluation?
...According to class of machine or measurement uncertainty?

As far as I know for class 1,0 machines maksimum allowable error of the testing machine is %1 of reading force
Thanks
 
G

George Weiss

Some review of the Class of machine statement has determined that a class 1 tester has a specification of 1.00%
Your calibration stated Uncertainty listed is 0.42%, 0.66%, and 0.70%
It is generally desirable to have TUR, (test uncertainty ratios), of better than 4:1, (0.25%). You are showing a worst case here of 1.4:1, which is the test uncertainty of 0.70%
The uncertainty reported should never be more than the specification being tested.
Is this calibration ISO-17025? uncertainties being reported, suggests this. With large uncertainties, the possiblity of determining the device is working correctly or not requires a guard banding approach to evaluating PASS/FAIL of test points. The magnitude of guard banding will determine the level of confidence, false pass, and false reject possibility %
If you have a question about the source and magnatude of the calibration facility's uncertainties, then I would suggest asking if they might allow you to view, or send you a copy, of the uncertainty budget calculations they used to report the 0.42%, 0.66% and 0.70%
The tension/compression measurement indication display resolution would limit the device’s best possible accuracy, and be considered in an uncertainty evaluation.
Example: 100kN, 100.0kN, 100.00kN, and etc.
In these examples the best possible accuracy/resolution would be half the least significant digit.
In cases of high uncertainty an accepted practice is the use of the equipment with current calibration data.
This would mean correcting for the errors with correction factors developed from current calibration data.
 
Last edited by a moderator:
Top Bottom