Fourier modelling, analysis and interpretation of high-resolution mixograph data

High resolution mixograph data (HRMD) has the potential to provide improved information on dough rheology. A designed experiment involving nine cultivars grown under two levels of nitrogen fertilization and three levels of irrigation motivated the method of analysis that is based on time-varying Fourier terms. The procedure involves a preliminary time series analysis in the frequency domain based on the periodogram, to determine the main cyclic patterns in the data, followed by modelling of the cyclic patterns found. The estimates of the peak value in dough development and the associated time based on the resulting major cyclic frequency are shown to correlate highly with the existing method of analysis using Mixsmart software. The measure of bandwidth is based on the maximum of the time-varying amplitude of the sinusoidal term corresponding to the major frequency over time. The trace of this amplitude provides a clear single maximum, improves the ability to separate varieties in terms of bandwidth, and has the potential for characterizing dough extension across the full mixing process.