Fabrication of porous polymeric-based scaffold for dental tissue repair in fracture healing: RVE simulation and ANN optimization

•Using biodegradable polymers with a freeze-drying technique leads to producing a porous scaffold.•After preparing the porous scaffold, SEM and (FTIR analysis were used.•The mechanical and biological behaviors of the samples were evaluated.•Multi-scale finite element modeling (FEM) was promoted to predict the mechanical properties of the fabricated scaffold.•The numerical results indicated a good agreement with experimental results. In this study, a porous scaffold composed of polyvinylpyrrolidone (PVP) and alginate containing titanium nanoparticles (Ti-NPs or TiO2 nanoparticles) layer by layer was produced using biodegradable polymers and a freeze-drying technique. The prepared porous scaffold was evaluated for its mechanical and biological behavior, such as swelling, porosity, and biodegradability in phosphate buffer saline (PBS) and simulated body fluid (SBF), through scanning electron microscopy (SEM) analysis. The incorporation of hydrophilic polymers increased the amount of water absorption, reaching 811 ± 214% in the layer-by-layer scaffold, which was significantly different from pure PVP. Cell culture of rat bone marrow stem cells (rMSC) on the scaffold showed a cell survival percentage above 80% after 3, 5, and 7 days, and the morphology of the cells indicated optimal biocompatibility. Furthermore, multi-scale finite element modeling (FEM) was used to predict the mechanical properties of the fabricated scaffold with different porosity percentages. Four types of representative volume elements (RVEs) were generated with randomly distributed 74%, 79%, 81%, and 84% porosity. The tensile strength and stress–strain curves of the specimens were obtained using an ABAQUS/Explicit solver attached with VUSDFLD subroutines to evaluate the elastic modulus of simulation results. The numerical results demonstrated good agreement with experimental results, indicating the reliability of the developed FEM for predicting the tensile properties of the fabricated porous scaffold. Finally, an artificial neural network (ANN) was developed to determine the effect of weight percentages (wt. %) of loaded TiO2 nanoparticles into bone scaffolds on the porosity and elastic modulus. The predicted results from the ANN showed an increase in the porosity and elastic modulus with an increase in wt. % of TiO2. The generated ANN can be used to determine the optimal amount of nanoparticles for various applications of dental scaffolds.