Reduction of the Sample to Analytical Size
The two common methods of reducing bulk samples to a practical size are riffling and quartering. A diagram of a riffle is shown in figure 10. The sample is fed onto the top of the riffle and as it falls through the device the sample is divided equally into two bins. When the operation is complete, the contents of one bin is discarded and the other passed through the riffle again. In this way the sample is progressively and randomly halved until its bulk is reduced to that required for laboratory work up.
The riffle, due to its design is extremely difficult to clean and therefore it is practical to arrange for an individual riffle should be permanently kept for use with a specific material.
Figure 10. The Sampling Riffle
The quartering procedure is diagrammatically represented in figure 11.
Figure 11. Sample Quartering
The total sample is placed on a clean impervious surface and formed into a cone by shoveling. Each shovel full is placed on the top of the last so that the material runs down the side of the pile and is thereby distributed around the sides of the pile as evenly as possible (if larger pieces of material roll away from the base of the cone they must be pushed back to the edge). From the first cone two successive cones are sequentially made in the same manner (this is to mix the sample well). The third cone is then flattened to a uniform thickness and then quartered along two diameters. One diagonal pair of quarters is rejected. These four stages are repeated until a sample of the required size is obtained. If the initial bulk sample is very course then the average particle diameter made need to be reduced by grinding between each quartering procedure.
Sample Size
Some samples, e.g. forensic samples, such as material from an incinerated container are obviously limited in size, but it is generally important to obtain an adequate quantity of sample. If bulk material is available then the sample size is no problem. If the amount of material available is limited either by cost or availability, however, then a quantity sufficient for at least three replicate analyses should be taken. If possible, enough extra sample should be taken for reference purposes and to allow at least three additional samples to be analyzed should the initial results become challenged.
