# Mechanical behavior of tooth-class II restoration complex with various restorative materials using linear and non-linear finite element analysis

## TL;DR

This study uses computer modeling to compare how different dental restoration materials affect tooth stress, finding that amalgam and gold cause more stress than composite resin and ceramic.

## Contribution

The study introduces non-linear finite element analysis with non-bonded contact conditions to better simulate real-world dental restoration mechanics.

## Key findings

- Amalgam restorations caused the highest enamel and dentin stresses.
- Ceramic restorations resulted in the lowest stresses in tooth structures.
- Non-bonded contact conditions in amalgam and gold restorations contributed to higher stress levels.

## Abstract

This study aimed to investigate the mechanical behavior of the tooth–class II restoration complex restored with composite resin, ceramic, amalgam, and gold using three-dimensional linear and non-linear finite element analyses (FEA). Previous FEA studies generally assumed linear mechanical properties and bonded interfaces between tooth and restorative materials, which differ from clinical reality, particularly for non-bonded restorations such as amalgam and gold. Therefore, in this study, non-linear mechanical properties and non-bonded contact conditions were adopted for different restorative materials. The results showed that the highest enamel and dentin stresses occurred in amalgam restorations, while the lowest were found in ceramic restorations. Amalgam and gold produced higher stress in enamel and dentin compared with composite resin and ceramic. The higher stress observed in amalgam and gold was mainly associated with the non-bonded (frictional) contact condition, because the stresses in these restorations remained below their reported yield strengths under the applied loading conditions. This study provides mechanical insights into the behavior of the tooth–class II restoration complex and offers a possible mechanical interpretation of previously reported clinical observations. These findings should be interpreted within the limitations of the finite element model and do not directly represent clinical outcomes.

The online version contains supplementary material available at 10.1038/s41598-026-40204-3.