Page 228 - Introduction to Mineral Exploration

10: EVALUATION TECHNIQUES 211

where C is the heterogeneity constant (Box collect a minimum of 1000 particles. This
10.1) and K equals 125,000. will seem a trivial amount when collecting, for
From this a series of equal variance lines can instance, a froth flotation product, but plants
be drawn with decreasing values of K, such as have accurate sampling systems built into the
K/2, K/4, etc., which correspond to increasing process system. It becomes important when
variances. For each stage, knowing M and the dealing with larger particles, such as RC or
2
increased value of So (FE) the related coarser blasthole rock chips, rocks in stockpiles, or
grain size (d) can be calculated from the Gy on trucks. Invariably the amount required will
formula. be larger than anticipated, but don’t forget if
the primary sample is unrepresentative, every-
Variation of the safety line thing else is a waste of time.
The heterogeneity constant (C) contains the
liberation factor β (Box 10.1). The safety lines Hand or mechanical shovel splitting
above are calculated on the basis that this Splitting samples is best completed by the pro-
factor is at its maximum value of 1.0 which cess of alternative shovelling which minimizes
is a safe assumption but only correct when the the splitting and preparation error contribution
top particle size is smaller than or equal to the to the total error. The sample is spread either
liberation size (d lib ) of the valuable compon- on a sheet of thick plastic or sheet iron which
ents. If d is larger than d lib then β becomes is on a flat, clean, and smooth floor which
0.5
[d lib /d] , which is less than 1.0. can easily be swept. The material is piled into
When the liberation size of the mineral of a cone (step 1) which is then flattened into
interest is not known the value of 1.0 is used as a circular cake (step 2) and the first step is
a safe procedure. This provides for the continu- repeated three times for maximum mixing of
ously straight equal variance lines as described the sample. Studies have shown that rolling on
above, with a slope of 3.0. When the liberation a cloth or plastic sheet does not necessarily
size is known, the above formula can be used cause mixing. If too slow the material just
with a β value of less than 1.0 but approaching slides, and if too fast, there is a loss of the finer
this value as the grain size d is reduced. In this particles thereby causing bias. From the final
case each sampling stage has a unique safety cone, shovelfuls of material are taken con-
line slope from about 2.5 increasing to 3.0 as d secutively around its circumference and placed
approaches d lib , with the less steep sections in alternatively into two separate piles. In this
the coarser grain sizes. François-Bongarçon and way the original cone, and sample, is split into
Gy (2002) point out that there are still common two separate but equal smaller cones and the
errors being made when applying Gy’s formula. choice as to which is the new sample is decided
They describe the main variable, the liberation by the toss of a coin and not by the sampler.
size, and illustrate the importance of modeling Further weight reduction is continued in
it with a worked example. exactly the same manner on the new sample.
Alternatively, the cone may be quartered with
a shovel and each quarter is further reduced in
Splitting of samples during reduction
the same way. This technique is referred to as
In each stage after reduction in grain size the cone and quartering.
sample mass is correspondingly reduced by The width of the hand, or small mechan-
splitting. The sampling described in this text is ical, shovel should be at least three times the
concerned with small batches of less than 40 t width of the diameter of the coarsest sample
and usually less than 1 t which are stationary particle size which provides for a maximum
in the sense that they are not moving streams possible size of about less than 100 mm with
of material on, say, a conveyor belt. In explora- the former and less than 500 mm with the
tion, stationary samples include chip samples latter tool. Shovels should always be over-
produced by percussion or reverse circulation flowing and working with them partially filled
drilling (section 10.4). is not a good practice. Obviously, towards the
In the field it may be difficult to calculate end of a splitting operation with finer grain
the weight of sample required. In such cases, sizes a smaller sized tool has to be used.