Testing of industrial products - Aggregates for construction

Definitions

The following definitions are taken from the Final Draft of prEN 932-6 (April 1998) Tests for general properties of aggregates. Part 6 Definitions of repeatability and reproducibility) that sets out definitions of measures of precision applicable to aggregates. (The subscript "1" is appended to these measures to indicate that sample reduction variation is one of the sources of variation that contributes to their size.)

Critical range Wc
Value less than or equal to which the range of n single determinations made to obtain a test result is expected to lie within a probability of 95%.
Repeatability r conditions
Conditions where test results are obtained with the same test method on identical test portions of aggregate, in the same laboratory, by the same operator, using the same equipment, and within short intervals of time.
Repeatability r1 conditions
Conditions where test results are obtained with the same test method on different test portions of the same laboratory sample of aggregate, in the same laboratory, by the same operator, using the same equipment, and within short intervals of time.
Repeatability standard deviation Sr or Sr1
Standard deviation of test results obtained under repeatability r or r1 conditions.
Repeatability value r or r1
Value less than or equal to which the absolute difference between two test results, obtained under repeatability r or r1 conditions, is expected to be within a probability of 95 %.
Reproducibility R conditions
Conditions where test results are obtained with the same test method on identical test portions of aggregate, in different laboratories, by different operators, using different equipment.
Reproducibility R1 conditions
Conditions where test results are obtained with the same test method on test portions of different laboratory samples of the same bulk sample, in different laboratories, by different operators, using different apparatus.
Reproducibility standard deviation SR or SR1
Standard deviation of test results obtained under reproducibility R or R1 conditions.
Reproducibility value R or R1
Value less than or equal to which the absolute difference between two test results, obtained under reproducibility R or R1 conditions, is expected to lie with a probability of 95%.

The following terms are also used.

Crushed or broken surfaces
Facets of a particle of gravel produced by crushing, or broken by natural forces, and bounded by sharp edges. If the surface edges of a particle are worn or weathered then these surfaces shall be considered as rounded for the purposes of the test method.
Crushed or broken particle
Particle with more than 50 % of its surface crushed or broken.
Totally crushed or broken particle
Particle with more than 90 % of its surface crushed or broken.
Rounded particle
Particle with 50 % or less of its surface crushed or broken.
Totally rounded particle
Particle with more than 90 % of its surface rounded.
DSC units
The percentage by mass of the test portion that passes the test sieve after the DSC test.
Fines
Aggregate passing a 0.063 mm sieve
FD units
The percentage by mass of the test portion that passes a 1.60 mm sieve after the FD test.
FI units
The percentage by mass of flaky particles in the test portion.
LA units
The percentage by mass of the test portion that passes a 1.60 mm sieve after the LA test.
MB units
The mass in grams of anhydrous methylene blue adsorbed per kilogram of the 0/2 mm fraction of the aggregate.
MDE units
The percentage by mass of the test portion that passes a 1.60 mm sieve after the MDE test.
PSV units
The degree of polish, measured by a standard friction tester, of stone that has been subjected to polishing in a standard polishing machine.
SE units
The ratio of the height of the column of sediment to the height of the flocculated material, expressed as a percentage.
SI units
The percentage by mass of non-cubical particles in the test portion.
SZ value
The test result obtained from the Schlagversuch. It is the average of the determinations on three test specimens prepared from one laboratory sample.
Sub-sample
Used when the participants were required to divide each laboratory sample of aggregate into two parts which were then reduced further to produce test-portions.
Test portion
Sample used as a whole in a single test. (Typically in the experiments reported on this web site, the test portion was 1kg to 5kg of a coarse aggregate, or 100g to 500g of a fine aggregate).
Test specimen
Sub-sample used in a single determination of a property when a test method requires more than one determination of a property.
Aliquot
A measured volume (typically 100mL) of an aqueous extract prepared from a test portion.
Determination
A result obtained by applying the test method once to an aliquot or a part of a test portion.
Test result
When more than one determination is obtained on a test portion, the test result is the average of the determinations.
Laboratory average
The average of the data obtained by one laboratory in one level of a cross-testing experiment.
Between-determination range
When two determinations obtained by a laboratory on one test portion, the between-determination range is calculated as:
the highest determination - the lowest determination
Between-test-specimen standard deviation
When three specimens are used to give a test result, the between-test-specimen standard deviation is calculated as the standard deviation of the three determinations on the specimens.
Between-sub-sample range
When two test results are obtained on each of two sub-samples, the between-sub-sample range is calculated as the absolute value of:
the average of the two test results obtained on sub-sample 1 - the average of the two test results obtained on sub-sample 2
Between-test-portion range
For two test results obtained by a laboratory on one laboratory sample, the between-test-portion range is calculated as:
the highest test result - the lowest test result
Laboratory bias
For the data obtained in one level of a cross-testing experiment, the bias of a laboratory is estimated by:
the average of the data from the one laboratory - the average of the data from all the laboratories
Standardised range (k-statistic)
For the data obtained in one level of a cross-testing experiment, a standardised range is a measure of the size of the range from one laboratory relative to the general size of ranges from all the laboratories.
Standardised laboratory bias (h-statistic)
For the data obtained in one level of a cross-testing experiment, a standardised laboratory bias is a measure of the size of the bias of one laboratory relative to the general size of the biasses of all the laboratories.
Sensitivity ratio
A measure of the ability of a test method to discriminate between two materials (under either repeatability or reproducibility conditions).
Straggler/outlier
A member of a set of values that is inconsistent with the other members of the set. (A straggler is a value that is shown, in a statistical test, to differ from the other members of the set at the 5% significance level. An outlier is a value that is shown to differ at the 1% significance level.)