Deflection fatigue of Ti-6Al-7Nb, Co-Cr, and gold alloy cast clasps

Ahmad Mahmoud, BDS,a Noriyuki Wakabayashi, DDS, PhD,b Hidekazu Takahashi, DDS, PhD,c and Takashi Ohyama, DDS, PhDd Division of Oral Health Sciences, Graduate School, Tokyo Medical and Dental University, Tokyo, Japan Statement of problem. There is little information about the deflection fatigue of clasps in relation to stress

Purpose. The aim of this study was to investigate the fatigue resistance and permanent deformation of cast
clasps made of titanium and other dental alloys and to relate the fatigue resistance with the calculated stress values. Material and methods. Twenty-five Ti-6Al-7Nb, 25 Co-Cr, and 15 Type IV gold alloy clasps were subjected to cyclic deflection of preset values of 0.25 mm, 0.50 mm, or 0.75 mm, for 106 cycles (n = 14). Finite element models were created to calculate principal stresses within the specimens. Fatigue life, retentive force, and permanent deformation were recorded, and the fracture locations were determined microscopically. The results were characterized in relation to the stress within the clasps. One-way analysis of variance and Tamhane’s posthoc tests were used to compare the results of the 9 material-deflection groups (a=.05). Results. Ti-6Al-7Nb clasps exhibited significantly less permanent deformation than the other clasps under relatively greater deflections, indicating better adaptation to the tooth surface. However, the fatigue life of the Ti-6Al-7Nb clasps under 0.75-mm deflection, with the stress above the alloy’s 0.2% yield strength, was significantly shorter than those under smaller deflections. The gold clasps showed significantly longer fatigue life than the other clasps under the 0.50-mm deflection. High-stress areas within the fatigue clasp specimens coincided with the fracture locations. The probabilities of fatigue fracture and permanent deformation were closely related to the material strengths and the preset deflections. Conclusion. To minimize fatigue failures, the cast clasp should be designed with consideration of the stresses distributions within the clasps. (J Prosthet Dent 2005;93:183-8.)

The Ti-6Al-7Nb and gold clasps demonstrated fatigue resistance that allows placement in undercut greater than 0.25 mm, which is suitable for Co-Cr alloy clasps. These clasps may be indicated when esthetics or periodontal health is a primary concern.


i-6Al-7Nb alloy has been developed as an alternative to Co-Cr and other existing removable partial denture alloys because it offers excellent biocompatibility,

This study was supported by Grants-in-Aid for Scientific Research, 12771170 (N.W.) and 14571840 (N.W.), from The Ministry of Education, Science, and Culture of Japan. Presented at the 110th Scientific Meeting of the Japan Prosthodontic Society, Nagano, Japan, October 2003. Presented at the 82nd General Session and Exhibition of the International Association for Dental Research; an Arthur R. Frechette Research Award Finalist, Hawaii, March, 2004. a Resident, PhD student, Removable Prosthodontics, Department of Masticatory Function Rehabilitation. b Research associate, Removable Prosthodontics, Department of Masticatory Function Rehabilitation. c Associate Professor, Advanced Biomaterials, Department of Restorative Sciences, Division of Oral Health Sciences. d Professor and Chair, Removable Prosthodontics, Department of Masticatory Function Rehabilitation.

good resistance to abrasion, and other advantages for relatively low cost.1-4 Like other titanium alloys, the Ti-6Al-7Nb alloy has an elastic modulus that is approximately half of Co-Cr alloy, which increases its flexibility.5 The increased flexibility allows the retentive clasp arms to be placed into larger undercuts on abutments than Co-Cr alloy clasps, and this may be important when esthetics or periodontal health is a primary concern. Clasps undergo permanent deformation and fatigue fracture under repeated flexures caused by denture insertion and removal and mastication.6-9 The fatigue life of cast clasps made of commercially-pure titanium was reported to be shorter than that of Co-Cr and gold alloy clasps.5 However, the fatigue fracture and permanent deformation of cast clasps made of titanium alloy has not been sufficiently assessed in relation to stress distribution, and little is known about how these




Fig. 1. A, Schematic illustration of fatigue test specimen. Specimen consists of clasp arm, loading sphere, and plate for fixation to testing machine. B, Finite element model of clasp arm. Dimensions of each model were based on test specimen.

(Degulor M; Degussa), both with a phosphate-bonded investment (Biosint-supra; Degussa). The compositions and material properties of these alloys are shown in Table I. The Ti-6Al-7Nb was cast using argon arc melting technology with a centrifugal vacuum pressure casting machine (Valcan-T; Shofu, Kyoto, Japan). The Co-Cr alloy was cast using high-frequency induction melting technology with a centrifugal casting machine (SA-2000; Sankin, Tokyo, Japan). The gold alloy was cast using electric resistance melting technology with a centrifugal casting machine (TS3; Degussa). The casting procedures were determined following the manufacturers’ instructions for the metals and investments. Recovered castings were lightly cleaned with airborne-particle abrasion using 80-mm aluminum oxide particles. Each gold alloy clasp was heat-treated according to the manufacturer’s instructions.

Fatigue test
Twenty five Ti-6Al-7Nb, 25 Co-Cr, and 15 gold alloy cast specimens were divided into different experimental groups according to preset clasp tip deflections, 0.25 mm, 0.50 mm, and 0.75 mm, which were designed to compare the clasps’ functions under different undercuts to the tooth surfaces. The 3 groups of the Ti-6Al-7Nb clasps based on deflections were denoted by codes T25 (n = 5), T50 (n = 10), and T75 (n = 10). The 3 groups of Co-Cr clasps denoted by C25 (n = 5), C50 (n = 10), and C75 (n = 10), and G25 (n = 5), G50 (n = 5), and G75 (n = 5) were used for the gold alloy clasps. Each specimen was fixed to the testing machine with screws and subjected to a sinusoidal cyclic deflection generated by the radial direction force at the tip of each clasp arm at a frequency of 5 Hz. The load/deflection curve was monitored, and the test was terminated when the maximum force was reduced to less than 15% of the initial load, or when 106 cycles15 were completed. The permanent deformation was calculated as the difference in piston tip-loading sphere distance between the position at the beginning of each cycle and the position at the first cycle. The actual deflection was thereafter obtained by subtracting the permanent deformation from the preset deflection. One-way analysis of variance (ANOVA) and Tamhane’s T216,17 post-hoc tests were conducted to determine the differences in the permanent deformations, applied forces, and number of loading cycles between the 9 material-deflection groups were significant or not. After the fatigue test, an optical microscopic (VH-5000; Keyence Corp, Osaka, Japan) examination was performed to locate the fracture site in each clasp specimen.

clasps would function in long-term clinical use. Permanent deformation and fatigue fracture are caused by the stress created in the clasp.10,11 The stress distribution may depend on the elastic modulus of the alloy, dimensions and curvature of the clasp,12,13 and the amount and direction of defection in relation to the abutment undercut.14 The purpose of this in vitro study was to investigate the fatigue of the Ti-6Al-7Nb clasps compared to Co-Cr and gold alloy clasps. The fatigue life, retentive force, and permanent deformation of clasps under cyclic deflections of different magnitudes were evaluated, and deflection fatigue was characterized in relation to the stress distributions within the clasps.

A metal cylinder with a 10-mm outer diameter was used for the adaptation of a pre-formed tapered wax pattern (Rapid-Flex; Degussa, Dusseldorf, Germany). The clasp arm pattern originated from a plate and curved around the cylindrical surface 120 degrees in a single plane (Fig. 1, A). Average width and thickness of the clasp arm at 30 degrees were 0.92 mm and 0.97 mm, respectively, while those at 120 degrees were 0.73 mm and 0.79 mm, respectively. The plate served as an attachment for fixation to a fatigue-testing machine (250N; Shimadzu Corp, Kyoto, Japan). The original pattern was duplicated using a split mold to make acrylic resin patterns (Pattern resin, GC, Tokyo, Japan). A spherical bead 0.5 mm in diameter was glued onto the inner surface of each clasp tip to provide a point of force application. The patterns were invested and cast using Ti6Al-7Nb (T-alloy Tough; GC) with a magnesia-based investment (Selevest CB; Kobelco, Hyogo, Japan), Co-Cr (Biosil L; Degussa) and a Type IV gold alloy

Finite Element Analysis
By the use of the preprocessor of the finite element method (FEM) computer software (ANSYS 7.1 FEM; ANSYS Inc, Canonsburg, Pa), a 3-dimensional finite



Table I. Composition and mechanical properties of alloys evaluated
Ti-6Al-7Nb Co-Cr Type IV gold

Composition (wt %) Modulus of elasticity (GPa) 0.2% Proof stress (MPa) Tensile strength (MPa) Percentage elongation (%)

86.5 Ti, 7 Nb, 6 Al, 0.5 others 123 890 950 5

62.5 Co, 30.5 Cr, 5 Mo, 1 Si, 0.4 Mn, 0.3 C, 0.3 N 220 710 900 6

70 Au, 4.4 Pt, 13.5 Ag, 8.8 Cu, 2 Pd, 0.1 Ir, 1.2 others 90 620 740 17

All data were based on information provided by manufacturer for each respective material except for modulus of elasticity of Type IV gold alloy.18 Note: recorded yield strength value is affected by offset value and can be substantially different from proportional limit.24

element model for each fatigue test clasp was created based on the width and thickness of each specimen, which were measured at 7 representative locations using a measuring microscope (MM-60; Nikon, Tokyo, Japan) (Fig.1, B). Each model was meshed by 8-node hexahedral elements. In the simulation, each clasp tip was displaced by the average of actual deflections in the radial direction while the base area at the clasp’s shoulder was fixed. Elastic moduli of 90 GPa,18 220 GPa, and 123 GPa (Table I) were input into the program to simulate the gold, Co-Cr, and Ti-6Al-7Nb alloys, respectively. A Poisson ratio of 0.33 was used for all of the alloys.19 The friction coefficient between the rod and the clasp surface was assumed to be 0.2.14 The principal stress distribution within each clasp model was calculated using the post processor of the FEM computer software.

RESULTS Fatigue resistance
Figure 2 displays the fatigue life of each specimen. For each test group, a representative specimen was chosen, and its permanent deformation and force required for deflection are respectively shown in Figures 3 and 4 as functions of loading cycles. In the early stage, instant increases in permanent deformations were recorded for all specimens. Thereafter, all of the clasps with 0.25mm deflections, except for 1 Co-Cr clasp, sustained constant forces and permanent deformations and survived 106 cycles without fracture. One Co-Cr clasp with a 0.25 mm deflection and all of the other clasps with larger deflections also sustained almost constant forces and permanent deformations after the early stages. However, those clasps that fractured before the 106 cycles demonstrated sudden force decreases accompanied by dramatic deformation increases shortly before fracture. The average numbers of loading cycles for failure, the average loads for deflection, and the average permanent deformations are listed in Table II. One-way ANOVA showed that the differences in load required for deflection, the permanent deformation, and the number of cycles to failure among the 9 material-deflection test groups were significant (P,.05). The results

Fig. 2. Fatigue life distributions of specimens as function of number of cycle. Each vertical line segment represents clasp specimen. Bar with asterisk was considered outlier and not included in statistical analysis. Left superscript letters of each test group represents homogeneous subsets resulting from post-hoc comparison. Each subset represents groups with insignificant differences. (P>.05).

of the post-hoc comparison between the 9 materialdeflection experimental groups are shown in Figure 2 and Table II. It should be noted that one C50 clasp with the exceptionally greater number of cycles to failure was considered an outlier and was excluded from the statistical analysis (Fig. 2). Postoperative microscopic observations indicated that the fracture sites of all of the failed clasps were located between 48 and 91 degrees from the clasp arm shoulder, with a mean value of 71 6 11 degrees (Fig. 5). There were no significant differences in the fracture sites among the 9 test groups.

Finite element analysis
In the finite element models, the sites where the maximum principal stresses occurred were located between 39 and 85 degrees from the clasp shoulder, with an average of 70 6 11 degrees. The average principal stress distribution for each test group is illustrated in Figure 6. The stress values detected between 35 and 90 were relatively constant and above 85% of the maximum stress. There were no significant differences in the sites of the maximum stress among the clasp materials or



Table II. Means (standard deviations) of fatigue test and FEM results
Material Preset deflection (mm) Ti-6Al-7Nb Gold alloy Co-Cr

Test group code / Number of specimens*

0.25 0.50 0.75 0.25 0.50 0.75 0.25 0.50 0.75 0.25 0.50 0.75 0.25 0.50 0.75

T 25 / 5 T 50 / 9 T 75 / 9 No failure 22,469(9,876) 2,875(1,000) 9.6 (4.0) a 27.4 (3.9) c 64.7 (11.4) d 6.0 (0.8) 13.9 (1.9) 21.9 (1.8) 362 (17) 706 (47) 1070 (26)***

G 25 / 5 G 50 / 5 G 75 / 4 No failure 120,500(18,014) 32,375 (8,845) 13.4 (2.4) 40.0 (8.9) 132.0 (9.6) 6.3 (0.8) 12.3 (1.4) 18.5 (1.9) 296 (14) 559 (27) 773 (23)***
a, b b, c, d e

C 25 / 5 C 50 / 8 C 75 / 9 4 did not fail 28,929(10,358) 15,806 (5,312) 14.0 (3.5) a, b 114.4 (12.7) e 298.1(21.2) f 12.0 (0.9) 21.0 (1.6) 25.9 (2.9) 640 (23) 1095 (52)*** 1254 (56)***

Cycles to failure

Permanent deformation (mm)**

Load to deflection (N)

Max. principal stress (MPa)***

*Exclusion of some specimens due to errors in testing is reason for deviation of number of specimens from originally assigned 5 or 10 specimens. **Subscript letters (a to f) indicate homogeneous subsets resulting from post-hoc test. Insignificant difference in mean permanent deformation was evidenced among groups of identical subset (P>.05). ***Recorded maximum stress value exceeding respective material’s 0.2% yield strength.

Fig. 3. Change of permanent deformation for representative specimen from each group. Solid lines: Ti-6Al-7Nb clasps; dotted lines: gold alloy clasps; dashed lines: Co-Cr clasps.

Fig. 4. Load required for deflection as function of deflection cycles for representative specimen from each group. Solid lines: Ti-6Al-7Nb clasps; dotted lines: gold alloy clasps; dashed lines: Co-Cr clasps.

among the preset deflections. The average maximum principal stresses calculated are shown in Table II.

Based on the loads required to deflection, both Ti6Al-7Nb and Type IV gold alloy would be clinically ad186

vantageous due to low rigidity when compared to the Co-Cr clasps, which is expected to have a minimum possibility of traumatic overloading to the abutment tooth during insertion and removal.12 Fatigue fracture is commonly believed to occur at cyclic stress levels much lower than that needed to cause failure on a single application of load.19 However, it was previously claimed that the



Fig. 5. Fractured Co-Cr specimen tested under 0.50 mm preset deflection with enlarged view.

Co-Cr alloy could withstand a stress slightly above its proportional limit without failure over infinitely many cycles.20 Results of the present study were not in conflict with this claim. Four out of 5 Co-Cr clasps tested under 0.25-mm preset deflections, which had maximum tensile stresses (average of 640 MPa) slightly below the alloy’s reported 0.2% yield strength (Fig. 6), survived over 106 cycles. However, all of the other Co-Cr clasps under larger preset deflections with stresses above the yield strength failed long before reaching 106 cycles. All of the Ti-6Al-7Nb and gold alloy clasps with 0.5mm deflections and maximum stresses below the materials’ yield strengths failed before reaching 106 cycles. Furthermore, the fatigue lives of Ti-6Al-7Nb clasps under 0.75 mm deflection, with the maximum calculated stress above its yield strength, remarkably decreased. These results indicate that the fatigue limits of these alloys are well below their yield strengths. At relatively high deflection amounts, the material differences greatly affected the magnitude of the permanent deformations of the clasps. The Ti-6Al-7Nb clasps had considerably smaller permanent deformations when compared to the Co-Cr clasps under all preset deflections greater than 0.25 mm. They also showed much smaller permanent deformations than the gold alloy clasps under the 0.75 mm deflection. This may be because of the material’s high ratio of yield strength to elastic modulus,21 which indicates a relatively larger working range (7.24 for the Ti-6Al-7Nb, 3.23 for Co-Cr, and 6.89 for gold alloy). These results indicate better adaptation of Ti-6Al-7Nb clasps to tooth surfaces when designed to engage undercuts greater than 0.25 mm. A relatively higher load was needed to generate a larger permanent deformation in the clasp (21.9

Fig. 6. Average principal stress distribution for each test group. Each volume was divided into 9 different colors according to stress levels. Red zone indicates greatest stress region (tension), blue zone indicates smallest. Scales below show maximum and minimum scale values, as well as boundary values between levels. Bold numbers indicate reported 0.2% yield strengths (Table I).

N for T75) than the other materials tested because of the greater yield strength of the Ti-6Al-7Nb alloy. This provides the Ti-6Al-7Nb clasps better resistance to permanent deformation, which is advantageous in situations of accidental overloading.9,22 Greater permanent deformation of Co-Cr clasps led to reduction of the actual deflections, and this contributed to the extension of the fatigue lives because smaller actual deflections would reduce the maximum tensile stress within the clasps.12 Permanent deformation was observed in the clasps that had maximum stresses below the yield strengths of the alloys (Table II). It is suspected that these deformation values were partially due to the surface flattening of the load application sphere at the clasp tip, which was confirmed in postoperative microscopic observations. Although the calculated average maximum principal tensile stresses in clasps of the C50, C75, T75, and G75 groups exceeded the yield strengths of the respective alloys (Table II), no specimen in these groups failed during the first cycle, and all survived at least 1500 cycles. The permanent deformations did not have notable increases after the first few cycles, and the deformations remained stable until just before fracture. The residual stresses and strain hardening induced by the permanent structure deformations after the deflection can explain these findings.19,23 The residual stresses generated in the clasp after the first few cycles should have comprised compressive stress on the inner surface



and the tensile stress on the outer surface of the clasp.19,23 Besides strain hardening, these stresses could further increase the clasp’s apparent yield strength and resistance to further permanent deformation.19,23 It should also be noted that the stresses were calculated in the models by inputting the actual deflections without considering these residual stresses. This may result in higher calculated tensile stresses than those actually present in the specimens. Post-insertion adjustment or re-adaptation of a clasp to the tooth surface may cause permanent deformation in the opposite direction to that seen in this study and may generate some tensile residual stress on the inner surface. This may subsequently increase the susceptibility to further permanent deformation with the reduction of actual deflection, which could apparently extend the clasp’s fatigue life.5 The fracture locations in all of the failed specimens were found to be within the range where the stress was more than 85% of the maximum stress. This implies that the stress distribution within the clasp was closely related to the fracture location. However, since an area of more than 85% of the maximum stress is relatively wide and stress is almost constant in this area, the site of the maximum stress could not be used as a good predictor of the fracture site. It should be noted that the experimental condition in this study may differ from that encountered clinically. For example, fixed direction and magnitude were used for the experimental loading in this study, although the real direction may be different and the loading conditions can actually fluctuate.14 Also, the fatigue test in this study was conducted in dry air atmosphere which is different from the oral invironment.19 Further studies are needed to develop a sound base for a computerized methodology of design optimization that should minimize the failure possibilities of cast clasps.

The gold alloy clasps exhibited significantly longer fatigue lives, while the Ti-6Al-7Nb clasps showed significantly greater resistance to permanent deformation under cyclic deflections. The results of the fatigue analyses suggest that the Ti-6Al-7Nb and gold clasps are suitable for use with undercuts greater than 0.25 mm. However, a reduction in the fatigue resistance of the Ti-6Al-7Nb clasp was clear when the tensile stress within the clasp exceeded its yield strength.
1. Wang TJ, Kobayashi E, Doi H, Yoneyama T. Castability of Ti-6Al-7Nb alloy for dental casting. J Med Dent Sci 1999;46:13-9. 2. Watanabe K, Miyakawa O, Takada Y, Okuno O, Okabe T. Casting behavior of titanium alloy in a centrifugal casting machine. Biomaterials 2003; 24:1737-43.

3. Iijima D, Yoneyama T, Doi H, Hamanaka H, Kurosaki N. Wear properties of Ti and Ti-6Al-7Nb castings for dental prostheses. Biomaterials 2003;24: 1519-24. 4. Kobayashi E, Wang TJ, Doi H, Yoneyama T, Hamanaka H. Mechanical properties and corrosion resistance of Ti-6Al-7Nb alloy dental castings. J Mater Sci Mater Med 1998;9:567-74. 5. Vallittu PK, Kokkonen M. Deflection fatigue of cobalt-chromium, titanium, and gold alloy cast denture clasp. J Prosthet Dent 1995;74:412-9. 6. Keltjens HM, Mulder J, Kayser AF, Creugers NH. Fit of direct retainers in removable partial dentures after 8 years of use. J Oral Rehabil 1997;24: 138-42. 7. Saito M, Notani K, Miura Y, Kawasaki T. Complications and failures in removable partial dentures: a clinical evaluation. J Oral Rehabil 2002;29: 627-33. 8. Hofmann E, Behr M, Handel G. Frequency and costs of technical failures of clasp- and double crown-retained removable partial dentures. Clin Oral Investig 2002;6:104-8. 9. Carr AB, McGivney GP, Brown DT. McCracken’s removable partial prosthodontics. 11th ed. St. Louis: Elsevier; 2005. p. 79-115. 10. Vallittu PK. Fatigue resistance and stress of wrought-steel wire clasps. J Prosthodont 1996;5:186-92. 11. Gapido CG, Kobayashi H, Miyakawa O, Kohno S. Fatigue resistance of cast occlusal rests using Co-Cr and Ag-Pd-Cu-Au alloys. J Prosthet Dent 2003;90:261-9. 12. Yuasa Y, Sato Y, Ohkawa S, Nagasawa T, Tsuru H. Finite element analysis of the relationship between clasp dimensions and flexibility. J Dent Res 1990;69:1664-8. 13. Sato Y, Yuasa Y, Akagawa Y, Ohkawa S. An investigation of preferable taper and thickness ratios for cast circumferential clasp arms using finite element analysis. Int J Prosthodont 1995;8:392-7. 14. Sato Y, Abe Y, Yuasa Y, Akagawa Y. Effect of friction coefficient on Akers clasp retention. J Prosthet Dent 1997;78:22-7. 15. Wiskott HW, Nicholls JI, Belser UC. Stress fatigue: basic principles and prosthodontic implications. Int J Prosthodont 1995;8:105-16. 16. Tamhane AC. Multiple comparisons in model I one way ANOVA with unequal variances. Comm Statist Theory Methods 1977;6:15-32. 17. Tamhane AC. A comparison of procedures for multiple comparisons of means with unequal variances. J Am Statist Assoc 1979;74:471-80. 18. Anusavice KJ, Cascone P. Dental casting and soldering alloys. In: Anusavice KJ, editor. Philips’ science of dental materials. 11th ed. St. Louis: Elsevier; 2003. p. 604. 19. Dieter GE. Mechanical metallurgy. 3rd ed. New York: McGraw-Hill; 1986. p. 48, 375-431. 20. Bates JF. Studies related to the fracture of partial dentures; flexural fatigue of a cobalt-chromium alloy. Br Dent J 1965;118:532-7. 21. Brantley WA. Wrought alloys. In: Anusavice KJ, editor. Philips’ science of dental materials. 11th ed. St. Louis: Elsevier; 2003. p. 623. 22. Warr AJ. Numerical system of clasp design. J Prosthet Dent 1961;11: 1105-11. 23. Beer FP, Johnston ER, Dewolf JT. Mechanics of materials. 3rd ed. New York: McGraw-Hill; 2002. p. 209-306. 24. Morris HF, Asgar K, Rowe AP, Nasjleti CE. The influence of heat treatments on several types of base-metal removable partial denture alloys. J Prosthet Dent 1979;41:388-95. Reprint request to: DR NORIYUKI WAKABAYASHI REMOVABLE PROSTHODONTICS, DIVISION OF ORAL HEALTH SCIENCES GRADUATE SCHOOL, TOKYO MEDICAL AND DENTAL UNIVERSITY 1-5-45 YUSHIMA, BUNKYO TOKYO 113-8549 JAPAN FAX: 181-3-5803-584 E-MAIL: 0022-3913/$30.00 Copyright Ó 2005 by The Editorial Council of The Journal of Prosthetic Dentistry.




Sign up to vote on this title
UsefulNot useful