Page 182 - Vogel's TEXTBOOK OF QUANTITATIVE CHEMICAL ANALYSIS

of the sample rnay be completely wasted. If the material is more or less
homogeneous, sampling is comparatively simple. If, however, the material is
bulky and heterogeneous, sampling must be carried out with great care, and
the method will Vary somewhat with the nature of the bulk solid.
The underlying principle of the sampling of material in bulk, Say of a truckload
of coal or iron ore, is to select a large number of portions in a systematic manner
from different parts of the bulk and then to combine them. This large sample
of the total weight is crushed mechanically, if necessary, and then shovelled into
a conical pile. Every shovelful must fa11 upon the apex of the cone and the
operator must walk around the cone as he shovels; this ensures a comparatively
even distribution. The top of the cone is then flattened out and divided into
quarters. Opposite quarters of the pile are then removed, mixed to form a smaller
conical pile, and again quartered. This process is repeated, further crushing
being carried out if necessary, until a sample of suitable weight (say, 200-300 g)
is obtained.
If the quantity of material is of the order of 2-3 kg or less, intermixing rnay
be accomplished by the method known as 'tabling'. The finely divided material
is spread on the centre of a large sheet of oilcloth or similar material. Each
corner is pulled in succession over its diagonal partner, the lifting being reduced
to a minimum; the particles are thus caused to roll over and over on themselves,
and the lower portions are constantly brought to the top of the mass and
thorough intermixing ensues. The sample rnay then be rolled to the centre of
the cloth, spread out, and quartered as before. The process is repeated until a
sufficiently small sample is obtained. The final sample for the laboratory, which
is usually between 25 and 200 g in weight, is placed in an air-tight bottle. This
method produces what is known as the 'average sample' and any analysis on
it should always be compared with those of a second sample of the same material
obtained by the identical routine.
Mechanical methods also exist for dividing up particulate material into
suitably sized samples. Samples obtained by these means are usuaiiy representative
of the bulk material within limits of less than + 1 per cent, and are based upon
the requirements established by the British Standards Institution. Sample
dividers exist with capacities of up to 10 L and operate either by means of
a series of rapidly rotating sample jars under the outlet of a loading funnel, or
by a rotary cascade from which the samples are fed into a series of separate
compartments. Sample dividers can lead to a great deal of time-saving in
laboratories dealing with bulk quantities of powders or minerals.
The sampling of metals and alloys rnay be effected by drilling holes through
a representative ingot at selected points; al1 the material from the holes is
collected, mixed, and a sample of suitable size used for analysis. Turnings or
scrapings from the outside are not suitable as these frequently possess superficial
impurities from the castings or moulds.
In some instances in which grinding presents problems it is possible to obtain
a suitable homogeneous sample by dissolving a portion of the material in an
appropriate solvent.
Before analysis the representative solid sample is usually dried at 105-1 10 OC,
or at some higher specified temperature if necessary, to constant weight. The
results of the analysis are then reported on the 'dry' basis, viz. on a material
dried at a specified temperature. The loss in weight on drying rnay be determined,
and the results rnay be reported, if desired on the original 'moist' basis; these