Measurement of the perceived odour intensity of building materials with multi gas-sensor systems

In recent years multi gas-sensor systems (electronic noses) have been developed in order to mimic the human sense of smell. These systems comprise of an array of gas sensors with a broad and partly overlapping selectivity, and a data processing unit. At the Hermann-Rietschel-Insitute the multi gas-sensor system "Kamina of the FZ Karlsruhe is used to determine perceived intensity of odours from building materials. The data processing uses pattern recognition and regression methods to correlate the signals to the perceived intensity of odours measured by a trained human panel. The building materials were placed in ventilated emission chambers (climpaqs). The contaminated exhaust air was assessed by a trained human panel and the emission load was measured by the sensor system. The investigations were done at different concentration levels for each material which was realized by adding pure air to the sample air after the climpaq. The measurements and the parallel assessments of the panel were processed using a principal component regression (PCR) method. The results show good correlations between the perceived intensity and the sensor signals for the investigated materials.