This paper investigates the in fl uence of surface preparation treatments of dental implants on their potential (mechanical) fatigue failure, with emphasis on grit-blasting. The investigation includes limited fatigue testing of implants, showing the relationship between fatigue life and surface damage condition. Those observations are corroborated by a detailed failure analysis of retrieved fracture dental implants. In both cases, the negative effect of embedded alumina particles related to the grit-blasting process is identi fi ed. The study also comprises a numerical simulation part of the grit blasting process that reveals, for a given implant material and particle size, the existence of a velocity threshold, below which the rough surface is obtained without damage, and beyond which the creation of signi fi cant surface damage will severely reduce the fatigue life, thus increasing fracture probability. The main outcome of this work is that the overall performance of dental implants comprises, in addition to the biological considerations, mechanical reliability aspects. Fatigue fracture is a central issue, and this study shows that uncontrolled surface roughening grit-blasting treatments can induce signi fi cant surface damage which accel- erate fatigue fracture under certain conditions, even if those treatments are bene fi cial to the osseointegration process.
Dental implants offer a highly successful solution for missing teeth, contingent upon the well-known osseointegration process. Albrektsson et al. (1981) stated that the implant's surface properties affect the successful course of osseointe- gration. Those properties can be addressed from three differ- ent directions: Mechanical, topographic, and physicochemical ( Albrektsson and Wennerberg, 2004 ). The effect of surface topograp hy on the biological reaction and on bone-implant contact has been studied extensively in the dental implant research fi eld, for over a decade. Average height deviation parameters ( R a and S a ) between 1 and 2 m m, which de fi ne a “ moderately rough surface ” ,werefoundtobe optimal for a successful osseointegration process ( Albrektsson and Wennerberg, 2004 ; Elias and Meirelles, 2010 ; Wennerberg and Albrektsson, 2009 , 2010 ). A great variety of surface treatments exist today, in order to achieve a desired degree of surface roughness. Those include machining, plasma spray and laser peening (LST), acid etching, grit blasting followed by acid etching, anodizing and biomimetic coating. Among those, grit blasting is one of the most common dental implant surface treatments ( Elias and Meirelles, 2010 ; Wennerberg and Albrektsson, 2010 ). Blasted surface roughness with S a values ranging from 0.6 and 2.1 m m is deemed ideal for the implant's osseointegration ( Wennerberg and Albrektsson, 2009 ). During this process, implants made of pure titanium (CP – Ti) and titanium alloy (Ti – 6Al – 4V) – which are the most widely used biomaterials for fabrication of dental implants ( Elias et al., 2008 ) – are blasted with air – propelled hard ceramic particles (Al 2 O 3 , TiO 2 or Ca 2 P 2 O 7 )( Guéhennec et al., 2007 ). Depending on the size and shape of the ceramic particle, which is polyhedral with sharp corners ( Barriuso et al., 2014 ), and on its velocity, erosion and material tearing of the titanium surface, is in fl icted. The result is different surface roughness levels that can be produced on the implant's surface.
he numerical results provide a detailed characterization of the residual pressure and strain fi eld underneath the impact. They indicate that for the speci fi c simulated systems (particle size and target material), there is a threshold impact velocity ( V th ) beyond which signi fi cant damage of the impacted plate (tearing, holes, pits, cracks) will develop. The adiabatic nature of the impact causes a high temperature rise, especially in the damaged plate areas that sustain large plastic strains. These high temperatures might cause melting, or at least signi fi cant softening of the plate material, thereby promoting embedding and adhesion of the ceramic particles to the target. Finally, the evolution of the damage in the impacting sphere is fully characterized and found to be more signi fi cant in the harder Ti – 6Al – 4V substrate.
he present study addresses systematically the effect of grit blasting on the fatigue strength of commercial dental implants, using a combination of mechanical fatigue testing, failure analysis and numerical modeling. The results of our study strongly suggest that the grit blasting treatment can shift from bene fi cial to detrimental if it is carried out in an uncontrolled manner. In the present case, the post- treatment analysis reveals tearin g of the titanium/titanium alloy, which will become increasingly severe as the particle velocity increases for a given substrate. The resulting surface condition is largely affected by the particle size, particle and target mechan- ical properties, nozzle diameter and blasting pressure. While those parameters can be controlled to some extent, it is still important to remember that the particle velocity has an inherent statistical distribution. Those particles that impact at the higher end of the velocity spectrum are likely to cause the observed surface defects that facilitate fatigue crack initiation.dental laser tips
The numerical analysis reveals that for the speci fi c simu- lated treatment parameters ,a velocity threshold can be identi fi ed. This threshold can be used in order to evaluate the mechan- ical effect of the surface treatment, and de fi ne the optimal treatment parameters (particle velocity and size) that will roughen the surface of the implant (osseointegration) without destroying it (cracking). It is important to emphasize that the numerical model which was presented is based on certain assumptions, such as adia- batic conditions and spherical shape alumina particles. In reality, the particles might be polyhedral, as shown in earlier work ( Barriuso et al., 2014 ). We deliberately simpli fi ed the situation by adopting a spherical shape. This assumption does not detract fromtheabsolutenatureoftheimparteddamage,withthe proviso that sharp particles might even in fl ict more damage than spherical ones, with an increased tendency to early particle disintegration upon impact. Another point is that the real blasting process involves impact by tens of thousands of particles of varying shape and unknown velocities. Such a situation cannot be modeled realistically, so that the current simulation addresses a single im pact and its associated effects. In that sense, the obtained results are a sort of “ lower bound ” for the actual evolution of the implant's surface. However, it is believed that the numerical model retains the salient features of the problem at hand, namely that as the impact velocity increases, the nature and the extent of the damage vary, with the existence of a velocity threshold below which no signi fi cant damage will result from the surfa ce treatment. It is therefore advised to carefully re-assess the various parameters of a planned grit-blasting treatment, simulate it to optimize it, and validate the process by performing a thorough surface charac- terization. At this stage, this assessment must be largely experimental.
Considering the nature of the imparted damage, SEM obser- vations clearly showed the presence of embedded ceramic particles on the implants' surface. During the fabrication process of the implants these particles should be removed either by chemical treatments done according to the ISO recommenda- tions ( ISO B600, 2007 ; ISO F86, 2007 ), or during the etching surface treatments done using a strong acid. The very fact that these particles remained embedded to the surface, whether intact or shattered indicates that these particles have adhered to the surface as a result of the aggressive surface treatment. In that case too, the numerical simulations show that for a certain velocity, the surface temperature can approach or even exceed the melting temperature of the surface, thereby apparently promoting a strong particle-substrate adhesion process. The embedded particles are strong stress raisers, creating a singular stress distribution in th eir vicinity. This singularity is created as a result of the geometry, created by the adhered particle and the different elastic properties of the metal and the ceramic ( Hein and Erdogan, 1971 ). One can expect that the resulting “ in fi nite ” tensile stresses will generate immediate micro-cracks that can then subsequently grow by a fatigue mechanism. The embedded particles and their effect on fatigue crack nucleation were unambiguously identi fi ed for both labora- tory tested and retrieved dental implants. One can now trace one of the fatigue crack initiation causes to the presence of embedded ceramic particles, a point that was not previously identi fi ed. Finally, another misconception dealt with in this research, is that grit blasting surface treatment, with residual surface embedded particles, has no adverse effect, on both processes of osseointegration and implant surface integrity. This study showsthatthetreatmentitselfhasaprofoundeffecton the mechanical reliability of the dental implants. The main conclusion of this work is therefore that implant manu- facturers and practitioners alike should be aware of the potential fatigue failure, and should take into consideration mechanical considerations related to surface preparation, in addition to the biological ones, to achieved increased mechanical reliability of dental implants.