|dc.description.abstract||In Chapter 1, a literature review was undertaken to identify the etiological factors of failure of soldered bars in bar attachment systems for removable implant overdentures. Clinical studies with prosthodontic maintenance complications with implant overdentures, where bar attachment systems had been used as the suprastructure, were identified to establish the perceived etiology of failure. A further search of the fixed prosthodontic literature was then done to identify specific factors affecting the strength, fatigue resistance and quality of gold solder joints related to those used for bar attachment systems. This showed low failure rates of inter-implant bars, but higher failure rates for distal-extension bars. There was no explanation or description of the nature of the failure in the clinical studies reviewed. Conversely, the fixed prosthodontic literature identified multiple factors that could potentially have affected the solder joints. Two potential sites of failure initiation in distal bar cantilevers are identified and a simple estimate of the tensile stresses at the solder joint indicates values comparable to the fatigue failure stresses reported in the literature. Further research was deemed necessary to establish the fatigue resistance; physical and metallurgical properties of bar attachment system solder joints for implant overdentures.
In Chapter 2 a series of clinical case studies were followed up to investigate the failure process of 6 failed soldered bars, 4 intact soldered bars and 1 intact cast milled bar supporting either maxillary or mandibular implant overdentures. A total of 11 overdenture bars, divided into 2 groups, were removed for scanning electron microscopy observations from patients. The failed bar group (n= 6) had failed overdenture bars which were recovered from patients following up to 2 years of wear before requiring prosthodontic maintenance and repair. The intact bar group (n=5) had soldered bars as well as a single cast milled bar, which had been worn by patients for 2-5 years prior to receiving other aspects of prosthodontic maintenance. All bars were examined using scanning electron microscopy to establish the possible mode of failure or to identify evidence of potential failure in the future. Evidence of a progressive failure mode of corrosion fatigue and creep were observed on all the failed bars and in various stages of corrosion. Fatigue and creep was also observed in all the intact bars. Where the level of corrosion was substantial, there was no evidence of wear from the matrices of the attachment system. Evidence of an instantaneous failure mode, ductile and brittle overload was observed on the fracture surfaces of all the failed bars. In all failures, the fractures occurred within the solder and within the non-oxidising gold cylinder abutment at the solder / cylinder interface. The intact bars showed evidence of progressive failure modes around the solder areas and non-oxidising gold cylinder abutments. It was concluded that corrosion followed by corrosion fatigue appears to be a key factor in the onset of the failure process for overdenture bars supporting overdentures. The clinical implications relate the recommendation of the use of connecting bars to support overdentures, as well as using distal cantilevers.
Chapter 3 involved an in vitro investigation of the nature of the observed clinical failure process of bar attachment systems and cantilevered bar attachment systems by comparing this with a simple flexure and fatigue cycling simulation of similar cantilever soldered joints. 6 non-oxidising gold cylinders were soldered to Dolder bars using 3 different solders and a flame soldering technique, one with a round (filleted) shaped design and the other with a straight (non-filleted) shape design. Force was applied with a universal testing machine and deflection of the bars was plotted in relation to the load applied. Additional fatigue tests were conducted on the cantilever beams using a simple cam activated force loading system (150N) that lowered a dead weight onto the cantilever beam at a frequency of 1.2 Hz. The time to fracture and number of cycles to fracture was recorded. SEM observations were made at 8000 and 16000 cycles to identify the onset and propagation of the fatigue crack extension as well as the failed joints. It showed that the soldered joints used with bar attachment systems had relatively low yield stresses and that distal cantilever bars are prone to plastic deformation under biting forces similar to those developed in vivo. Also, it was shown that the yield stress depends on the solder alloy type. Stress corrosion in conjunction with masticatory cyclic loading appeared to be the primary factor in the etiology of observed clinical failure.
In Chapter 4 the influence of flame and furnace soldering methods was investigated on the stress corrosion, fatigue resistance and fracture toughness of a Dolder bar soldered to a non-oxidising gold cylinder using a gold solder. Two bars were soldered to opposite sides of a cylinder using the two techniques. Fatigue cycling one side of the bars was undertaken, with one group in a wet corrosive environment and the other in air. The system compliance and energy loss (mJ), was calculated. The critical stress intensity factor, K1c, of the solder / cylinder interface was established using a single-edge-notched-beam test. Mode of failure was evaluated using SEM. The results showed that no corrosion was observed in the wet environment. During fatigue cycling, intermittent flexural evaluation showed a loss of energy within the soldered joints. Both soldering methods produced a thin brittle diffusion layer within the surface of the cylinder. The K1c of this layer was similar to that of a brittle ceramic. No embrittling reaction occurred at the solder / bar junction. It was concluded that both soldering techniques produce a thin brittle diffusion layer that created a brittle intermetallic phase within the surface of the cylinder that created an inherent weakness within the joint and that this interfacial reaction has clinical relevance for soldered bar attachment systems using non-oxidising cylinders.
Chapter 5 investigated the metallurgical properties of flame and furnace soldered joints produced with three different solders in terms of their elemental composition. Three Dolder bars were soldered, each with a different solder, to the non-oxidising gold cylinder using a flame and furnace soldering technique. These, plus non-heat treated reference specimens were embedded and polished for metallographic analysis using a field emission SEM and electro dispersive spectrography analysis (EDS). Images were acquired using either a secondary detector or backscattered electron detector (BSD) to identify different areas and metallic phases within the samples for analysis. The width of the reacted phases at the soldered interface was measured. The study found that distinct phases at the non-oxidising gold cylinder / solder junction formed during the soldering process in the solder matrix and the cylinder, but not at the solder / Dolder bar junction. The width of diffusion zone in the edge of the cylinder is dependent upon the elemental composition of the solders used and the soldering technique, with the flame technique producing a narrower zone width compared to the furnace technique. The homogeneous diffusion zone within the cylinder contained the elements Au, Pd, Pt, Ag , Cu and Zn, that was different to the bulk material which contained only Au, Pd and Pt. Zn, Cu and Ag had migrated from the solder towards the cylinder interface while Pd and Pt had diffused from the cylinder into the solder to form eutectic and precipitated spheroid phases. The eutectic structure formed within the solder adjacent to the cylinder was narrower with the flame technique compared to the furnace technique.
Chapter 6 investigated the same joints prepared for EDS analysis in Chapter 5 in terms of their mechanical properties before and after the soldering process. Indents were carried out in a UMIS nanoindentation system. Indents were placed across the interface of the non-oxidising gold cylinder / solder / Dolder bar for each of the solder types and soldering techniques. In addition, a specimen of non-heat treated non-oxidising gold cylinder, three types of solder and Dolder bar were also indented for purposes of comparison. Subsequently, 3 parallel rows of indents were repeated at a decreased force across these interfaces. The study found that as a result of the soldering process the non-oxidising gold cylinder showed a slight increase in E-modulus and a slight decrease in hardness for the heat-treated materials compared to the non-heat treated control, while the Dolder bar shows the opposite with a small decrease in E-modulus in the heat-treated materials and a small increase in hardness for the heat-treated materials. The E-modulus and hardness values of the diffusion zone, identified in Chapter 5, within the edge of the non-oxidising gold cylinder at the non-oxidising gold cylinder / solder interface were significantly higher then those of the surrounding non-oxidising gold cylinder and the solder. The increase in E-modulus and hardness values of the diffusion zone was caused by the diffusion of elements into and out of the non-oxidising gold cylinder with a resultant formation of a different alloy to that of the original non-oxidising gold cylinder alloy.
Overall Conclusion of this thesis: The soldering process, when joining a non-oxidising gold cylinder to a Type IV gold Dolder bar, creates a hard brittle diffusion zone within the non-oxidising gold cylinder at the non-oxidising gold cylinder / solder interface that may lead to catastrophic failure if crack propagation, that occurs within the more ductile solder, reaches this zone.
Recommendations - When fabricating bar attachment systems using a soldering technique, the dental technician should avoid using the non-oxidising gold cylinders as supplied and recommended by the implant component manufacturers. Instead, the dental technician should cast plastic burnout abutments / cylinders in a Type IV gold alloy, thereby avoiding the development of the brittle diffusion layer. Manufacturers of implant cylinders for use in the fabrication of bar attachment systems, should introduce an additional cylinder to their product line for the soldering technique made from Type IV gold alloy, the same that is used to manufacture the Dolder bar.||