2016 Volume 57 Issue 12 Pages 1993-1997
In spinal fixation devices, the activity of the patient can cause fretting of the metal-to-metal contacts between the rod and plug, which may result in failures. In this study, compressive fatigue tests were conducted with rods made of Ti–29Nb–13Ta–4.6Zr alloy (TNTZ) with oxygen contents of 0.06 mass% (06O) and 0.89 mass% (89O) and Ti–6Al–4V extra low interstitial alloy (Ti64) as comparison in both air and saline solution. The fatigue strength increases in the order of 06O < 89O < Ti64 in both air and saline solution. These results indicate that solid-solution strengthening by oxygen improves the fretting fatigue resistance of the TNTZ rod.
Titanium (Ti) alloys have been used as biomedical materials because of their good biocompatibility and high corrosion resistance. Ti alloys are often used for biomedical applications, including in implant devices with metal-to-metal contacting parts.
Metallic implants inserted into the body can eventually undergo some degradation via various mechanisms including fretting1). Spinal fixation devices consist of three components; a rod, screw, and plug. The rod is fixed to the screw by mechanical fastening using the plugs. In these devices, the activity of the patient can cause fretting of the metal-to-metal contacts between the rod and plug1). Failures have consequently occurred because of fretting in this system, which consists of many mechanical joints2).
To evaluate the fatigue life of Ti spinal implants, two types of Ti alloys were tested as rods, in this study. The first was a conventional (α + β)-type Ti–6Al–4V ELI (extra-low-interstitial) (Ti64), which accounts for approximately 45–60% of the total Ti alloys produced globally3), and the second was a newly developed β-type Ti alloy for biomedical applications, Ti–29Nb–13Ta–4.6Zr (TNTZ). It has been reported that vanadium in Ti64 is toxic to the human body4), and stress shielding can occur because the Young's modulus of Ti64 (approximately 110 GPa) is higher than that of cortical bone (approximately 10–30 GPa)5). However, TNTZ exhibits good biocompatibility because of its nontoxicity, and a lower Young's modulus (approximately 60 GPa) than that of Ti646).
However, based on the results of previous studies7–9) on wear properties of TNTZ and Ti64, it has been concluded that TNTZ exhibits significantly more severe subsurface deformation than Ti64, which is caused by its lower resistance to plastic shearing deformation. Furthermore, the fatigue strength of TNTZ is not sufficient to satisfy the requirements of a long service life for biomedical implants, even though this alloy exhibits good biocompatibility and a low Young's modulus. It is therefore necessary for TNTZ to further increase the resistance to fretting fatigue and concomitantly reduce the Young's modulus such that it approaches that of human bone.
The Young's modulus is related to the crystal structure, which is not significantly changed upon increasing the interstitial element content. From this viewpoint, interstitial elements are promising because their addition to TNTZ is expected to improve the plastic shear resistance, which could suppress the fretting fatigue failure via solid-solution strengthening.
For the effect of solid-solution strengthening, low-cost oxygen (O) is used as an interstitial element in this study. The use of interstitial element O affords several advantages, including no risk to the human body, larger solid-solution strengthening compared with conventional substitutional alloying elements, and minimal effect on the specific gravity of the material10). Therefore, the mechanical performance of the spinal construct consisting of the TNTZ rods with O contents of 0.06 and 0.89 mass% in air and saline solution was investigated in this study. Conventional Ti64 was also used as a rod material for comparison.
The screws and plugs were constructed and assembled as per the manufacturer's recommendations. Ti64 was used as a plug and screw in this study. The materials used as a spinal rod in the present study were hot forged bars of the Ti64, TNTZ containing 0.06 mass% O (06O), and TNTZ containing 0.89 mass% O (89O) with a diameter of 5 mm and a length of 100 mm. The chemical compositions of the bars of 06O and 89O are listed in Table 1. The bars of 06O and 89O were subjected to a solution treatment at 1063 K and 1323 K (β transus temperature + 50 K) for 3.6 ks in vacuum followed by water quenching. For the vertebrectomy testing, ultrahigh molecular weight polyethylene (UHMWPE) test blocks were also used as simulating bone.
| Element Alloy |
Ti | Nb | Ta | Zr | O | C | N |
|---|---|---|---|---|---|---|---|
| 06O | bal. | 29.4 | 13.0 | 4.64 | 0.06 | 0.01 | 0.008 |
| 89O | bal. | 30.5 | 12.7 | 5.01 | 0.89 | 0.01 | 0.007 |
Testing of the spinal implant assemblies was based on a simulated vertebrectomy model using a large gap between two UHMWPE test blocks based on American Society for Testing Materials (ASTM)-F1717. The ASTM-F1717 specification is the standard testing method used to evaluate the mechanical performance of spinal implants. The UHMWPE test blocks are designed to minimize the effects of the variability of bone properties and morphometry.
Figure 1 shows an experimental model prescribed by ASTM-F1717 that mimics a segment composed of the functional spinal units. The ASTM-F1717 tests were carried out in both air at room temperature and saline solution (0.9 mass% NaCl solution). The temperature of saline solution was kept at 310 K during the tests. Each fatigue test was performed at a frequency of 5 Hz with a stress ratio, R = 0.1. The rod breakage was an indication of failure in this fatigue testing. The ASTM standard for the fatigue test indicates the run-out number of cycles (5 × 106 cycles). Therefore, the fatigue test was retained until failure or 5 × 106 cycles. All the screw systems were tightened by 8 Nm using the plugs.
Specification of experimental setup according to ASTM-F1717 model.
In addition, the fretted surfaces of the plugs and rods were analyzed using a scanning electron microscopy (SEM) after the ASTM-F1717 tests. Electron probe micro-analyzer (EPMA) observations were also performed to enable analysis of the O intensity within the fretted surfaces.
The maximum cyclic compressive load–the number of cycles to failure (S–N) curves obtained from the compressive fatigue testing of the Ti-based spinal construct using the Ti64, 06O, and 89O rods in air and saline solution are shown in Fig. 2. The symbols with diagonal lines represent data from the rods fractured at the middle part, and the other symbols represent data from the rods fractured at a position under the plug. The Ti64 (tested at a maximum compressive cyclic load of 350 MPa in air) and 06O (tested at maximum compresssive cyclic loads of 200 and 250 MPa in air, and 250 MPa in saline solution) rods are fractured at the middle part in a low cycle fatigue life region (under 105 cycles).
Compressive fatigue test results for Ti-based spinal constructs containing Ti64, 06O, and 89O rods using an ASTM-F1717 model, in (a) air and (b) saline solution.
Moreover, the fatigue strength increases in the order of 06O < 89O < Ti64 in both air and saline solution. This result indicates that solid-solution strengthening by O improves the fatigue resistance of the TNTZ rod material. However, no sharp difference is observed between the air and saline solution conditions for any of the rod materials.
3.2 Cracking observation at fretted surfaceAfter the rod breakage in the ASTM-F1717 tests, the cross-section of the unbroken fretted surface of the 06O rod was observed using an SEM, as shown in Fig 3. The crack propagates from the fretted surface to the inside of the rod. Generally, the first fretting fatigue crack develops with an inclination of approximately 45° to the fretted surface during plain fretting fatigue tests11). The fatigue cracks in Fig. 3 also develops with an inclination of approximately 45° to the fretted surface in this study. Moreover, the crack vertically develops to the contacting surface at increasing depths. Therefore, it is considered that this crack from the fretted surface is the fretting fatigue crack, which occurs when contacting components experience small amplitude relative motion12); it is too difficult to examine the cross-sections of the Ti64 and 89O rods using an SEM because of the small widths (approximately 480 and 450 μm for the Ti64 and 89O rods, respectively) of the fretted surfaces.
SEM micrograph of cross-section of 06O rod after ASTM-F1717 test at a maximum cyclic compressive load of 175 N.
In addition, Fig. 4 exhibits SEM micrographs of fretted surfaces between the plugs and rods at broken parts in the Ti64, 06O and 89O rods after the ASTM-F1717 tests. All rod materials (Figs. 4 (a)–(c)) exhibit that fatigue failure occurs at the edge of the fretted surface. According to another study, a fretting fatigue crack originates from the edge of the fretted surface where the strain energy is the highest in a plain fretting fatigue test11). Thus, the results shown in Fig. 4 are also strong evidence for the occurrence of fatigue failure for the Ti64, 06O, and 89O rods by fretting.
SEM micrographs of fretted surfaces between plugs and rods at broken parts in (a) Ti64, (b) 06O, and (c) 89O rods after ASTM-F1717 tests at a maximum cyclic compressive load of 175 N.
The Ti64 and 06O rods are fractured at the middle part of the rod in the low cycle fatigue region in both air and saline solution as shown in Fig. 2. Besides for those data, all the other data from the rods broken at a position under the plug indicate that the fatigue failure is accelerated by the fretting in the high cycle region (over 105 cycles). This result indicates that the effect of fretting fatigue is negligible in the low cycle fatigue life region where the number of cycles to failure is under 105 cycles.
The EPMA assessment of the O intensity maps on fretted surfaces of the Ti64, 06O, and 89O rods against the Ti64 plug was performed after the ASTM-F1717 tests in air, as shown in Fig. 5. The widths of the fretted surfaces are approximately 480, 900, and 450 μm for the Ti64, 06O, and 89O rods, respectively.
EPMA maps of O on fretted surfaces of (a) Ti64, (b) 06O, and (c) 89O rods against Ti64 plug after ASTM-F1717 tests at a maximum cyclic compressive load of 175 N in air.
In particular, it is observed that all the Ti rods exhibit high O intensities at the edge parts of the fretted surfaces in both air and saline solution (not shown here), which can be explained by the slip between the plug and rod occurring at mainly the edge part of the fretted surface during the ASTM-F1717 testing. This result signifies that the fretting and fretting damage are the highest at the edge part.
Fretting is viewed as providing a local site that causes elevation of strain, which ultimately causes crack nucleation1). As previously discussed, fretting damage accelerates the nucleation and early growth of the fatigue crack, which may cause failure of the Ti64, 06O, and 89O rods in a spinal fixation system.
The plug slips on the rod at small amplitude during the ASTM-F1717 testing, which leads to the relative oscillatory tangential movement between the plug and rod, as observed in Fig. 6. The fatigue crack originates from the edge part of the fretted surface on the rod. In particular, the fretting between the plug and rod mainly occurs at both ends of the edge of fretted surfaces, according to the results in Figs. 4 and 5.
Schematic drawing of spinal construct containing plugs, rods, and screws and fretting between plug and rod by cyclic slip during ASTM-F1717 testing.
Generally, the fretted surface exhibits electrochemical oxidation in moist air and simulated body fluid13). Under partial slip conditions, periodic passive film removal followed by re-passivation of the metal surface leads to metal loss according to the anodic reaction:
| \[{\rm Ti} + 2{\rm H}_2{\rm O} \rightarrow {\rm TiO}_2 + 4{\rm H}^+ + 4{\rm e}^-\] |
This is the reason why O intensity is higher at the edge parts of the fretted surfaces, as shown in Fig. 5.
Figure 7 exhibits a schematic drawing of fretting damages at the fretted surface of the rod. The fretted surface can be divided into slip and stick regions. The fretting fatigue cracks generally initiates near the boundary between the slip and stick regions14–16), which provides evidence that the fatigue crack initiated from the edge part of the fretted surface caused fretting between the plug and rod. As shown in Fig. 7, the inward edge of the fretted surface exhibits more severe fretted damage than the opposite edge because of the compressive load occurring in the spinal construct. The fatigue crack also initiates at the inward edge during the ASTM-F1717 testing, in this study.
Schematic drawing of fretting damage at fretted surface of rod during ASTM-F1717 testing.
In addition, Fig. 8 exhibits the distribution of shear stresses at increasing depths below the fretted surface of the rod, based on a stress model of the fretting fatigue11). The position of the fretting fatigue crack in this study can be also predicted, as shown in Fig. 8. Moreover, the initial direction of the crack is approximately 45°. Between the plug and rod, normal pressure is applied via the torque pressure of the plug. In this case, the local normal pressure is maximum at the center of the contact area and minimum at both edges of the contact area17). Consequently, when the plug slips on the rod by cyclic compressive stress during the ASTM-F1717 testing, the highest alternating shear stresses occur at both ends of the edge, which exhibit the minimum local normal pressure.
Distribution of shear stresses at increasing depths below fretted surface of rod during ASTM-F1717 testing.
Therefore, the initial direction of the fretting fatigue crack is approximately 45° to the fretted surface and it vertically develops to the fretted surface at increasing depths, as shown in Figs. 3 and 8.
4.2 Effects of solid-solution strengthening on fretting fatigueFretting drastically decreases the fatigue strengths of spinal fixation devices. This reduction in fatigue life is attributed to the introduction of shear stress on the fretted surface through contact18).
Smaller width of the fretted surface in the 89O rod (approximately 450 μm) compared to that of the 06O rod (approximately 900 μm) indicates that the resistance to plastic shearing substantially increases because of the solid-solution strengthening effect caused by intersitial O. Therefore, it can be considered that the solid-solution strengthening by O improves the plastic shear resistance of the TNTZ rod, which leads to the higher fatigue strength of the 89O rod compared to that of the 06O rod in both air and saline solution. In conclusion, one of the most effective treatment against fretting fatigue failure in implants is to increase the plastic shear resistance.
4.3 Differences in fatigue properties between air and saline solutionIn plain fretting fatigue tests19), the fatigue behaviors of Ti alloys in solutions of simulated body fluid differ from those in air. The ASTM-F1717 tests were thus performed in not only air but also saline solution, in this study.
One of the effects of saline solution on fretting is the lubricant effect. It has been reported that the fretting fatigue strength of TNTZ specimens in Ringer's solution is slightly higher than that in air in the low cycle fatigue life region because Ringer's solution can act as a lubricant between the fretting areas16). In this case, the fatigue strength can increase because the maximum friction force can be reduced, leading to a decrease in the strain energy at the edge part of the fretted surface.
In contrast, in this study, no significant difference is observed in the fatigue strengths in both air and saline solution determined using ASTM-F1717 tests for any of the rod materials (Fig. 2). This finding can be explained by the reduction of the lubricant effect of the saline solution resulting from the strong contact pressure between the plug and rod.
Another effect is the corrosion effect of the saline solution. Fretting–corrosion results in metal degradation because of the simultaneous action of mechanical wear and chemical oxidation20). Furthermore, slip or partial slip conditions in fretting fatigue tests can lead to anodic metal oxidation13). In particular, saline solution containing chlorine can act as an electrolyte and accelerate oxidation by dissolved O. However, there is no significant difference in the fatigue strengths of any of the rod materials between the air and saline solution conditions, as shown in Fig. 2. Therefore, the corrosion effect of saline solution is considered negligible in these ASTM-F1717 tests.
In this study, fatigue strengths of three Ti-based spinal constructs using Ti64, 06O, and 89O rods were evaluated according to the ASTM 1717 standard and analyzed. The primary findings of this study are the following:
This work was supported in part by the Industrial Technology Research Grant Program in 2009 from the New Energy and Industrial Technology Development Organization (NEDO), Grant-in-Aid for Scientific Research (A), Young Scientists (A), Challenging Exploratory Research from the Japan Society for the Promotion of Science (JSPS), and the Inter-University Cooperative Research Program “Innovation Research for Biosis–Abiosis Intelligent Interface” from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan.
