Uncertainty assessment of the characteristic value of the gray standard sample for the coating resistance test of architectural coatings

0 Foreword
In order to ensure the accuracy and comparability of the results of the test analysis of stain resistance of building coatings, in order to improve the scientificity and accuracy of stain resistance data of building coating coatings, GB/T 9780-2013 "Building coatings coating stain resistance test method" standard has a better promotion significance, developed the building coating coating stain resistance test ash standard samples. Uncertainty is a parameter that is indicative of the dispersion of the measured value and the measurement results. In order to reasonably assess the uncertainty of a measurement result, it is necessary to confirm its traceability characteristics and
the
. In assessing the measurement uncertainty of standard sample characteristic values, in addition to the uncertainty part introduced by the measurement procedure, the uncertainty part introduced by sample unevenness, long-cycle stability and short-cycle stability is also evaluated. In this paper, the reflectivity value/grey cake of the gray standard sample used in the coating resistance test of architectural coating is used as an example, and the uncertainty assessment of the characteristic value of the standard sample is briefly introduced.1 Standard material preparation and uniformity inspection
the electric duster or cloth bag dust collector the first electric field separated by all the thick ash goal grinder for grinding, grinding fine coal ash through 0.9 mm square hole sieve, according to the requirements of the test qualified, mixed into a closed bottle of polypropylene material. Each bottle has a net content of 200 g and the bottle is labeled with the appropriate information. The minimum number of packaging units is approximately 1,000 bottles, and the extract is calculated in units (N for the total number of units). A sample of 20 bottles of a batch number standard sample is randomly selected in a prepared sample for uniformity testing. The uniformity test was performed using the F method, with 3 independent tests per bottle, resulting in a total of 60 data. During the development of standard samples of coating resistance to contamination, uniformity inspection data show that there is no significant difference between the average measurement between bottles and their standard deviations, so that the standard substances can be determined to be uniform. Table 1 lists the homogeneity test results of the standard sample reflectivity value/grey cake.
Reflectivity value/grey cake uniformity test results
2 stability test
stability test at intervals using the principle of first secretion, then dredging, on the day of sample preparation, and then placed 1 month, 3 months, 6 months, 1 2 months, 24 months, 36 months, each random sampling 1 group, each group repeated measurement 3 times, calculate its mean and standard deviation. Table 2 shows the stability test results of the standard sample reflectivity value/grey cake.
Reflectivity value / ash cake stability test results
Because the standard value of the characteristics of the standard sample is unknown, the average consistency test method is used to evaluate the stability of the standard sample, using the formula expressed as:
reflectivity value / ash cake stability test results
3 standard sample evaluation

reflectivity value / gray cake using GB/T 23981-2009 "white and light paint ratio determination." The value-setting principle is that multiple laboratories collaborate on the value-setting in an accurate and reliable way. Each laboratory uses a uniform measurement method. Laboratories participating in the scale have passed metering certification or have analyzed units with high technical levels within the industry (see Table 3 for a summary of data per unit).
the reflectivity value/grey cake test results of the standard samples in table 3
using the Dixon test method, the suspicious data are excluded. The normality of the data was tested by the bias coefficient and the peak coefficient. Under the condition of obeying normal distribution or approximate normal distribution, the average value of the data measured by each measurement method of each laboratory is regarded as a single measurement value, which constitutes a new set of measurement data and calculates its total average. The total average of the reflectivity value/greycake measurement data for the standard sample is 35.3.
4 uncertainty is calculated
the fixed value results are represented by standard values and uncertainties.
the numerical repair shall be carried out in accordance with the provisions of GB/T 8170-2008 "The statement and determination of numerical repair rules and limits". Uncertainty is calculated using GB/T 15000.3-2008/ISO Guide 35:2006 General Principles and Statistical Methods for Standard Sample Work Guide (3) Standard Sample Value.
(1) Uncertainty calculation of fixed statistical averages: uncertainty calculation of
constant statistical averages
evaluation of the uncertainty of unevenness between bottles of reflectivity values/grey cakes, uncertainty of unevenness between bottles see Table 4.The uncertainty assessment caused by the long-term instability of the
(3) characteristic value between
bottles is
it can be said that the standard sample stability test is qualified, according to the stability statistical test results, it can be guaranteed to be stable and qualified within 3 years, and the uncertainty caused by its long-term instability is negligible.
the composition uncertainty of the CRM attribute values can be found in Table 5.
standard values and uncertainties of standard samples for stain resistance to building coatings
5 Conclusion
all factors that contribute significantly to the uncertainty of standard sample characteristic values should be included. In order to ensure completeness, the synthetic uncertainty of a standard sample should include factors such as indetermination of technical capabilities in participating in the fixed-value laboratory, material unevenness, long (short) cycle stability, and factors introduced by batch normalization (in certain cases, the uncertainty introduced by the system error inherent in the measurement procedure should also be included).