Accuracy and Precision
NB: In cases where standards compliance is required, the full original current standard should be consulted
Reproducibility and Accuracy of Measurement
Reproducibility (ASTM): ASTM standard 4417-03 cites levels of reproducibility for both “X-Coarse” and “Coarse” grade replica tape. (A revision, with precision and bias data appropriate to Press-O-Film HT, is in development and will likely become available in early 2012. Preliminary results suggest that the associated, newer, testing will indicate that replica tape has much smaller levels of error than is reflected in the numbers below.)
For X-Coarse grade replica tape:
“Two results, each the mean of four replicates, obtained by operators in different laboratories should be considered suspect if they differ by more than 37%.”
For Coarse grade replica tape, the equivalent level of reproducibility is cited as 28%.
According to these ASTM criteria, the following errors are the maximum that should be expected (according to the current version of the standard:
|1.0 / 25||0.3 / 8||not applicable|
|2.0 / 50||not applicable||0.8 / 18|
|3.0 / 75||not applicable||1.1 / 27|
|4.0 / 100||not applicable||1.5 / 37|
Accuracy (NACE): NACE Standard RP0287-95 addresses the issue of accuracy of measurement, reporting the results of round robin tests in which 14 blasted panels were measured by 7 laboratories. Replica tape and focusing microscope measurements agreed within their 95% confidence limits in 11 of 14 cases. The average difference between the two types of measurement was 0.2 mils (5 microns).
Reproducibility and Accuracy (Testex): By using a series of surfaces machined to known roughness, it is possible to test the accuracy and reproducibility of profile measurements made with replica tape and a micrometer gage. The surfaces used in Testex' internal testing were produced by making casts of grooves of known depth. The casts then consist of sharp-topped parallel "V-shaped" ridges of known height.
A blasted surface is highly irregular, unlike the regular grooved structure used in our tests. For this reason, the proper way to assess the accuracy of replica tape for evaluation of blasted surfaces is via the techniques used by the standards organizations concerned with such surfaces. In general, measurements of profile for highly irregular surfaces will be a strong function of the methods used to measure them.
Nevertheless, tests using casts of machined (as opposed to blasted) surfaces show micrometer gage measurements of Press-O-Film to have an accuracy and reproducibility of 0.25 mil (6 micrometers), across the ranges of Coarse and X-Coarse material. The casts of machined surfaces were measured with a light section microscope. These represent relatively idealized circumstances. Agreement between methods may or may not be as good for the case of blasted surfaces.
Testex' own testing on grit-blasted steel surfaces characterized with a number of fully electronic digital stylus instruments tends to support the grooved-surface error ( 0.25 mil or 6 micrometers) except in the profile range extending from 1.5 mils to 2.5 mils (38 to 64 um). The error associated with measurements in this range has a more complicated behavior.
Pass/Fail determinations of profile around the values 1.5, 2.0 and 2.5 mils (38, 50 and 64 um) will tend to be quite accurate. Pass/Fail assessments centered on other values in this range will be less accurate. Testex recommends that specifiers and inspectors agree, before beginning a job, on profile limits that take this circumstance into consideration.
In general, Testex replica tape determinations of profile will be similar in magnitude to the parameter Rt measured with fully digital electronic stylus devices. Replica measurements generally display statistical error behavior comparable to or better than that obtained with fully digital electronic stylus devices.