This paper presents different ways of modelling the strength of corrugated fibreboard (CFB) subjected to different levels of pre‐crushing. The strength performance was measured through four‐point bending loading and edge crush test (ECT). The models used in this study were an analytical solution, an equivalent flute model, and detailed flute geometry models that consisted of idealized sine geometry and real geometry. The study found that the bending performance was dependent on the calliper of CFB rather than the flute geometry. All models showed a similar trend in predicting the drop in bending stiffness as the level of pre‐crushing increased, albeit with different absolute value. It was found that the real geometry model of the board predicted ECT performance better than the other models. However, at severe pre‐crushing levels (>50%), there was a significant drop in the experimental ECT force not predicted by the models. For these cases, there was evidence of delamination of the flute, a failure mechanism that was not included in any of the models. The analytical solution model provides the quickest prediction but could not predict the crushed ECT performance due to not considering the calliper variable in the equation. The equivalent model showed faster solving time compared with both real and idealized geometry models, although these microgeometry models predicted ECT the most accurately.

This work was aimed at examining the overall contributions to displacement of panels of compressed boxes, such as panel compression strain and flap and crease displacements. 3D digital image correlation (DIC) was used to analyse motion of side panels of corrugated fiberboard regular slotted containers. The vertical displacement and the vertical component of strain of the panel face during box compression test were examined. Measuring displacements of the whole compressed box with DIC enables measurement of the in‐plane compression of a panel in isolation from the horizontal fold or crease zone displacement, without having to test tube sections of the box to infer or extract the in‐plane panel compression behaviour. Detailed study of two box designs in three representative test cases was presented, which in future could be extended to other box designs. At peak load, the in‐plane compression of the panels, calculated from the average vertical component of the strain along the right edge of the long panel of the box, was 3% to 6% of total crosshead displacement for the test cases. At peak load, the portion of the box compression associated with bottom box flaps or crease zone crushing was 48% to 59% of total crosshead displacement for the different cases. The analysis showed that the majority of the vertical displacement of the box occurred in the top and bottom creased folds and that these folds are responsible for the low apparent in‐plane stiffness of a box.

This research presents a technique to quantify morphological damage to flutes in corrugated fibreboard (CFB). The method involves laser cutting thin samples and analysing digital images of the flute profiles. The surface profiles of creased CFB before and after laser cutting were measured using fringe projection and showed that the sample preparation does not significantly affect the flute profile. After imaging the laser cut samples, skeleton analysis was used to derive a digitised profile of the flute shape. To characterise the level of damage to the flute profile, a similarity factor (SF) was introduced to quantify the relative difference between test sample and reference flute profiles. Validation of this analysis technique was done by generating known images of flute profile with variations that include distortions that could occur to CFB. These images were then fed into the skeleton analysis, and the results were compared with the original profile. This comparison showed good agreement between the initial and skeleton‐analysed flutes. A demonstration of the skeleton analysis on purposefully damaged actual CFB flute profiles shows that the SF reduces as the level of crushing increases, showing that the technique could be used to enumerate morphological damage to CFB during manufacture, conversion, and use.

This paper presents experimental work, finite element (FE) model, and analytical solution for predicting the four‐point bending on C‐flute corrugated fibreboard (CFB) when oriented at different angles. The angles of the CFB samples used in this research study were 0° (cross‐machine direction) and 30°, 45°, 60°, and 90° (machine direction). The CFB was assumed as an orthotropic shell element in the FE model and was validated by comparing the bending stiffness, maximum bending force, and failure formation from the experimental test. It was found in the experiment that the 90° sample had the highest bending stiffness with the lowest maximum bending force while the 0° sample had the opposite. An interesting finding was that the 30° and 45° samples improve the bending stiffness than does 0° without significantly affecting the maximum bending force. Both the FE model and analytical solution predicted the bending stiffness trend of the board from 0° to 90° with good agreement compared with experimental results. The maximum bending force in the FE model showed reasonable agreement with the experimental findings. The failure regions on the samples showed similar patterns in both experiments and the FE model. The accurate response in the FE model justify that it is a good tool to predict the bending behaviour of CFB.

The hybrid genetic algorithm (HGA) was used to optimise box design to maximise cooling performance, mechanical performance, pallet footprint, and container packing efficiency and to minimise cardboard usage. These factors are normally investigated independently, but the industry requires combined functionality. Here, we present a case study, for optimisation of regular slotted carton boxes filled with wrapped beef mince. After packing, individual boxes are chilled to a desired storage temperature and then palletised for shipping. Four models were developed to predict design performance, including cooling rate, mechanical performance, cardboard usage, and box stacking on pallets. The combination of the model results was used to score the average performance of box designs. The models were solved by Comsol Multiphysics, Ansys APDL, and Cape Pack. The overall design generation and optimisation were developed with Matlab that controls all these software packages, evaluates the interactions between results, and runs the HGA for box optimisation. The HGA was conducted for 10 generations each with a population of 100 individuals. The optimisation routine successfully found optimum dimensions for the box for the defined conditions with relative short simulation times (about 3 hours per generation). This paper demonstrates how overall optimisation of packaging can be achieved through combining the strengths of multiple simulation software packages.