The Rapid Visco-Analyser (RVA) normally does not indicate the actual t
emperature inside the sample can during operation, but rather that of
the heated block. Sample cans were modified by incorporating a thermoc
ouple that was immersed in the contents, and its temperature could the
n be monitored during operation. Three different starch types and four
temperature profiles of differing ramping rates were used. It was fou
nd that the actual paste temperature lagged behind the block temperatu
re, and the lag was a function of the heating/cooling rate and the app
arent peak viscosity of the starch sample. A second-order multiple-reg
ression equation (r = 0.94) was developed to predict that lag. The res
ults can be used to predict the actual liquid temperature for any give
n point on the pasting curve.