An 'electronic nose' based on mass spectrometer and discriminant function analysis (DFA) was used to evaluate the grade of coffee beans. The data obtained from the electronic nose was analyzed by DFA. The discriminant function first score (DF1) of natural coffee beans showed a greater decrease than the different processing methods. Defective coffee beans were separated well from non-defective coffee beans by DF1, which correlated with a weaker flavor than that of the others. Flavor patterns of the defective and the non-defective coffee beans were determined as complementary information. The flavor patterns obtained in this study can explain, in a simplified way, the differences between the defective and the non-defective coffee beans.

The linear polarization resistance method is one of the widely used techniques for the corrosion rate monitoring in the water circulating systems of plants. The measurement is simple and rapid, so that a continuous on-line monitoring is possible without any shutdown of plants. A 2-electrode polarization corrosion rate measurement system was installed in a laboratory using a data acquisition board and PC. The signal processing parameters were optimized for the accurate corrosion rate measurement, and the polarization resistance was compensated with the solution resistance measured by the high frequency sine wave signal of an output channel. The precision of corrosion rate data was greatly improved by removing the initial noise signals on measuring the polarization resistance.

No abstract

No abstract