A sub-sample of aggregate is dried at 110 ± 5 ° C, and sieved on a 2.00 mm sieve to produce a test portion containing at least 200 g of the 0/2 mm fraction. This is placed in a beaker with 500 mL of demineralised water and, if necessary, 30 g of kaolinite. The mixture is kept agitated and methylene blue dye solution (10 g of anhydrous dye/L) is added progressively until the stain test shows that the test portion has finished adsorbing the dye.
The stain test is carried out by dipping a glass (or plastic) rod into the mixture and then allowing a drop of the mixture to fall onto a filter paper. If the test portion has adsorbed all the dye, the drop on the filter paper appears as a spot of blue-stained grains surrounded by a colourless halo. When more dye has been added than the test portion can adsorb, the central spot is surrounded by a blue halo.
The addition of kaolinite is not mandatory, but it helps the halo to be distinguished from the central spot, particularly when the sand contains a low proportion of fines. If kaolinite is added, it has to be tested to establish the volume of dye solution that it adsorbs.
The Methylene Blue value is calculated as:
MB = 10 × (V1 - V') ÷ Md
with units of g (of anhydrous dye) per kg (of 0/2 mm fraction of the aggregate), where
V1 = the volume of dye solution added (mL)
V' = the volume of dye solution adsorbed by 30 g of the kaolinite
Md = the mass of the test portion (g)
The CEN test method has been developed from the French Standard P 18-592 (AFNOR, 1990. P 18-592. Essai au bleu de méthyline méthode à la tache). In the French method, the test is carried out on water-extracted fines (the test portion is washed over a 0.080 mm sieve using a specified volume of water, and using a 0.500 mm sieve as a guard sieve). The CEN method differs from this in that the test portion is prepared by dry-sieving the aggregate on a 2.00 mm sieve. Another difference between the methods is that in the French test the aggregate that is to be used in the dye test is not dried - its moisture content is determined on a separate test portion.
Participants in the cross-testing experiment were sent copies of the method that was prepared in August 1996 for formal voting by CEN members. Some laboratories may have used the version prepared in their own language, and may have been given additional information by their National Co-ordinator.