Abstract
The jaggedness of signatures, such as the force-displacement curves of brittle cereal-based foods, can be a manifestation of physical processes. Consequently, it should be regarded as an inherent property of the system rather than noise that should be discarded. Changes in the degree of jaggedness can provide useful information on the system's performance or on how it is affected by external conditions, such as exposure to moisture. Quantitative assessment of the degree of jaggedness can be done by methods that include statistical analysis, conversion of the record into a Fourier power spectrum, and determination of its apparent fractal dimension. Most can be performed with microcomputers and commercially available software. It is recommended that the jaggedness of any experimental curve or signature be determined by at least two methods or algorithms simultaneously for mutual verification. The magnitude of any single measure of jaggedness is determined by the signature's fluctuations amplitude and frequency. While the amplitude effect can be accounted for by normalization, the frequency cannot. Consequently, only curves having about the same number of points at the pertinent interval can be meaningfully compared. There is, however, a mathematical method, at this time still empirical, that allows meaningful comparison of the jaggedness of signatures obtained at different sampling rates. When several or numerous jagged signatures are recorded simultaneously, as in the case of brittle particulates compressed in bulk, the resulting combined signature is considerably smoothed. This is primarily the result of averaging caused by mutual cancellation of peaks in different directions and reducing the relative weight of any large peak. However, there are mathematical methods, some partly empirical, that enable restoration of the jaggedness of the original signatures.
Collapse