Upon analyzing the diffraction patterns it was found that the bone and dentin samples showed very similar trends for the constant creep experiments. Immediately upon loading, strong compressive strains were measured in both WAXS and SAXS; however, with increasing time under constant load, the strain on the HAP phase decreased and the strain on the fibrils increased. In the case of dentin, the strain rate on either phase was linear with values of 2.0·10-5 İ/min for the HAP and -7·10-6 İ/min for the fibrils. Bone on the other hand had a much greater changes in strain in the early stages of loading followed by plateaus. The maximum strain rate for the bones samples were 1.8·10-5 ε/min for HAP and -2.3·10-5 ε/min for the fibrils. dŚĂƚ ƚhe fibrillar strain, as a representation of the nanoscale composite strain, increased with creep time was expected. However, the strain trends for HAP are not. Reinforced composites, much like bone and teeth, usually exhibit the opposite behavior. It is expected that with increasing creep time, more and more load should be transferred from the weaker phase to the stiffer reinforcement phase, therefore causing an increase in elastic strain on the reinforcement. The exception to this rule occurs when there is delamination between the matrix and the reinforcement. As delamination of the interface occurs, the matrix can no longer transfer load to the reinforcement; as a result the matrix must carry more load and the reinforcement is allowed to elastically unload. Such a delamination scenario seems highly probable in both bone and dentin as the HAP-collagen interface is a weak one dominated by Van der Waals and electro-static forces [20, 21]. However, damage such as interfacial delamination will result in a change of composite stiffness as well as apparent stiffness due to a change in the ability to transfer load. Such a change would be seen in the creep-load-unload experiments. The bone and dentin samples showed the same creep trends in the creep-load-unload experiments, as in the constant creep experiments. However, the apparent elastic moduli (Eapp=ıapplied/İphase) of both the HAP (Eapp HAP) and the fibrils (Eapp fib) did not show significant changes after one, two or three hours of creep. The average Eapp HAP and Eapp fib for bone were 41±3.0 GPa and 13±0.94 GPa, respectively. For dentin, average Eapp of 31±3.8 GPa and 9.6±0.5were measured for HAP and collagen. These values are higher than those usually measured in bone or dentin (Unpublished Data,[18]). This consistency in the apparent modulus values suggests that the extent of debonding at the time of modulus measurement was always the same. However, the increase in the apparent elastic modulus of HAP in bone and dentin may be due to delamination damage incurred in the first hour of creep. Although this supports the hypothesis that there is delamination damage, it does not explain why the apparent elastic moduli remain constant. There may be other mechanisms at play that might explain the unusual strain partitioning during creep. The HAP platelets in dentin and bone are known to fall into two separate populations: intra-fibrillar and extrafibrillar HAP. Although the exact distribution of HAP between these populations is debatable [22-24], it is possible that their different locations within the hierarchy of these mineralized biomaterials could cause vast differences in their strains. Therefore it is possible that the measured decrease in the overall HAP strain might actually be caused by a large decrease in strain on the extra-fibrillar HAP which overwhelms a small increase in strain on the intra-fibrillar HAP population. However, if this were true, it would cause a broadening of the peak due to the constant increase in HAP strain distribution or microstrain. Although there was an initial increase in broadening upon loading, such a continuous increase is not consistently seen from the broadening analysis thus ruling out a bi-modal strain distribution as a possible explanation. The unexpected strain partitioning between HAP and collagen may also be due to a HAP platelet tilting during loading. If more HAP platelets were to become longitudinally aligned upon loading, this would cause a better load distribution among the aligned platelets which would be recorded as a decrease in HAP strain. However, the broadening analysis results show that upon loading the HAP platelets tilt away from the longitudinal direction. This results in a decrease in the population of longitudinally aligned platelets and should cause an increase in the measured HAP strain. Tilting is therefore not a valid explanation for the decrease in HAP strain with increasing creep time. Having eliminated other explanations for the unusual load portioning behavior between HAP and collagen in dentin and bone during creep, it seems that delamination damage is still the most reliable hypothesis. The only obstacle is the consistency of the apparent elastic moduli of HAP and collagen throughout creep. The obstacle however, may be overcome by considering the types of interactions at the HAP collagen interfaces. Although the interfacial behavior of mineralized biomaterials such and bone and teeth is not well understood, it is thought that the main bonding mechanisms are weak bonds such as Van der Waals and electrostatic forces [20, 21]. These interactions suggest not only that it would be easy for interfacial damage to occur, but also that the interfacial damage would be easily repaired upon unloading. This interfacial ‘healing’ process may be responsible for the constant apparent elastic moduli. Upon unloading the damaged interface would reform the broken bonds allowing 323
RkJQdWJsaXNoZXIy MTMzNzEzMQ==