Particle Sampling

How can you ensure your bowl of muesli contains the right portion of each and every ingredient, each and every time? Due to free-flowing ingredients, segregation occurs, and so the size of cereal pieces in the first bowl varies greatly from the last bowl from the box. Typically, larger sized ingredients are at the top and the fines at the bottom of the packet.

This problem is analogous to sampling of particulate materials for particle size analysis – how do you know that the sample you have is truly representative of the material you wish to test?

Sticking with muesli, to obtain a representative portion a reliable method of sampling is required.  You could cone and quarter the cereal. This involves flattening a cone shaped heap of cereal before dividing it into four and dividing each quarter further, until only 1/16^th of the original volume is left. Nonetheless, large pieces tend to fall to the outer edges of the heap while finer ones adhere to the work surface, and when combined with other factors give an estimated maximum error (EME) value >20%.

Can this be improved? You could try scoop sampling, whereby scoops of cereal are taken straight from the box. However, the samples would not be representative of variations across the box. Whilst better than cone and quartering, the EME is still poor at >17%

Fortunately we can apply the ‘golden rules’ of sampling, which state that particles should be sampled when in motion, from the whole stream and in many short increments. One technique that upholds these rules and is called spin riffling. Muesli is poured into a hopper, flows down a vibrating chute and free-falls into containers rotating at a constant speed. With an EME value < ½% it is the most reliable sampling technique.

So for a superior breakfast to the last particle in the box, spin riffle your cereal,