Biomaterials 22 (2001) 1675}1681

Bond strength of binary titanium alloys to porcelain

Masanobu Yoda

*,Tatsuhiko Konno ,Yukyo Takada,Kazunori Iijima,

Jason Griggs

,Osamu Okuno,Kohei Kimura ,Toru Okabe

Division of Fixed Prosthodontics, Department of Oral Rehabilitation and Materials Science, Graduate School of Dentistry,

Tohoku University, 4-1 Seiryou-machi, Aoba-ku, Sendai 980-8575, Japan

Division of Dental Biomaterials, Department of Oral Rehabilitation and Materials Science, Graduate School of Dentistry,

Tohoku University, 4-1 Seiryou-machi, Aoba-ku, Sendai 980-8575, Japan

Department of Biomaterials Science, Baylor College of Dentistry, Texas A&M University System Health Science Center,

3302 Gaston Ave., Dallas, TX 75246, USA

Received 8 February 2000; accepted 10 October 2000

Abstract

The purpose of this study was to investigate the bond strength between porcelain and experimental cast titanium alloys. Eleven

binary titanium alloys were examined: Ti}Cr (15,20,25 wt%),Ti}Pd (15,20,25 wt%),Ti}Ag (10,15,20 wt%),and Ti}Cu (5,10 wt%).
As controls,the bond strengths for commercially pure titanium (KS-50,Kobelco,Japan) and a high noble gold alloy (KIK,Ishifuku,
Japan) were also examined. Castings were made using a centrifugal casting unit (Ticast Super R,Selec Co.,Japan). Commercial
porcelain for titanium (TITAN,Noritake,Japan) was applied to cast specimens. The bond strengths were evaluated using
a three-point bend test according to ISO 9693. Since the elastic modulus value is needed to evaluate the bond strength,the modulus
was measured for each alloy using a three-point bend test. Results were analyzed using one-way ANOVA/S-N}K test (

"0.05).

Although the elastic moduli of the Ti}Pd alloys were signi"cantly lower than those of other alloys (p"0.0001),there was a signi"cant
di!erence in bond strength only between the Ti}25Pd and Ti}15Ag alloys (p"0.009). The strengths determined for all the
experimental alloys ranged from 29.4 to 37.2 MPa,which are above the minimum value required by the ISO speci"cation
(25 MPa).

2001 Elsevier Science Ltd. All rights reserved.

Keywords: Binary titanium alloys; Porcelain fused to metal; Bond strength; Three-point bend test; Metal}ceramic bonding

1. Introduction

Titanium is an attractive dental restorative material

because of its excellent characteristics such as biocom-
patibility,corrosion resistance,light weight,and high
strength. It has been increasingly used for dental applica-
tions because of progress in dental casting technology
such as in investment materials and casting machines [1].
However,there are some inherent problems that must be
overcome for titanium to be used successfully in den-
tistry. The fact that pure titanium has a high melting
temperature and high reactivity with oxygen and impu-
rities at elevated temperatures makes it di$cult to cast
[2]. One method of solving these problems is to use
titanium alloys. It is known that

-titanium alloys exhibit

solid solution hardening and have lower fusion temper-

* Corresponding author. Tel.: #[phone]; fax: #[phone].
E-mail address: [e-mail] (M. Yoda).

atures and better ductility than commercially pure (CP)
titanium [3,4]. Takada et al. [5] tested the mechanical
properties and corrosion behavior of experimental bi-
nary titanium alloys (Ti}Ag,Ti}Cr,Ti}Co,Ti}Cu,
Ti}Fe,Ti}Mn,and Ti}Pd) and reported that some of
these alloys have the potential for use in dental applica-
tions. For dental clinical use,these alloys must be able to
bond to dental porcelains because of the esthetic de-
mands of dentistry. The purpose of this study was to
evaluate the bond strength of experimental binary tita-
nium alloys to dental porcelain.

2. Materials and methods

2.1. Materials

The chemical compositions of the experimental alloys

and the commercial alloys used as controls (KS50 pure

0142-9612/01/$ - see front matter

2001 Elsevier Science Ltd. All rights reserved.

PII: S 0 1 4 2 - 9 6 1 2 ( 0 0 ) 0 0 3 2 9 - X

Table 1
Experimental and commercially available alloys and porcelains used in this study

Alloys

Chemical compositions (wt%)

Alloy phases

Melting ranges (3C)

Porcelains

Ti}Ag

10,15,20

1640}1530

Ti}Cr

15,20,25

1600}1460

Super Porcelain

Ti}Cu

5,10

, #TiCu

1600}1240

TITAN

Ti}Pd

15,20,25

#,

1570}1340

(Noritake)

Pure titanium (KS50)

1683

Gold alloy (KIK)

1170}1200

Vita VMK68 (Vita)

titanium,Kobelco,Japan and KIK high noble gold alloy,
Ishifuku,Japan) are shown in Table 1. This table also
includes the alloy phases,melting ranges and porcelains
applied to these alloys. In this study,four alloying
elements (Ag,Cr,Cu,and Pd) out of eight tested in
a previous study [5] were investigated. The chemical
compositions of each alloy system were chosen according
to their bulk hardness and corrosion behavior found in
the previous study. The pure titanium used was JIS grade
II,which corresponds to ASTM grade 2. A commercial
porcelain for titanium (Noritake Super TITAN,Nori-
take,Japan) was applied to the experimental alloys and
pure titanium. A commercial porcelain for gold alloy
(VITA VMK68,VITA Zahnfabrik,Germany) was ap-
plied to the KIK gold alloy.

2.2. Methods

Titanium alloy buttons for casting were made in the

same manner as in the previous study [5].

2.2.1. Casting

A thin acrylic plate pattern (1.0 mm

;4.0 mm;30 mm)

was invested in a magnesia-based investment material
(Selevest CB,Selec Co.,Japan); then the invested mold
was burned out as the manufacturer recommended,i.e.,
the mold was heated to 8503C at 63C/min and held at this
temperature for 1 h. The mold was furnace-cooled to
3503C and subsequently placed in a casting machine. The
molten metal was cast into the mold in a centrifugal
casting machine (Ticast Super R,Selec Co.,Japan) at
a rotational frequency of 3000 rpm. Immediately after
casting,the mold was quenched in water. Nine castings
were made for each alloy composition. Three of them
were used for elastic modulus evaluation,and six were
used for bond strength evaluation.

2.2.2. Grinding and polishing

To ensure complete removal of the

-case layer [5],

a thickness of at least 200

m was ground from all six

surfaces

of

the

cast

specimens

with

SiC

paper

(120}600 grit). All dimensions were controlled by multi-
point measurements with a digital caliper (Digimatic
NTD15P-6

C,Mitutoyo,Japan). After polishing,each

specimen measured 0.45}0.55 mm

;3.5 mm;25 mm.

2.2.3. Elastic modulus

To evaluate the bond strengths between alloy and

porcelain according to ISO 9693 [6],the elastic moduli
of the alloys are needed. In this study,three of the
nine polished specimens were used to evaluate the elastic
modulus for each alloy. Porcelain "ring cycles were
applied to these specimens but no porcelain was applied.
These "ring cycles were as follows: (1) preheating
for oxidation,(2) "ring to bonding porcelain,(3) "ring
to opaque porcelain,(4) "ring to dentin porcelain,and
(5) glazing. They were carried out according to the
manufacturers'

recommendations

as

described

in

Table 2.

After the "ring cycles,the elastic moduli of the speci-

mens were evaluated with a three-point bend test. The
specimens were tested in an Instron universal testing
machine (model 1125,Instron Co.,USA) at a crosshead
speed of 1 mm/min. The specimens were placed on a cus-
tom-made loading jig with rounded supporting rods
(H1.8 mm) 20 mm apart and loaded in the center with
a rounded loading rod. The load and de#ection at the
mid-span of the specimens were recorded. The elastic
moduli in bending were calculated using the following
formula [7]:

E+"

¸

P

4bh

where E+ is the elastic modulus in bending, L the span

length (20 mm), h the specimen thickness (0.45}0.55 mm),
b the specimen width (3.50 mm),

P the load increment

measured from preload,

the de

#ection increment at

mid-span as measured from preload.

These data were analyzed using one-way ANOVA

followed by Student}Newman}Keuls pairwise tests
(

"0.05).

2.2.4. Bond strength between alloy and porcelain

The remaining six specimens of each alloy were tested

for bond strength. The surface of each polished porcelain
specimen was sandblasted with 50

m diameter alumina

particles at a pressure of 4 kg/cm

and cleaned with

acetone in an ultrasonic cleaner for 10 min. The porcelain

"ring procedure was carried out according to the manu-

facturers' instructions (Table 2).

1676

M. Yoda et al. / Biomaterials 22 (2001) 1675} 1681

Table 2
Porcelain "ring conditions of Super Porcelain TITAN (Vita VMK68)

Start temp. (3C)

Heat rate (3C/min)

Final temp. (3C)

Holding time (min)

Vacuum (cm/Hg)

Oxidation

500 (600)

50 (50)

800 (980)

3 (3)

74 (74)

Bonding

500 (*)

50 (*)

790 (*)

* (*)

72 (*)

Opaque

500 (700)

50 (25)

770 (940)

* (*)

72 (72)

Dentin

500 (700)

40 (40)

750 (950)

* (*)

72 (72)

Glazing

500 (700)

50 (40)

760 (950)

* (2)

* (*)

Fig. 1. An example of specimen and diagram of three-point bend test.

After oxidation,bonding porcelain was applied to the

center of the sandblasted surface over a length of 8 mm.
After the bonding porcelain was "red,it was ground to
a thickness of 0.15 mm,sandblasted with 50

m diameter

alumina at a pressure of 2 kg/cm

and cleaned. Then

opaque porcelain was applied to the polished bonding
porcelain particles, "red,and ground to a thickness of
0.15 mm.

Dentin porcelain was "red onto the opaque porcelain

using a custom-made jig that controlled the position and
thickness of the dentin porcelain. To compensate for "ring
shrinkage,another application of dentin porcelain was
made. After "ring,the dentin porcelain was ground to
a thickness of 0.7 mm,so that the total porcelain thickness
was 1.0 mm and the total specimen width was 3.0 mm.
Finally,the specimens were subjected to a glaze "ring. The

"nal geometry of the specimen is shown in Fig. 1.

After glazing,the bond strength between alloy and

porcelain was tested. The specimens were tested in the
same testing machine and jig as used to evaluate the
elastic moduli. Each specimen was positioned with the
porcelain side opposite the center support,and force was
applied at a crosshead speed of 1 mm/min until fracture.
The maximum in the force displacement curve was re-
corded as the breaking force, F.

The bond strength was calculated according to the

formula in ISO 9693 [6]. In this standard,the bond

strength is determined by failure load (F)

;coe

$cient (

k).

These data were statistically analyzed using one-way
ANOVA followed by Student}Newman}Keuls pairwise
tests (

"0.05).

2.2.5. Coezcient of thermal expansion

Because the coe$cient of thermal expansion of each

alloy and porcelain has a great in#uence on the net
interfacial stress [8],the linear thermal expansion of each
material used in this study was measured according to
the procedure speci"ed in ISO 9693. Rectangular
beam specimens (15 mm

;5 mm;5 mm) were made

from three types of Super Porcelain TITAN (bonding,
opaque,and dentin) and two types of VITA VMK68
porcelain (opaque and dentin). Four specimens were
made for each type. After "ring,they were ground and
polished. Two of the specimens were "red again (i.e.,
a total of two "rings). The remaining specimens were

"red three more times (i.e.,a total of four "rings). The

alloys were cast,ground,and polished in the same man-
ner as for the specimens used in the bond strength evalu-
ation. Two specimens were prepared for each alloy.
Before evaluation of the expansion behavior,the speci-
mens were heated according to the same procedure of
preheating for oxidation as for the bond strength speci-
mens.

M. Yoda et al. / Biomaterials 22 (2001) 1675} 1681

1677

Fig. 2. Elastic moduli of experimental and commercially available alloys.

Thermal expansion was tested using a push-rod dila-

tometer (model DL-1500RH,Shinku-Riko,Japan) at
a heating rate of 53C/min over temperature ranges of
25}5503C and 253C to the softening point for alloys and
porcelains,respectively. For each specimen,the coe$c-
ient of thermal expansion was determined between 25
and 5003C (or between 253C and the glass transition
temperature if that temperature was lower than 5003C)
from the plotted curve of expansion versus temperature.

3. Results

3.1. Elastic modulus

The elastic modulus results are shown in Fig. 2. The

alloys tested were divided into four groups by one-way
ANOVA followed by Student}Newman}Keuls pairwise
tests (p"0.0001). The elastic moduli of the Ti}Pd alloys
(15,20,and 25 wt%) were classi"ed as the lowest
modulus group (d),and these values were signi"cantly
lower than those of the other groups (a}c). The elastic
moduli for the Ti}Pd alloys were approximately 80 GPa,
and they were 62}67% of that for Ti}10 wt% Cu
(122 GPa). The values for Ti}10 wt% Cu,Ti}20 wt% Ag,
KS50,and Ti}25 wt% Cr were in the range of
120}122 GPa. These were classi"ed as the highest
modulus group (a). As for the Ti}Cr alloys,the elastic
modulus values had a tendency to increase as the chro-
mium content increased.

3.2. Bond strength

After

porcelain

"ring,the

Ti}15 wt%

Pd

and

Ti}20 wt% Pd specimens de#ected toward the porcelain
side,and the degree of de#ection was larger in the

Ti}15 wt% Pd specimens. The other specimens,including
Ti}25 wt% Pd,did not de#ect prior to loading.

Calculated bond strengths are shown in Fig. 3. The

mean bond strengths ranged from 29.4 MPa (Ti}25 wt%
Pd) to 37.2 MPa (Ti}15 wt% Ag) including KS50
(32.2 MPa) and KIK (30.5 MPa) as controls. The bond
strengths of the alloys tested were divided into two
groups by one-way ANOVA (p"0.0009). However,the
only signi"cant di!erence was between Ti}15 wt% Ag
and Ti}25 wt% Pd.

In the Ti}Cr alloys,bond strength tended to increase

as the Cr component increased. In contrast,the Ti}Pd
alloys followed an opposite trend from the Ti}Cr alloys.
The Ti}Pd alloy bond strengths had a tendency to in-
crease as the Pd component decreased. None of the
di!erences in bond strength between compositions of the
same alloying element were statistically signi"cant. There
was no consistent trend in bond strength based on com-
position for the Ti}Ag alloys.

3.3. Thermal expansion coezcient

Thermal expansion results are shown in Table 3. The

specimens "red two times and the specimens "red four
times exhibited almost the same values for each type of
porcelain. The coe$cients of thermal expansion of the
experimental alloys ranged from 9.8 ppm/3C (Ti}5 wt%
Cu and Ti}10 wt% Cu) to 11.3 ppm/3C (Ti}25 wt% Cr)
and were higher than that for pure titanium (KS50)
(9.5 ppm/3C).

4. Discussion

In the present study,the bonding strength to porcelain

was investigated by removing a thickness of at least

1678

M. Yoda et al. / Biomaterials 22 (2001) 1675} 1681

Fig. 3. Bond strengths of experimental and commercially available alloys to dental porcelains.

Table 3
Thermal expansion coe$cient of experimental and commercially available alloys and porcelains (ppm/3C)

Ti}Cr wt%

Ti}Pd wt%

Ti}Ag wt%

Ti}Cu wt%

Pure
titanium

Gold
alloy

TITAN

VMK68

15

20

25

15

20

25

10

15

20

5

10

KS50

KIK

B

O

D

O

D

10.7

10.9

11.3

10.5

10.0

10.3

10.0

10.1

10.3

9.8

9.8

9.5

14.3

9.7

7.4

7.3

13.4

12.9

10.7

7.1

7.2

14.0

13.0

Mean (n"2),B; bonding porcelain,O; opaque porcelain,D; dentin porcelain (upper row; "red two times specimen,lower row; "red four times

specimen).

200

m of the -case layer from all surfaces of the alloy

specimens. In the clinical situation,grinding the inside
surfaces of cast crowns is not recommended because the

"t of the cast crown to the prepared tooth structure must

be maintained. However,the outer surfaces of the crown
must be ground. Since the thickness of the

-case layer on

alloys was previously found to be 150}200

m [5],all of

the specimen surfaces were ground so that the bond
strength would not be a!ected by the surface layer.

To evaluate the bond strength between porcelain and

metal,various methods have been tried,including the
push-pull test,shear test,and tensile test. Anusavice et al.
[9] used "nite element stress analysis to show that
a three-point bend test is an appropriate method among
the

many

methods

available.

More

recently,

Schwickerath [10],Schwarz et al. [11],and Lenz et al.
[12] performed "nite element stress analyses and de-
veloped a method of evaluating bond strength that was
subsequently adopted as a standard for ISO 9693 [6].
The thicknesses of metal and porcelain speci"ed in this
standard are similar to those found in the clinical situ-

ation; thus,this method appears to be clinically relevant.
Therefore,ISO 9693 was chosen for use in this study to
evaluate metal}porcelain bond strength.

4.1. Modulus of elasticity

The modulus of elasticity for CP-Ti (KS-50) was

120 GPa,which was higher than previously reported
values of 104}120 GPa [1,4,13]. The discrepancy among
values may be due to di!erences in test methods,as the
previously reported values were measured using a tensile
test instead of three-point bending. The modulus of elas-
ticity obtained for the gold alloy control (KIK) (107 GPa)
was considered to be acceptable since modulus values for
a gold alloy with a similar chemical composition were
reported to be 89}118 GPa [12].

In the present study,the moduli of elasticity of all

the Ti}Pd alloys examined were found to be approxim-
ately 80 GPa,which is lower than that for the other
alloys. It is not known why these alloys exhibited lower
moduli.

M. Yoda et al. / Biomaterials 22 (2001) 1675} 1681

1679

Two factors may have contributed to the low modulus

of elasticity of the Ti}Pd alloys: (1) the Ti}Pd alloys
contained a greater proportion of

-titanium phase and

(2) the

-titanium phase of Ti

}Pd alloys itself might have

the lower modulus compared to the

-titanium phase of

the other titanium alloys. According to the equilibrium
phase diagram of the Ti}Pd alloy system [14],the Ti}Pd
alloy system has the lowest transition temperature
(630}7803C) for the

}

phase change among the alloys

examined in the present study (Ti}Cr,715}9553C; Ti}Ag,
860}9153C and Ti}Cu,820}8753C). In addition,the eu-
tectoid temperature of the Ti}Pd alloy system is 5953C
(Ti}Cr,6673C; Ti}Ag,8553C; and Ti}Cu,7903C). In the
present study,the alloy specimens were thermocycled
several times (from room temperature up to 8003C) in
order to simulate the steps that the cast and polished
specimens undergo in the porcelain "ring procedure.
Some of the

phase retained in the grains in the cast

microstructure might have transformed to the

phase.

However,in the present three Ti}Pd alloys and the
15 wt% Cr alloy,the highest thermocycling temperature
was higher than the

}

phase transformation temper-

ature and therefore,more

phase should have been

retained than in the other alloys. On the other hand,the
phase should have formed in other alloys since the
thermocycling temperature is lower than that of the

}

phase transformation temperature. The

-stabilizing

power values for the elements Cr and Pd estimated from
the equation by Ja!ee [15] were 18 and 23,respectively,
indicating that this higher

-stabilizing power by Pd was

responsible for the higher amount of

phase retained in

this type of alloy.

4.2. Bond strength

Including the bond strength for the control,the

strengths determined for all the experimental alloys
ranged from 29.4 to 37.2 MPa,which is above the min-
imum value speci"ed in ISO 9693 (25 MPa). The strength
of the gold alloy (KIK) control was 30.5 MPa,and the
corresponding values reported earlier for similar gold
alloys were 26.6}39.2 MPa [12]. The bond strength
found in the present study for CP-Ti was 32.2 MPa,
which falls within the range of values reported by
Probster et al. [16] (21.4}34.0 MPa). Thus,the values
determined in the present study were considered to be
reasonable.

The strengths of all the experimental alloys were not

signi"cantly di!erent compared to those of the two con-
trols,indicating that all the experimental alloys de-
veloped an acceptable bond strength to the porcelain.

4.3. Deyection of some of the Ti}Pd alloys used

It was observed that the surfaces of the 15 and 20 wt%

Pd alloy specimens on which the porcelain was "red

curved slightly toward the porcelain side upon cooling.
The de#ection was greater for the 15 wt% Pd alloy; no
bending was seen in the 25 wt% Pd alloy. This de#ection
was probably caused by residual stress from the thermal
expansion mismatch between the alloy and porcelain.
The thermocycled specimens without porcelain did not
bend,and the direction of de#ection was always toward
the porcelain surfaces. The phenomenon observed here
could be an important "nding since it would a!ect the "t
of a dental crown. Warping usually occurs because of
a di!erence in the coe$cient of linear thermal expansion
between the alloy and porcelain. However,the thermal
expansion value of the Ti}Pd alloys in which the de#ec-
tion was observed was not much di!erent from that of
the other alloys in the present study. Thus,the de#ection
observed was not totally related to the thermal expansion
coe$cient.

Since these alloys may have de#ected noticeably,and

the de#ection did not correct itself after the porcelain was
debonded during the bonding test,it can be said that the
alloys underwent creep while the porcelain was cooling
to room temperature from the glass transition temper-
ature. Tuccillo and Nielsen [17] studied the de#ection of
metals and porcelains during the entire "ring process
(from heating to cooling). They pointed out that the
bending was in#uenced more by the incompatibility of
a speci"c portion of the thermal expansion curves of
metals and porcelain than by the di!erence in the average
coe$cients of thermal expansion. They also indicated
that the de#ection was largely a!ected by the creep
resistance of both metals and porcelains.

As described earlier,the phase transformation temper-

ature of the Ti}Pd alloys was lower than that of the other
alloys examined. Thus,the

P phase transformation

occurred in the alloy during the solidifying process at
temperature lower than ¹ (the dilatometric softening

point,6003C) of the porcelain (TITAN) or ¹ (glass

transition temperature,4943C). In addition,the alloy
could de#ect because of the stress developed from the
shrinkage of the porcelain. In cases where the stress level
diminishes because of the phase transformation,the alloy
could also be permanently deformed (i.e.,it undergoes
creep),resulting in a de#ected specimen. In general,an
increasing amount of alloying element slows down the
rate of phase transformation [18]. The 15% Pd alloy
exhibited the highest de#ection because the

P phase

transformation was the easiest to accomplish,since the
amount of Pd was the lowest of the Ti}Pd alloys studied.
In contrast,the least de#ection was found in the 25% Pd
specimen.

The Ti}15 wt% Cr alloy was heated at temperatures

higher than its phase transformation temperature. The
eutectoid temperature of this alloy is higher than that of
the Ti}Pd alloy system. In addition,the alloy specimen
cooled at a relatively fast rate after removal from the
furnace at 760}8003C. Therefore,the degree of the phase

1680

M. Yoda et al. / Biomaterials 22 (2001) 1675} 1681

transformation of this alloy was minimal,and yet no
deformation occurred.

In any case,the degree of de#ection depends on the

di!erence in the thermal contraction at the temperature
when the porcelain begins to solidify. Further study is
needed in this area.

The de#ection found in some specimens before the

bond test is believed to have a!ected the results of the
strength calculation. However,in the present study,cal-
culations were made without taking the bending into
consideration. It is estimated that the calculated bond
strength would be no more than 5% greater if the e!ect
of the geometrical de#ection were considered [19]. In the
future, "nite element analysis studies should be used to
determine the e!ect of pre-load de#ection on strength.

4.4. Topics for future study

In the present study,four di!erent binary titanium-

alloying elements out of the seven elements studied by
Takada et al. [5] were evaluated for bond strength to
porcelain. The other three elements in Takada's study
should also be examined. In addition,an SEM/EPMA
examination of the metal}porcelain interface will be per-
formed,and the e!ect of the

-case layer on bonding

strength should be examined.

5. Conclusions

The titanium alloy composition had a signi"cant e!ect

on alloy elastic modulus. The Ti}Pd alloys had the
lowest elastic moduli,and the Ti}Cu alloys had the
highest elastic moduli.

The titanium alloy composition had little e!ect on the

bond strength to porcelain. The bond strengths for all the
experimental alloys were similar to those of the two
control alloys (KIK and KS50),which indicates that the
experimental alloys have su$cient bond strength for
clinical use.

Acknowledgements

This research was partially supported by NIH/NIDR

grant DE 11787. The authors thank Mrs. Jeanne Santa
Cruz for editing the manuscript.

References

[1] Okabe T,Her+ H. The use of titanium in dentistry. Cells Mater

1995;5:211}30.

[2] Okuno O,Shimizu A,Miura I. Fundamental study on titanium

alloys for dental casting (in Japanese). J Jpn Dent Mater
1985;4:708}15.

[3] Shimizu A. Studies on titanium alloys for dental casting. Part 1.

E!ects of Pd and Cr on titanium properties (in Japanese). J Jpn
Dent Mater 1986;5:122}32.

[4] Okuno O,Hamanaka H. Application of beta titanium alloys in

dentistry. Dentistry Jpn 1989;26:101}4.

[5] Takada Y,Okuno O,Nakajima H,Okabe T. Evaluation of

binary metastable

-titanium alloys for dental applications.

J Dent Res 1997;76:402,Abstract 3111.

[6] ISO 9693: 1999(E). Metal}ceramic dental restorative systems,2nd

ed. Switzerland,International Organization for Standardization,
1999.

[7] Guha A. Bending strength tests. In: Murray JL,editor. ASM

metals handbook,9th ed,vol. 8. Ohio: ASM International,1985.
p. 132}6.

[8] Lenz J,Kessel S. Thermal stresses in metal-ceramic specimens for

the ISO crack initiation test (three-point #exure bend test). Dent
Mater 1998;14:277}80.

[9] Anusavice KJ,Deho! PH,Fairhurst CW. Comparative evalu-

ation of ceramic-metal bond tests using "nite element stress anal-
ysis. J Dent Res 1980;59:608}13.

[10] Schwickerath H. Zur verbundfestigkeit von metallkeramik

(Fusion strength of metal ceramics) (in German). Dtsch Zahnarztl
Z 1980;35:910}2.

[11] Schwarz S,Lenz J,Schwickerath H. Zur Festigkeit des metal-

lkeramischen verbundes bei der biegeprufung. (Metalloceramic
bond strength in bending tests) (in German). Dtsch Zahnarztl
Z 1988;43:1152}8.

[12] Lenz J,Schwarz S,Schwickerath H,Sperner F,Schafer A. Bond

strength of metal}ceramic systems in three-point #exure bond
test. J Appl Biomat 1995;6:55}64.

[13] Smithells CJ. Elastic properties and damping capacity. In: Metals

reference book,4th ed,vol. 3. London: Butterworths,1967.
p. 708}11.

[14] Murray JL. Binary alloy phase diagrams. In: Baker H,editor.

Alloy phase diagrams. Ohio: ASM International,1987.

[15] Ja!ee RI. The physical metallurgy of titanium alloys. In: Chal-

mers B,editor. Progress in metal physics,vol. 7. London: Per-
gamon Press,1958. p. 65}163.

[16] Probster L,Maiwald U,Weber H. Three-point bending

strength of ceramics fused to cast titanium. Eur J Oral Sci 1996;
104:313}9.

[17] Tuccillo JJ,Nielsen JP. Shear stress measurements at a dental

porcelain gold bond interface. J Dent Res 1972;51:626}33.

[18] Kimura K. The recent titanium and titanium alloys (in Japanese).

Bull Jpn Inst Met 1970;9:620}38.

[19] Roark RJ. Curved beams. In: Young WC,editor. Roark's for-

mulas for stress strain,6th ed. New York: McGraw-Hill,1989.
p. 236}343.

M. Yoda et al. / Biomaterials 22 (2001) 1675} 1681

1681