Heinrich J.G., Aldinger F. (Eds.) Ceramic Materials and Components for Engines

gas flow rates of the propane process

(-

factor

1.6)

com-

pared to the acetylene process and the less degree of parti-

cle fusion higher compressive stresses can be measured. In

general

it

can be said that with increasing compressive

stresses in coating and interface the hardness and the bond

strength of the layer coatings increase.

Mo

Mo/NiCrBSi-30/70

200,o

-

3-4 1000

-

1200

6

-

8

5

700

-

1100 1000

-

1200

-

g

100.0

0,o

f

-100.0

f

-200.0

a

g

300.0

dM

n

8.APS;

wilbul

mofing

+HVOF.

pmpivm:

CO2

moling

-w>"

*

0

0,05

0.1

0.15 0.2

0.25

0.3

0.35

0.4

ddlllng

depth

[mm]

Fig.

3

Measured residual stresses in Ti02 coated cylinder

bores, variation of the thermal spraying processes

(APS, HVOF)

The investigations of the bond strength are performed on a

universal testing machine. A steel tension rod is glued on

the coating surface. As soon as delamination of the coat-

ing

occurs,

the tension load is measured and the bond

strength is calculated. Figure

4

shows the measured

bonding strength results for the Ti02-A1Si9Cu layer com-

posite manufactured with different spray processes and

cooling parameters.

60

,---.I_.-,.

CQlidr

(pacum

mdlng

ding

aid'

Fig.

4

Measured bonding strength of TiOz-AlSi9Cu layer

composites

In general

it

can be said that with increasing compressive

stresses

in

the interface also the bond strength increases

[5].

For

the HVOF propane oxygen sprayed Ti02 system

no bond strength can be measured because of the lower

yield strength of the used glue. It

just

can be mentioned

that the bond strength of the HVOF propane oxygen

sprayed Ti02 coatings compared to the APS Ti02 coatings

is by a factor of at least 1.8 higher.

MECHANICAL COATING

CHARACTERIZATION

To assess coating quality, manufacturing repro-

ducibility and the operation behaviour of the layer com-

posites metallographic examinations were performed for

all samples. By image analysis the coating porosity, ex-

pressed by the relative pore volume content Vp

[%],

is

defined using cross section evaluation of metallografic

samples. The hardness HV0.05 is measured with an auto-

mated universal hardness indenter equipment (F

=

500

mN).

Table

2

shows the measured porosity and hardness

characteristics

for

both coating processes.

I

Vp[Oh]

I

VP~O]

I

HV0.05

I

HV0.05

1

APS

HVOF

APS

HVOF

Coating

CrB/NiCr-75/25

I

18

I

13

I

1500

I

1300- 1600

I

Mo/Cr-Steel-70/30

I

6

-

10

I

4

-

5

1

400

-

500

I

700

-

1100

I

Table

2

Measured coating porosity and coating hardness

WEAR AND FRICTION

The aim of the material screening is the detection

of tribological systems with low friction coefficients as

well as low wear rates on coating and counterpart (piston

ring). Low friction coefficients are required to reduce the

energy dissipation during operation.

A

decrease of the

wear rate during operation increases lifetime and quality

of the tribological system and reduces the pollution of the

lubricating oil with particles. To get an overview about the

wear and friction behaviour of the different coating sys-

tems, dry running oscillating pin on disc tests are per-

formed. The coatings are finished to a surface roughness

of

R,

-

0.05

pm. As a counterpart A1203 balls with a di-

ameter of

5

mm are used. The number of oscillating

strokes is up to

50.000,

sliding velocity

70

mds, length

of strokes

5

mm with a imposed measurement load of

10

N.

For all thermally sprayed coatings friction coefficients

p

below 0.9 can be measured, compare figure

5.

Very low friction coefficients between

0.1

an

0.2

can be measured for some Ti02 APS coatings. Addition-

ally a very interesting correlation between hardness and

friction coefficient can be detected for the Ti02 coating

202

systems. The coating systems with the low hardness

HV0.05 values and friction coefficients

are

sprayed with

an increased hydrogen percentage by optimized Cot and

air jet cooling. Therefor it can be assumed that non-

stoichiometric TinOzn., phases are responsible for this

excellent friction behaviour. The wear investigations show

similar results.

j

A

APSCr3C21NiCr

0

HVOF-Gr3CZNtCr

.........................

:

.........................

i

AAPSCR03 OHVOF-CQ03

0

HVOF-Ti02

I

......................................................................

.....................

j

........................

::

........................

4

0,1

I

0

1000

2000

3000

4000

coating hardness

HV0.05

Measured average friction coefficient

p

vs. coat-

ing hardness HV0.05

Fig.

5

CONCLUSION

Intentions to reduce manufacturing cost, fuel

consumption and waste emissions in the automobile in-

dustry result in increasing light weight design and appli-

cations. The poor tribological operation behaviour of light

metal surfaces can be improved by protective coating

systems. Beside galvanic coatings and reinforced materi-

als, thermally sprayed coatings offer a broad material

variety as well as flexible and cost effective manufactur-

ing processes for refined crankcases. By means of APS

and

HVOF

spraying wear resistant inside coatings are

deposited on aluminum tubes and cylinder bores

of

com-

bustion engines. During the material screening different

ceramic, cermet and metallurgical coating systems were

deposited and investigated. The thermophysical properties

cx

and cp were defined. Residual stress measurements have

shown that the stresses

in

thermally coated layer compos-

ites strongly depend on the temperature level and history

during the coating process as well as on the particle im-

pact velocity. Wear investigations show friction coeffi-

cients lower than

0.9

for all coatings under dry conditions.

In the case of plasma sprayed Ti02 coatings excellent

friction coefficients between

0.1

and

0.2

can be measured

under dry running conditions. The investigations during

the material screening and first bench tests reflect positive

results for most of the coating systems.

A

detailed coating

selection can be done after further test runs of combustion

bench tests

in

stationary reciprocating engines. It can be

assumed that thermally sprayed coatings will play an

important role as protective coatings for light metal engine

applications and will continue to become more and more

important in the near future.

REFERENCES

[I] Hinz

R.,

Schwaderlapp M.; "Potential

zur

Massenre-

duktion am Beispiel eines

4-Zylinder-Reihenmotors";

Leichtbau im Antriebsstrang (1 996), Expert Verlag, ISBN

3-8169-1336-9, pp. 162

-

173

[2]

Woydt M., Skopp A., Dorfel

I.,

Witke K.;"Wear Engi-

neering Oxides

/

Antiwear Oxides"; Tribology Transac-

tions, Volume 42

(1

999),

1,

pp.

2

1

-

3

1

[3] Woydt M.; "Materials-based concepts for an oil-free

engine"; New Directions in Tribology (1997), pp. 459

-

468

[4] Buchmann

M.,

Gadow R., Tabellion

J.;

"Experimental

and numerical residual stress analysis of coated

CORIPOS-

ites"; E-MRS Spring Meeting Strasbourg 1999,

in

print

in

Mat.Sci.Eng.

[5] Buchmann M., Friedrich C., Gadow R.

;

"Residual

stress characterization of thermal barrier coatings

-

com-

parison of thermally sprayed, EB-PVD and CVD based

coatings"; 24th Annual Cocoa Beach Conference

&

Expo-

sition

2000,

in print

203

This Page Intentionally Left Blank

TRIBOLOGICAL BEHAVIOR

OF

SILICON NITRIDEBTEEL CONTACTS

UNDER LUBRICATED CONDITIONS

Hyung

K.

Yoon, Seung-Kun

Lee,

Frank

A.

Kelley, Michael

J.

Readey

Advanced

Materials

Technology

Caterpillar Inc.,

USA

ABSTRACT

The tribological properties of some silicon nitrides for

valve train components were evaluated under

lubricated conditions. The silicon nitrides tested are

SN235, SN237 (Kyocera, Japan) and AS800

(AlliedSignal Inc., USA). For comparison purposes,

limited tests were also conducted with a hardened

52100 steel (60 Rc). These materials were tested

against

SAE

4620 steel in a block-on-ring tester with a

polyalphaolefin (PAO) oil. The results show that

AS800 gives lower friction and wear compared to other

silicon nitrides at higher loading conditions. All silicon

nitrides exhibited much higher wear resistance

compared to a hardened 52100 steel. The effects of

contact pressure

(P)

and sliding speed

(V)

on friction

and wear characteristics were also examined. The

friction coefficient and wear increase

as

the contact

pressure increases. Wear also increased with sliding

speed, however, the coefficient of friction remains

about the same as the sliding speed increases. Silicon

nitrides exhibited mild polishing wear in most

conditions. However, a transition in the wear mode

from polishing to abrasive wear seems to occur when

PV

is higher than 763 MPa*m/s. For all silicon nitrides

tested, the friction transition was also observed when

PV

is higher than 763 MPa*m/s. Tribo-chemical films

formed on the sliding surfaces seem to play an

important role in the observed friction transition

behavior.

INTRODUCTION

Valve train components in heavy-duty diesel engines

operate at high temperatures, high peak cylinder

pressures and severe corrosive environments. These

severe conditions have resulted in various tribo-related

failures in valve train components in many diesel

engines. Therefore, there has been increasing interest

in

developing better valve train materials that are more

wear resistant to enhance

the

reliability and

performance of these components.

Advanced ceramics and emerging intermetallic

materials are promising materials candidates

for

valve

train components since they are highly corrosion and

oxidation resistant, and possess high strength and

hardness at elevated temperatures. In recent years,

advanced ceramics have been widely used in sliding

components such as journal bearings, cylinder liners,

piston rings and mechanical seals

[

1-31. Fine ceramics

have also been successfully used

as

rolling bearings in

machine tools and sliding bearings in water pumps 141.

In the past two decades, silicon nitride-based ceramics

have been targeted for valve train components, and

some commercial successes have been reported in both

automotive and diesel valve trains. They are currently

being used in some diesel engine valves, valve guides,

roller followers and tappet shims [5]. It was reported

[6]

that a significant wear reduction was achieved with

a Si3N4 rocker-arm pad coupled with a cast iron cam.

Silicon nitride ceramics are continually being

developed to improve their physical and mechanical

properties by controlling composition and

microstructure [7,8]. In order to establish a more

complete database for the performance characteristics

of these materials in a sliding contact, the tribological

properties for these ceramics need to be evaluated.

The objectives of this work are to investigate the

tribological properties of some silicon nitrideshteel

sliding contacts under lubricated conditions, and to

identify important properties to improve wear

resistance of these contacts. In addition, the effects of

contact pressure and sliding speed on friction and wear

transitions are also examined.

EXPERIMENTAL SETUP

Geometry

of

Contact

This work mainly focuses on a cam roller/follower

sliding contact in valve train components. To simulate

the contact geometry of this critical contact, the friction

and wear tests were conducted in a block-on-ring test

rig. A flat block specimen (silicon nitride) is loaded

against a ring specimen (4620 steel), which rotates at a

given speed for a given number of revolutions. The

ring specimen is partially submerged in the lubricant.

A

schematic

of

block-on-ring test geometry is given in

Fig.

1.

Materials

Tested

Three commercially available silicon nitrides were

tested in this study. For comparison purpose, limited

tests were also conducted with a hardened 52100 steel

(60 Rc). These materials were tested against 4620 steel

ring specimens, which have the hardness

of

58-62 Rc

and an average surface roughness of 0.23 pm Ra. Some

205

physical and mechanical properties for silicon nitrides tested are given in Table 1.

Materials

SN235

SN237

AS800

Suppliers Second Grain size Strength Toughness Hardness Thermal Young’s

Phase (Pm) (MPa) (MPa~rn”~) (GPa) Cond. Modulus

(W/m*K) (GPa)

Kyocera

Al, Y

Fine

(0.8)

780

k

18

6.3

14.7 27.7 305

Kyocera

Al, Y

Finer

(0.5)

835

k

56

5.8

14.5

305

AlliedSignal La Coarse (3.2) 685

k

21 7.2 14.2 72.5

300

Load

Block

(ceramic)

Fig.

1

-

Schematic of the contact geometry

(block-on-ring)

Test Procedure

All tests were conducted under submerged lubrication

conditions for the test duration of

30

minutes. The

coefficient of friction was monitored and recorded

continuously throughout the test by a computer-based

data acquisition system. The wear volume on the block

specimens was calculated by measuring the wear scar

width using an optical microscope. The worn surfaces

of both ceramic block and steel ring specimens were

also examined using a surface profilometer, optical

interferometer and SEM. To examine the effects of

contact pressure and sliding speed on the tribological

characteristics of silicon nitrides, various loads and

sliding speeds were used. The test conditions used are

summarized

in

Table

2.

Table 2

-

Testing conditions

Contact Pressures (MPa)

Sliding Speeds

(ds)

I

414,586,7 18,829

I

0.31,0.61,0.92, 1.23

Lubrication Condition

RESULTS

Effect

of

Contact

Pressure

Friction results

The friction data

for

silicon nitridedsteel contacts as a

function of contact pressures are given in Fig. 2. For

each condition, two tests were conducted. The results

show that, for given conditions, the average coefficient

of

friction increases

as

the contact pressure increases

for all silicon nitrides tested. AS800 that has a

relatively coarse microstructure gives slightly lower

friction coefficient at higher loading conditions (7

18

and 829 MPa).

300

400

500

600

700

800

900

Initial

Hatziaa

Conm

Ressure.

P

(MPa)

Fig. 2

-

Friction data for silicon nitrides/steel contacts

as

a

function of contact pressure

The typical coefficient of friction plots obtained at

various contact pressures are given

in

Fig. 3. In most

conditions, the friction coefficient tends to decrease

and then increase slightly with time. It should be noted

that, when a contact pressure of 829 MPa was used,

the

friction transitions were obtained with all silicon

nitrides tested. The initial increase in the friction data

can be due to the local breakdown of the lubricant, and

the following decrease is attributed to the formation of

tribo-chemical films on the sliding interface.

A

tribofilm formed on the worn surfaces of silicon

nitrides is given in Fig. 4. The SEM micrographs show

that the area of the worn surface, which is covered by

these films, increases as the contact pressure increases.

Based on the EDS analysis, these films are a

combination of silicon and iron oxides. It is known that

a tribofilm formed on the worn surface remarkably

alter the tribological characteristics of ceramic-steel

206

pairs. The study of Fischer and Mullins

[9]

has shown

that a film of chemical resultant, such

as

amorphous

SiOl

or

hydrated

Si02

will be formed on the wearing

surface

of

Si3N4. Consequently, the friction coefficient

is reduced to a low value. It is believed that the

tribofilms formed on the silicon nitride tested similarly

affect the friction behavior of this material.

0.30

0.25

C

.-

G

0.20

,k

L

2

0.15

5

8

.-

g

0.10

u

0.05

-P

=

414

MPn

P

=

586

MPa

P

=

718

MPa

0.00t"

'1"'

""/"'"I"

"'"'1

0

5

10

15 20

25

30

Time,

t

(min)

0.25

3

0.20

E

.-

B

L

0.15

tl

.-

g

0.10

0

P

=

586

MPa

P

=

718

MPs

P

=

829

MPa

-

0.05

L

........

~I

~~

....____,..

~.

....I

...........

i

...........

:

.........

0.00

b'

"

'

I

'

"

1'

'

"

1

'

"

'

I

1'

"

I'

I"

0

5

10

15

20

25

30

Time.

t

(min)

E

0.30

0.2.5

.8

0.20

c

._

-

i

;

0.1.5

8

0

u

.-

'=

0.10

0.0s

000

1

'

' '

'

1

'

*

' '

I

'

' '

'

I

'

0

5

10

IS

20

25

30

Time,

t

(min)

Fig. 3

-

Typical coefficient of friction plots for silicon

nitrideslsteel contacts obtained at various

contact pressures

Fig. 4

-

Tribo-chemical films formed on the worn

surfaces of silicon nitrides

(a)

P

=

414 MPa, (b)

P

=

829

MPa

Wear results

The wear data for silicon nitrideskteel contacts as a

function of contact pressures are given

in

Fig.

5.

As

previously indicated, the volume worn on a block

specimen was calculated based on the wear scar width

measured using an optical microscope. As expected,

wear increases as the contact pressure increases for all

silicon nitrides tested. For the exception

of

414 MPa,

the lowest wear was obtained with AS800. This is

attributed to its higher thermal conductivity as shown

in

Table

1.

Due to its higher thermal conductivity, the

surface temperature of

AS800

can be lower, resulting

in better wear resistance. The thermal conductivity for

SN237

is

not available, however,

it

is believed that

it

is

not significantly different from that of SN235.

. .

-2'.

. . .

.

:

. . .

.

300

400

so0

600

700

8M)

900

Initial

Henzian

Contact Pressure,

P

(MPa)

Fig.

5

-Wear data for silicon nitrides as a function of

contact pressure

Lubricant: SHF-41 (PAO)

The worn surfaces

of

both ceramic blocks and the

counterface steel rings were examined using a surface

profilometer, an optical interferometer and

SEM.

Surface roughness values of ceramic blocks before and

after testing are given

in

Fig.

6.

207

300

400

500

600

700

800

900

Initial

Henzian

Contact

Pressure.

P

(MPa)

Fig.

6

-

Change of surface roughness of silicon nitrides

as a function of contact pressure

The initial surface roughness

of

silicon nitrides was

about the same. It was

in

the range of

0.26-0.35

pm Ra.

However, except for

SN235

and

AS800

at

829

MPa,

surface roughness after testing decreased one order of

magnitude, indicating mild polishing wear. This is

mainly caused by plastic deformation and

microfractures of the asperities. However, surface

roughness sharply increases for

AS800

and

SN235

when a contact pressure of

829

MPa was used. It was

found that microgrooves were locally formed on the

worn surface as shown

in

Fig.

7.

This indicates that a

wear transition from polishing

to

abrasive wear seems

to occur at a given condition.

~w.3~’

ma*

Q*

24

22

2

1B

94

14

12

1

08

Ob

04

02

0

Pocehirng

.

I

.

Abrasive

Wear

00

02

04

06

08

10

12

Scm

Lcngth.

X

(mm)

Fig.

7

-

3D

surface map and surface profile of the worn

surface of silicon nitride

P

=

829

MPa,

V

=

0.92

ds

Surface roughness values

of

the counterface steel rings

are also given in Fig.

8.

Similar to silicon nitrides,

surface roughness of the steel ring decreased when the

contact pressure is lower than

718

MPa. The worn

surface shows that the initial machining marks are

removed during the rubbing process, and the tribo-

films are formed on the steel ring surface. However,

except for

SN237,

surface roughness values of the steel

counterface sharply increase when the contact pressure

of

829

MPa was used. Several local microgrooves were

also seen on the worn surface of the counterface steel

ring specimens at this condition. This indicates that a

transition in the wear mode is also taking place on the

counterface steel ring.

300

400

500

600

700 800

900

lnitial

Henzian

Contact

Pressure.

P

(MPa)

Fig.

8

-

Change of surface roughness of the

counterface steel ring as a function of contact pressure

Effect

of

Sliding

Speed

Friction results

The friction data as a function of sliding speeds are

given in Fig.

9.

As

previously indicated, the limited

friction and wear data were also obtained with a

hardened

52100

steel

(60

Rc) for comparison purposes.

For silicon nitrides, the coefficient of friction seems to

remain the same as sliding speed increases. The data

show that, for the exception of

0.31

ds,

AS800

gives

slightly lower. friction coefficient among silicon

nitrides tested.

For

52100

steel, the coefficient of

friction at

0.31

ds

is lower than those obtained with

silicon nitrides. However, the friction and wear data for

this material could not be obtained with higher sliding

speeds since

it

scuffed immediately when the sliding

speed of

0.62

m/s

was used. The coefficient of friction

plots as a function of time at various sliding speeds are

given in Fig.

10.

Similar to the friction behavior shown

in

Fig.

3,

the friction transitions were obtained during

the test when a sliding speed of

1.23

ds

was used. The

friction increase during the test at

1.23

ds

is also

attributed to the local breakdown of lubricant films at

the sliding interface.

208

.................................................................................

t

0.00

"'~'""''"~'"'~'"~

0.2

0.4

0.6

0.8

I

.o

I

.2

I

.4

Sliding

Speed.

V

(ds)

Fig. 9

-

Friction data for silicon nitrides and a

hardened 52100 steel as a function of sliding speed

Lubricant: SHF-41 (PAO)

V

=

0

91

m/s

V

=

0.31

m/s

.-

V

=

0.61

m/s

0.25

............,

,-

.............

........

0.30

,

,I,

I,,

,

I

2

SN23

......

I-q

....................................

......................

i.....

...............

I

.....................

~~"~~"~~""~~~'""''''~

0

5

10 I5

20 25

30

V

Time,

t

(min)

0.25

.....

......

Time,

t

(min)

-V

=0.31

mls

-V

=0.61

mls

-$

V

=0.92

m/s

0

0s

000

0

5

10

15

20 2s

30

Time,

t

(min)

Fig. 10

-

Typical coefficient of friction plots for silicon

nitrides obtained at various sliding speeds

Wear

results

The wear data for silicon nitrides and a hardened

52100 steel

as

a function of sliding speed are given in

Fig. 11. For the range of sliding speeds used, wear

increases as the sliding speed increases. It is also seen

that the lowest wear was obtained with AS800 for all

conditions used. As previously noted, this can be due

to its higher thermal conductivity compared to other

silicon nitrides tested. However, the ranking of wear

between SN235 and SN237 is inconclusive. At the

sliding speeds of 0.61 and 1.23

ds,

SN237 is clearly

better than SN235. However, SN235 is slightly better

than SN237 at the sliding speeds of 0.305 and 0.92

ds.

Note that, at a sliding speed of 0.31

ds,

30-40%

more wear was obtained with a hardened 52100 steel

compared to silicon nitrides tested even though its

friction coefficient is lower than those of silicon

nitrides as shown

in

Fig. 9. As previously indicated,

scuffing occurred immediately with a hardened 52100

steel when a sliding speed of 0.61

mls

was used.

Silicon nitrides, however, exhibited mild wear even

with much higher sliding speeds, indicating superior

scuffing resistance of these materials compared to a

hardened 52100 steel.

-

8

1

P=;1175MPq

,'

//

I

-I

3

000

'"IL

~~'~~~~~~'~~~~~~~~~'

'

00

02

04

06

08

10

12

14

Sliding Speed,

V

(ds)

Fig.

1

-

Wear data for silicon nitrides and 52100 steel

as a function of sliding speed

Lubricant: SHF-41 (PAO)

Surface roughness data for silicon nitrides before and

after testing, obtained at various sliding speeds, are

given in Fig. 12. Data show that surface roughness of

the worn surface decreases as the sliding speed

increases up to 0.92

ds.

Mild polishing wear was

obtained for all silicon nitrides at those conditions.

However, surface roughness of the worn surface

sharply increases as the sliding speed of 1.23

ds

was

used. Again,

a

wear transition

from

polishing

to

abrasive wear was observed at this sliding speed.

Surface roughness values of the counterface steel rings

before and after testing are also given

in

Fig. 13. Note

that surface roughness of the worn surface decreased

if

the sliding speed is less than 0.92

ds.

However,

surface roughness sharply increases when the sliding

speed of 1.23

ds

is used. Again, this is the indication

of a wear transition of steel ring at this condition.

209

z

v)

Fig. 12

-

Change of surface roughness of silicon

nitrides

as

a function of sliding speed

0.2

0.4

0.6

0.8

1

.o

1.2

1.4

Sliding

Speed,

V

(ds)

Fig.

13

-

Change of surface roughness of the

counterface steel ring as a function of sliding speed

CONCLUSIONS

Some silicon nitride ceramics for valve train

components were tribologically evaluated. The silicon

nitrides tested are SN235, SN237 (Kyocera Inc., Japan)

and AS800 (AlliedSignal Inc., USA).

For

comparison

purposes, limited tests were also conducted with a

hardened 52100 steel. These materials were tested

against

SAE

4620 steel

in

a block-on-ring tester under

lubricated conditions. The effects of contact pressure

and sliding speed on friction and wear characteristics

were examined. The results of this study are

summarized

as

follows:

(1)

For

a given sliding speed and the range of the

contact pressures used, the average coefficient of

friction increases as the contact pressure increases.

However, it remains about the same

as

the sliding

speed increases with a given contact pressure.

(2) AS800 ceramic gives slightly lower friction

coefficient at higher load and speed conditions.

(3) The friction transition was observed for all silicon

nitrides tested when

PV

is higher than 763

MPa*m/s.

(4)

For ail silicon nitrides tested, wear increases as the

contact pressure

or

sliding speed increases.

(5) For given conditions, AS800 generally gives better

wear resistance compared to other silicon nitrides

tested. It is attributed to its higher thermal

conductivity.

(6) In most conditions, the dominant wear mode for

(7)

silicon nitrides is mild polishing wear, which is

mainly caused by plastic deformation and

microfractures of the asperities. However, a

transition in wear mode

from

polishing to abrasive

wear seems

to

occur when

PV

is higher than 763

MPa*m/s. This transition might be related to the

friction transition observed at this condition.

Silicon nitride ceramics show a superior wear

resistance compared to a hardened 52100 steel (60

Rc) when they are tested against 4620 steel.

ACKNOWLEDGMENT

This research was sponsored by the Department of

Energy in USA, under Contract DE-FC05-970R22579.

REFERENCES

S.

Ying, Advanced Ceramics and

Applications of Ceramics",

Beijing Press

of

Science and Technology,

Beijing, (1990), 167-

170.

S.

Asanabe, Applications of Ceramics for

Tribological Components,

Tribological

International,

20 (6),

(I

987). 355-364.

M. Woydt and J. Schwenzien, Dry and Water-

Lubricated Sliprolling of Si3N4- and SiC-

Based Ceramics,

Tribological International,

G.

W. Stachowiak,

G.

B. Stachowiak, and A.

W. Batchelor, Metallic Film Transfer During

Metal-Ceramic Unlubricated Sliding,

Wear,

B. Dumont,

P.

J. Blau, and

G.

M.

Crosbie,

Reciprocating Friction and Wear of Two

Silicon Nitride-Based Ceramics Against Type

316 Stainless Steel,

Wear,

238, (2000), 93-

109.

M. Kano and I. Tanimoto, Wear Resistance

Properties

of

Ceramic Rocker Arm Pads,

Wear,

145, (1991), 153-165.

C-W.

Li and J. Yamanis, Super-Tough Silicon

Nitride with R-Curve Behavior,

Ceramic

Engineering and Science Proceedings,

10

(7-

C-W. Li, D-J. Lee, and S-C. Lui, R-Curve

Behavior and Strength

of

In-Situ Reinforced

with Different

Silicon Nitride

Microstructures,

J.

Am.

Ceram.

SOC.,

75 (71,

T.

E.

Fischer and W. M. Mullins, Chemical

Aspects of Ceramic Tribology,

J.

Phys.

Chem.,

96,

(1

992), 5690-5701.

26

(3),

(1993), 165-173.

132, (1989), 361-381.

8),

(1989), 632-645.

(1

992), 1777- 1785.

210

OPTIMIZATION OF THE BRAZILIAN DISC TEST FOR CERAMIC

MATERIALS

A. Borger",

P.

Supancic and

R.

Danzer

Department

of

Structural and Functional Ceramics, University

of

Leoben

and Materials Center Leoben,

A-8700

Leoben, Austria

ABSTRACT

The Brazilian disc test is widely used for strength

measurements of brittle low strength materials. It is also

considered to be a simple and low cost method for the

determination of the tensile strength of advanced

ceramics. Because of the very high compressive stress

components in the region of the loading point an

undesired failure mode may occur. In order to prevent

this mode a change of the test geometry is proposed and

the stresses and the effective volume of the modified

experimental set-up are evaluated.

INTRODUCTION

The Brazilian Disc Test (BDT) is a method for

testing the tensile strength

of

cylindrical specimen

[

1-41.

A

diametrical loading of a disc specimen via

two

parallel flat platens leads to a bi-axial stress state with

tensile and compressive stress components in the

specimen. Since the tensile stress maximum is in the

centre of the specimen, it is generally assumed that

failure occurs when the applied tensile stress

components exceed the materials strength.

In the region of the loading point very high

compressive stress components exist

[3].

Their

magnitude is several times higher than that of the tensile

stresses in the centre

of

the specimen.

An

undesired

failure mode caused by these compressive stresses is

sometimes observed. This occurs especially when

testing advanced ceramics even if these materials have a

much higher compressive then tensile strength. But

since this test is applied to measure tensile strength this

mode of failure has to be suppressed.

In this work the geometry of the loading device

has been modified by introducing a curved support in

order to reduce the compressive stresses in the area of

the loading contact and to provoke tensile failing from

the centre of the sample. The stress distribution in the

sample and in the platens is studied by a contact-

mechanical Finite Element (FE) analysis.

Also

the

dependence of the maximum tensile stress in the sample

on the curvature of the support is studied

[5].

STRESS DISTRIBUTION IN THE

STANDARD TEST ASSEMBLY

Figure

1

shows the geometry and the loaded area

of the specimen.

As

a result of loading an

inhomogeneous stress distribution exists in the

specimen.

Munz

and Fett

[6]

presented analytical

solutions for the stress distribution in BDT specimens

considered as a 2D object. For a=O (point load) it holds:

F

Fig.

1

Specimen geometry and loading status.

0,.

=--

(2)

(I

+y*)2x*

(3)

]

-

.?

r2

with:

X*

=slR; y'=ylR

,

5

r2

=(q

/R)~;

r2

.2

=(r,/~)'

.

(4)

These equations result in an infinite compressive stress

at

the loading point. The maximum tensile stress occurs

in the centre of the disc. Therefore, using the maximum

tensile stress failure criterion the disc would fail from

their centre along the y-axis where the stress

components are:

(5)

211

Heinrich J.G., Aldinger F. (Eds.) Ceramic Materials and Components for Engines

Подождите немного. Документ загружается.