Evaluation of the Scott bond method and testing of out-of-plane properties of paperboard

Abstract: The through-thickness mechanical behavior of paperboard is fundamental for converting operations and many end-use situations of paper-based products. High quality and uniformity of the through-thickness properties is ensured during paperboard manufacturing by use of on-line and laboratory quality control methods. One important test method used for quantifying the delamination strength of paperboard is the Scott bond type test [1]. In the Scott bond test, a swinging pendulum impacts an aluminum angle which is adhered to a paper sample. The loss of potential energy of the pendulum due to the rupturing of the paper piece is interpreted as the delamination strength of the paper material. Deviations and scatter of Scott bond values steer the production to maintain the desired strength of the material. The objective of this thesis is to minimize the scatter of data obtained from Scott bond type testing by building a better understanding of the method and the interaction between paperboard properties and Scott bond values. Scott bond testing of three-ply paperboard and splitted paperboard sheets were performed to enable testing of the whole construction as well as testing of the interfaces and the weakest ply separately. The Scott bond results were normalized with respect to the true delaminated surface of ruptured specimens to reduce the scatter of the results and to obtain a more accurate measurement of the delamination resistance of the material. The Scott bond sequence was also photographed using high-speed photography to explore the activated mechanisms during delamination of specimens. In addition, the 180 degree peel test and z-directional tensile test were performed to study the correlation between Scott bond values and the results from other quality control methods developed to measure the delamination strength of paperboard. The results show that normalizing the Scott bond data, by accounting for the true delaminated surface, and rejecting specimens that are impacted twice by the pendulum during the Scott bond sequence reduces the scatter of the data significantly. It also improves the correlation with the true delamination strength of the material measured using splitted paperboard sheets. Furthermore, the same procedure improves the correlation between Scott bond data and data obtained from peel testing and z-direction tensile testing. High-speed photography of the Scott bond sequence showed that variations of the delamination mechanism causes scatter of Scott bond data, which overestimates the measured delamination resistance of paperboard. The main conclusions from the thesis are that the scatter of Scott bond data can be minimized by normalizing Scott bond values with respect to true delaminated surface of ruptured specimens and by rejecting data from specimens that are impacted twice by the pendulum. The scatter of the data can also be reduced by being consistent with the chosen direction of the specimen by facing either the top ply or bottom ply upwards on the sample stage.