Cluster and principal component analysis in the assessment of fountain solution quality

Abstract

This study includes the analysis of 23 physico-chemical parameters of spent fountain solutions from 8 printing facilities (PFs) of Novi Sad, Serbia. Spent fountain solution samples were collected once per one hour during 8-hr working shift from sheet-fed and web-fed offset printing presses and analyzed using the standard procedures. Multivariate statistical methods, such as cluster analysis (CA) and principal component analysis (PCA), were applied for the interpretation of the physico-chemical dataset of the fountain solution samples. Based on the dataset, CA showed three groups of similarity between the fountain solution samples and the physico-chemical parameters. Cluster 1 (PF 1, 3, 5 and 8) and cluster 2 (PF 6 and 7) correspond to less polluted printing facilities in comparison to cluster 3 (PF 2 and 4). Three principal components (PCs) were identified as responsible for the data structure explaining 86.82% of the total variance of the dataset. The results obtained from this study could be very useful for the offset printing companies since they provide a complete quality overview of spent fountain solutions through the analysis of physico-chemical parameters and indicate the main sources of quality variation during the printing.