2nd Collaborative Study of the CORESTA on the Comparability of Three Selected Methods Measuring the Pressure Drop of Filter Rods/Auswertung der 2. CORESTA-Ringuntersuchung über die Vergleichbarkeit von drei ausgewählten Messmethoden für den Zugwiderstand von Filterstäben

Nineteen companies and institutes participated in the second collaborative study, using the measuring methods V_A, P_A and C in 3 series of filter-rods of 50 pieces each; 4 participants received 3 series and 3 CFO respectively. The differences of the mean values were much smaller than in the first collaborative study and did not exceed 28 mm water. Also the standard deviation of the single values and the standard deviation of reproducibility have decreased to less than half in comparison to the first collaborative study. There was no difference in the method-mean value over all participants and series. Significant differences arose however between the participant-mean values and the mean values as well of the standard deviations of the participants as also of their standard deviations of reproducibility. A regression between the standard deviation of the single values and the standard deviations of reproducibility could however not be shown. Comparing the values of the standard deviation between the samples of the same participants to the values of the residual variance of the single analysis of variance of the series over all participants, it shows that the latter is mostly based on the systematic participant influence. Also a regression between the standard deviation of the single values and the mean values of the pressure drop could not be noticed. An influence of temperature, relative humidity and atmospheric pressure on the value of the pressure drop could only be stated in series III of measuring method C corr. But it was so small that it could practically be neglected. By means of a rank correlation a connection between the mean values of the first and the second collaborative study could be shown for different parameters with most participants. This fact is another proof that the stated differences in pressure drop are mostly based on factors being characteristic for the individual participants. It is to be aimed at the fact that all these factors are checked more thoroughly in order to improve the agreement of the mean values of the pressure drop between the participants