Evaluation of biomechanical properties of jujube branches and analysis of prediction accuracy based on multi-scale artificially simplified model

The multi-scale model is key to the evaluation of shear or compression mechanical properties of jujube branches at the microlevel. In this study, multi-scale models of the microstructure of jujube branches, namely, the natural model and the artificially simplified model, were built using computed tomography and Python. The equivalent compression and shear properties of the microstructure were predicted by the finite element method. The prediction results of the artificially simplified model were compared with those of the natural model, and the prediction deviation of the artificially simplified model was analyzed. The results showed that the transverse shear modulus obtained by the artificially simplified model was slightly higher than the predicted value of the natural model, with an average deviation of 1.84%. The transverse shear strain energy obtained by the artificially simplified model was lower than the predicted result of the natural model, with a maximum deviation of 5.59%. The cell distribution direction had an important effect on the accuracy of prediction for the transverse elastic modulus, and the deviation was 16.53%, indicating that the further optimization of the cell distribution direction could enhance the prediction accuracy and the effectiveness of the model. The results can be used to evaluate the unit compression or shear mechanical properties of plant stalks with hierarchical structuring and to analyze the quantitative relationship between the microstructure and equivalent mechanical properties.