Heinrich J.G., Aldinger F. (Eds.) Ceramic Materials and Components for Engines
Подождите немного. Документ загружается.
think however that the test could be useful for more
types of components on condition that practical
solu-
tions for polishing and indentation can be obtained.
The condition that localized and well-defined
cracks are measured, makes it possible
to
evaluate the
crack growth after a specified number
of
thermal
quench cycles. In this way the indentation-quench test
could be used
to
study cumulative damage that might
arise from thermal cycling. This could then form the
basis of a method
to
estimate the lifetime of compo-
nents based on a specified maximum crack extension
in a critical region
of
a
component.
CONCLUSIONS
The indentation-quench technique has been studied
and the results confirm that the method is appropriate
for measurements of thermal shock resistance. Three
different materials were investigated and the ranking
according to thermal shock resistance was in good
agreement with results reported in the literature.
Measurements were made on
a
simplified compo-
nent in the shape of
a
tapered plate. The results showed
the greatest crack growth in the thickest part
of
the
plate. The crack growth pattern is effectively linked to
the stress pattern and a rough estimation of the thermal
stress pattern on the plate was made.
As a general conclusion the indentation-quench
test
has a variety of uses. It can be applied
to all
types
of
brittle materials in which it is possible to insert inden-
tation precracks. It is possible to study both stable and
unstable crack growth and
to
investigate the crack path
afterwards. Further, the test can normally be applied
directly to components and good statistics can be ob-
tained by making many indents in one single compo-
nent. The presence of residual stress makes the
test
sensitive compared
to
other methods and it is possible
to get information from materials with very good
thermal shock resistance. By combining measurements
of crack growth and calculations it is possible to
roughly estimate the thermal stress pattern in compo-
nents. Another potential application for the test is to
study the cumulative damage that might arise from
thermal cycling. In these ways the
test
can be
a
valu-
able tool for the component engineer.
Acknowledgement-This work has been performed
within the Center Inorganic Interfacial Engineering,
supported by the Swedish National Board for Industrial
and Technical Development (NUTEK) and the
following industrial partners: Erasteel Kloster AB,
Ericsson Cables AB, Hoganas AB, Kanthal AB,
OFCON
AB, Sandvik AB, Seco Tools AB and
Uniroc AB.
D.P.H. Hasselman, Thermal Stress Resistance Para-
meters for Brittle Refractory Ceramics: A Compen-
dium. Am. Ceram. SOC. Bull.,
49, (1970) 1033-1037.
D.P.H. Hasselman, Figures-of-Merit for the Thermal
Stress Resistance of High-Temperature Brittle Materi-
als:
A review, Ceramurgia Int.,
4, (1978) 147-150.
D.P.H. Hasselman, Strength Behavior of Polycrystal-
line Alumina Subjected to Thermal Shock,
J.
Am.
Ceram. SOC.,
53,
(1970) 490-495.
D.
Lewis, Fracture Mechanics of Ceramics
6,
ed. Bradt
and Evans. Plenum Press, New York,
(1983) 487-496.
H.Wang and R.N. Singh, Thermal Shock Behaviour
of
Ceramics and Ceramic Composites, Inter. Mater. Rev.,
W.J. Lee and E.D.Case, Thermal Fatigue
in
Polycrys-
talline Alumina,
J.
Mater. Sci.,
25, (1990) 5043-5054.
T. Anderson and D. J. Rowcliffe, Indentation Thermal
Shock Test for Ceramics, J. Am. Ceram. SOC.,
79,
(1996) 150S15 14.
T. Anderson and D. J. Rowcliffe, Thermal Cycling of
Indented Ceramic Materials, J. Eur. Ceram. SOC., 18,
M. Collin and D. Rowcliffe, Analysis
of
the
Indentation-Quench Test for Ceramics, Ceram. Eng.
Sci. Proc.,
20, (1999) 301-308.
M. Collin and D. Rowcliffe, Analysis and Prediction of
Thermal Shock in Brittle Materials, Acta Mater.,
48,
R. Munro, Evaluated Material Properties for a Sintered
a-Alumina,
80,
(1997) 1919-1928.
D. Richerson, Modern Ceramic Engineering, Marcel
Dekker Inc., New York,
(1992) 166-169.
Y.S. Touloukian, Thermal expansion
-
nonmetallic
solids. Thermophysical Properties of Matter, Plenum,
New York,
(1977) 1140.
F.Kreith and M.Bohn, Principles of Heat Transfer,
Harper
&
Row, New York,
(1986) 512-517.
P.Becher, D.Lewis, K.Carman and A.Gonzales, Ther-
mal Shock Resistance of Ceramics: Size and Geometry
Effects in Quench Tests, Am. Ceram. SOC. Bull.,
59,
K.Matsushita, S.Kuratani, T.Okamoto and M.Shimada,
Young’s Modulus and Internal Friction in Alumina
Subjected to Thermal Shock,
J.
Mater. Sci. Lett.,
3,
M. Collin and D. Rowcliffe, to
be
submitted.
T.Fett, D.Munz and J.Neumann, Technical Note. Local
Stress Intensity Factors for Surface Cracks
in
Plates un-
der Power-shaped Stress Distributions, Engng. Fracture
Mech.,
36, (1990) 647-65
1.
Y. Kim, W.-J. Lee and E.D. Case, The measurement
of
the surface heat transfer coefficient for ceramics
quenched into a water bath, Mat. Sci. Eng., A145,
N. Ramachandran and D. K. Shetty, Rising Crack-
Growth-Resistance (R-Curve) Behavior of Toughened
Alumina and Silicon Nitride, J. Am. Ceram. SOC.,
74
(1991) 26362641.
T. Ohji, K.Hirao and S.Kanzaki, Fracture Resistance
Behavior
of
Highly Anisotropic Silicon Nitride,
J.
Am.
Ceram. SOC.,
78 (1995) 3125-3128.
39, (1994) 228-244.
(1998) 2065-207 1.
(2000) 1655-1 665.
(1980) 542-545,548.
(1984) 345-348.
(1991) L7-Lll.
132
CERAMIC COMPONENTS FOR METAL FORMING TOOLS
Eckart Doege, Lutz Barnert, Torsten Hallfeldt, Steffen Kulp, Tobias Neumaier
Institute for Metal Forming and Metal Forming Machine Tools,
D-30167 Hannover, Germany
ABSTRACT
In metal forming processes tribological conditions
are the result of many factors. The tool material for
example is
of
major importance. Improvement of
existing ceramics has led to various applications in
production engineering. However, until now the use of
ceramics for hot massive or sheet metal forming tools
has not been broadly investigated.
In hot forging conventional methods to increase the
tool life of forging dies made from hot-work steel can
basically not avoid wear. Ceramics, however, are
expected to show better tribological characteristics like
greater resistance against wear. Investigations in flat
crush tests proved the basic suitability of ceramics as
material for forging tools. Further investigations with a
contoured tool indicated that silicon nitride ceramic
does not show any visible wear after
SO00
forging
cycles.
In sheet metal forming processes ceramics as tool
material are expected to improve the tribological
system. The process
is
influenced by various effects
which determine the quality of the drawing part.
Substantial influence on the forming process has the
drawing tool with its geometry, the accuracy of the
drawing tool, the tool material and their interaction in
the tribological system “tool material
-
lubricant
-
sheet
metal”. This interaction influences the beginning of tool
surface defects due to adhesion or abrasion. These
effects lead to aesthetic defects on surface of the
drawing part and fracture. An example of the deep
drawing process, using a ceramic die, is shown in
contrary to the often used hard metal die.
CERAMICS IN FORGING TOOLS
Tool wear is mostly the reason for the failure of hot
forging dies. Therefore, extra costs for finishing
processes of the tool arise in the following period. In
order to increase the lifetime of the tools it is necessary
to find new or improved materials for hot forging.
Because of this, ceramics have to be considered because
they are superior in terms of wear and thermal stability.
At the Institute for Metal Forming and Metal Forming
Machine Tools at the University of Hanover
(IFUM)
investigations are dealing with the application of
ceramics in forging dies.
The main loads on forging tools that cause wear on
the tool are thermal and mechanical loads especially
when combined. During each forging cycle the tool
surface is exposed to extreme thermal stresses. While
forging temperatures between 600°C
-
800°C occur on
the surface of the die. Therefore, cooling of the tools by
spraying water-based lubricants is mandatory in order to
reduce high thermal loads.
[
11.
In combination with high mechanical loads, the
forging tool made from hot-work steel shows abrasive
wear and deformation. The movement is initiated by
thermally caused tempering effects, which soften the
surface
of
the dies.
In order to achieve a notable progress in the tool life
of hot forging dies, the use of ceramics is preferable.
Ceramic materials are well-tried in various technical
areas, especially concerning critical wear. An example
is the use as valves in internal combustion motors [2].
They resist the high mechanical and thermal loads in the
combustion chamber. Because of the comparability of
the thermal shocks of valves and forging dies it seems
reasonable to test ceramics also as material for dies in
hot forging processes.
In 1990 Ohuchi [3] tested the application of ceramic
materials in hot massive forging. The wear
of
different
ceramic materials was tested on segmented tools during
isotherm forging. The material of the workpiece was the
titanium alloy Ti-6
Al-4
V.
In conventional drop forging
of steel materials the occurring loads are different from
those of isotherm forging,
so
the results can not be used
without further consideration.
TEST EQUIPMENT
The forging tests at IFUM are carried out on a
3.05
MN
eccentric forging press which is equipped with
an automated handling system for the parts. An
integrated heating device assures a constant tool
temperature of 200°C. The heating of the workpieces up
to forging temperature of 1100°C is performed in an
induction pusher type furnace. The frequency of the
forging cycle is 13 seconds.
Cold sheared carbon steel slugs
are
used for the
tests. In the cylinder upsetting tests the diameter was 20
mm and the original height of the ingot was ho
=
30 mm
(Fig.
1).
I1
I
1
Fig.
1
Test geometry:
left
upsetting test,
right
cone die
133
These are formed to the final height of hl
=
6,8 mm,
so
that in the end the strain was
cp
=
1,48 lengthwise. In
flash forging tests specimens with an initial height of
h,,=40mm and a diameter of 30mm are used. The
deformation ratio
Q
,,,
is roughly
25
s-'.
CERAMICS IN UPSETTING TEST
In order to get a good idea of the applicability of
ceramic inserts as material for dies cylinder upsetting
tests
are
preferable. In this process the mechanical and
thermal loads of cylinder upsetting are similar to those
loads of the forging process. A simple geometry for the
test specimen can also be used. The ceramic body is
fixed in the die by thermal shrinking. If a material fails
in this trial it will fail even more when
used
in
complicated dies.
The A1203-ceramic is already destroyed after 150
forging cycles probably due to thermal stress (Fig.
2,
left).
A
further investigation of oxide-ceramics does not
appear reasonable because of their physical properties
and the stress within the forging process.
Fig.
4
Fracture of silicon carbide after
14
forging cycles and
no wear even after
5000
forging cycles for silicon nitride
Fig.
2
Fracture of the Al~O~-ceramic after
150
forging cycles
and no wear after
lo00
forging cycles for silicon nitride
Contrary to the Al203-ceramic hot pressed silicon
nitride shows no visible wear after
lo00
strokes. In
order to raise the stress on the ceramic specimen other
tests were carried out with stress-loads pointing away
from the centre of the inlay. The great flow of material
causes extra strong mechanical abrasive loads on the
ceramic. Here also, no visible wear was detected on the
ceramic part, but the surrounding hot working steel
showed notable scarring (Fig.
2,
right).
The silicon nitride ceramic shows an extraordinary
resistance in the forging process. In order to find out
more about the possible way of applicating this material
a cylindric structure was produced in the ceramic. This
structure stands for a deformation-specific additional
element in forging processes. The mechanical load in
forging processes leads to tensile stress in the slot
caused by the pressure inside the structure.
mechanical load, the forging temperature was reduced
from 1100°C to 800°C. For this lower temperature the
ceramic broke after 400 cycles (Fig. 3).
FORGING DIE WITH
nASH
In
further tests hot pressed silicon nitride was
investigated along with variations of silicon carbide. In
these forging tests the three dies with silicon carbide
inlays were destroyed after 14, 40 and 160 cycles
(Fig.
4,
left). Probably a growth of cracks appeared or a
bending load acted on the ceramic. This bending load
may result from the inaccurate positioning of the work
piece into the die. This aspect will be further looked into
in relation with Finite Element Analysis investigation.
The silicon ceramic which had already shown good
results during the cylinder upsetting test also shows
positive results when forging with flash. After
5000
forging cycles no wear was visible, figure 4 right. Such
results can not
be
reached with conventional hot
working steel.
DETERMINATION OF
LOADS
USING FINITE
ELEMENT ANALYSIS
Finite Element Analysis is a good instrument to
discover loads and stress during the forging process at
ambient and working temperature. The rotationally
symmetrical ceramics
of this
investigation were bound
together by thermal shrinking with the parent tool made
from hot work steel. Considering the tensile stress
sensitivity of ceramic materials a sufficient pre-
compressive strain of the ceramic inlays is necessary.
This pre-compressive strain must not be relieved
otherwise the ceramic would fall out of the composite.
The silicon carbide ceramic was not destroyed in
the very first cycle. Therefore, the main cause for
fracture is not only a non-critical breakage but also a
transverse load acting on the inlay. When the work
piece is not placed exactly in the centre of the die an
unsymmetrical load on the upper die occurs. This results
in transverse strains which
are
presented in Fig.
5
for
silicon carbide.
Starting with the hypothesis of the transverse strain
for the determination of
the
reference stress by Mises
(0,
=
2
T-),
a transverse strain of
2-
=
170MPa
occurs after a deformation of only 8.5
mm.
Exceeding
the flectional resistance this causes a breakage
[4].
Fig.
3
Si3N4-ceramic inlay
with
depression, fracture after
400
forging cycles (forging temperature
800
"C)
Even after
lo00
forging cycles no visible wear or
fracture could be detected. In order to increase the
134
fracture of
sillcon
carbide
Fig.
5
Calculated transverse strain
in
forging cycle
for
silicon
carbide with decentralised load
With the help of Finite Element Analysis the reason
for the failure of silicon carbide and for the non-
breakmg of silicon carbide during the
5000
forging
cycles was found.
ANALYSIS OF
TOOL
FAILURE
An industrial partner tested a tool for hot extrusion
with a segment of silicon nitride (Fig.
6).
That ceramic
ring was applied in an area, where the work piece has to
maintain high precision.
armouring
ceramic
2ndarmouring
(steps
"38
(SN)
b
Fig.
6
Design of a hot extrusion tool with a silicon nitride
ceramic ring
The silicon nitride ring failed after
800
strokes
(Fig.
7).
The direction of the fractures indicates an
overload.
Fig.
7
Ceramic ring and fracture
To analyse the relevant loads and the interaction of
all components of the forging cycle the finite element
method was
used.
Relevant data are the maximum tool
temperature of
180°C
and the forging temperature of
1250°C.
Important for the analysis also was the oversize
of the ceramic and steel ring. They were mounted after
heating them to up
400°C
using the lower coefficient of
thermal expansion of ceramic compared to steel. Both
were then pressed into the armouring.
The pressure of about
1300
MPa in the die during
one forging cycle led to a main load of about
950
MPa
in the ceramic ring, which is much higher than the
bending strength of
SN.
A closer look to the contact pressure
of
the system
shows that a tool temperature of
180°C
compensates the
contact between armouring and ceramic ring (Fig.
8).
I
pressure
-160
MPa
)
~ die
(2OOC)
Fig.
8
Contact pressure at room
temperature
and
working
temperature
(1
80°C)
Variations to improve the armouring were carried
out. It was found that thermal shrinking is not suited to
assemble both components. The maximum temperature
of steel for thermal shrinking to avoid tempering effects
is
550°C.
The resulting oversize and armouring is not
sufficient to give an initial stress.
CERAMICS IN SHEET METAL
FORMING
Deep drawing is a technology often used in sheet
metal forming. In deep drawing processes parts are
produced for different industrial areas mostly with a
number
of
parts
amounting to one million, sometimes
even more. During a deep drawing process the sheet
material is normally formed with a tool system
consisting of a punch, a blank holder and a drawing die.
Especially the tribological system between sheet
metal and tool system has an important influence on the
deep drawing process, the quality of the formed
component and tool life
[5].
In figure
9
a conventional
tool system with three friction zones in the drawing die,
the blank holder and the punch is shown.
Fig.
9
Friction zones
in
a
deep
drawing
tool
system
135
In general at deep drawing processes the coefficient
of friction
p
in the punch head zone should
be
high to
allow a good transfer of the punch forces to the work
piece. In the flange area and
the
die radius a material
flow with small forces should
be
realised. Therefore,
low coefficients of friction
p
between sheet metal and
the drawing die are necessary [5]. In
this
context
different kinds of forming lubricants and several
coatings of the sheet metal material have been examined
[6]. Furthermore, besides conventional steel dies, new
tool concepts were developed and tested at
IFUM
[7,8].
For the processing
of
e.g. higher strength steels,
aluminium and other light weight alloys new tool
materials are of demand for the improvement of the
tribological system in deep drawing processes. The
main objectives are to reduce friction and forming
forces as well as the required lubricants. Another
important aspect is to avoid adhesion and pick-ups, that
occur e.g. when uncoated aluminium sheets without
lubricants are drawn with steel tools. Instead
of
these
tool materials, ceramics have favourable qualities
especially increased hardness, good wear resistance and
excellent tribological behaviour
[9.
10, 113,
On
the other
hand, ceramics have a reduced ductility
so
that
spontaneous cracks can appear. Furthermore, the
treatment of ceramic materials is expensive, mostly it
will be done with grinding methods
[
121 sometimes e.g.
for silicon-nitride ceramics with laser cutting
[
131.
At
IFUM
experimental examinations with different
kinds of ceramics and steel materials have been carried
out. To investigate the applicability of axisymmetric
ceramic drawing dies, silicon-nitride, circonium-oxide
and silicon-carbide ceramics have been applied. In
comparison conventional hard metal drawing dies have
been analysed under the same conditions.
Figure
10
shows pick-ups in the area of the die
radius, which affect the drawing results. Surface
failures, geometric deviations and in the worst
case
cracks on the work piece can be found.
Deep drawing dies
of
silicon-carbide ceramics have
been used with the same geometry.
As
shown in
Figure 11 no adhesion or pick-ups can be detected.
Fig. 11 Ceramic drawing die
The reason for this behaviour is the type of binding
of the ceramics, which differs from the metallic bonds
of steel materials. During the deep drawing process
adhesion results in higher punch forces. A comparison
of the maximum punch forces with the drawing die
materials silicon-nitride ceramic and hart metal
GE
50
without lubricants is shown in figure 12. The punch
forces with the silicon-nitride ceramic are up to 23
%
less compared to a hard metal tool system.
I
5"
hard
meta
g3
(GE50)
ceramic
Y
g2
51
(Si3NJ
P
do
lubrication:
E
B0
=
1.06,
SO
=
0.30
mm
material:
St4
LG
Ni
plat.
dry
referem
Ordder
@
IFUM
Fig.
12
Punch forces with different drawing die materials
In examinations with lubricants the punch forces
obtained with ceramics were
6
%
less compared to hart
metal drawing dies.
Another reason for these results are the roughness
of the surface
of
ceramic materials. They
are
lower than
of conventional steel drawing dies (Fig. 13).
Fig.
10
Surface failures
in
the
area
of
a
die radius
136
Fig.
13
Surface roughness
of
the examined materials
The examinations resulted in some advantages for
the usage of ceramic materials. The amount of
lubricants in deep drawing processes can be reduced.
Consistent ceramics are more environmentally friendly.
Due to the wear resistance the tool life is very high,
so
the expense for maintenance is smaller than for
conventional steel tools.
In deep drawing tools, ceramics are applied for
inlays in high loaded areas, e.g. comers
[14].
They are
also well suited in progressive tools or components, e.g.
drawing dies with small geometry. Ceramics also have
advantages for processing aluminium or titanium sheet
materials and other materials which show an adhesive
reaction to steel tools. The inlays should be placed in
areas with a high contact stress and a high relative
velocity between drawing die and sheet material.
Appropriate positions for ceramic inserts in large tool
systems are displayed in figure
14.
Fig.
14
Ceramic
inserts
located
in
high
loaded
areas
of
a
die
SUMMARY
Ceramic components are tested in a wide range of
technical applications. But
so
far, their use in forming
processes has not been broadly investigated. Obviously
ceramics offer an enormous potential in several fields.
In hot massive forming high thermal and mechanical
loads lead to wear of the forging tools made of hot-work
steel. In comparison a ceramic tool inlay of silicon
nitride showed no visible wear after even
5000
forging
cycles. The use of ceramics as die inlays results in a
considerably longer tool life. Only silicon nitride was
convincing in these test as a suited material.
In sheet metal forming ceramics as tool materials
improve the tribological system. Compared to
conventional dies, maximum punch forces are reduced
by
23%
for
an
non lubricated system. In consequence,
higher drawing ratios are possible, which means process
limits are extended. With ceramic dies reduced or no
lubrication is necessary. Ceramics lead to a longer tool
life. In the end maintenance costs are reduced.
In order to gain more experience in the application
of ceramics in forming tools further investigation are
necessary.
References
W. Stute-Schlamme: Konstruktion und thermo-
mechanisches Verhalten rotationssymmetrischer
Schmiedegesenke
198
1.
-
Universitat Hannover,
Fakultat fir Maschinenwesen, Dissertation
G.
Wotting,
G.
Leimer: Siliciumnitrid-Keramik,
deren Eigenschaften und Anwendungen in der
Umformtechnik.- In: Umformtechnik an der
Schwelle zum nachsten Jahrtausend.
16.
Umformtechnisches Kolloquium Hannover
25./26.
Februar
1999
K. Ohuchi,
S.
Sasaki, K. Matsuno: Isothermal
Forging with Ceramic Die on Industrial Basis.-In:
Advanced Technology of Plasticity
1990.
Proceedings of the 3rd Int. Conference on
Technology of Plasticity, July
1/6 1990,
Kyoto,
Japan
E. Doege, T. Neumaier, C. Romanowski: EinfluB
der Rohteilpositionierung auf die Werkzeug-
belastung beim Gesenkschmieden am Beispiel
eines KeramikMetall Verbundes. In: Marc
Benutzertreffen,
15.-16.
Oktober
1997,
Miinchen
E.
Doege, Wichtige EinfluBgroBen beim
Tiefziehen.
wt-Z.
fiir industrielle Fertigung
(66),
Heft
11,
S.
615-619,1976
W.
G.
Brazier,
R.
W. Thompson, The effect of
zinc coatings, die materials and forming
lubricants.
SAE
Technical Paper
860432, 1986
C. Frank, K. Droder, Reduced friction in deep
drawing by means of organic sheet coatings.
Birmingham, IDDRG Working Group Meeting,
E. Doege, C. Frank, Polymers as tool materials in
sheet metal forming.
2nd
Int. Assiut Conf. on
mechanical engineering, Advanced technologies
for production, Assiut, Egypt.,
02.-04.03.1999
W. D. Kingery, Introduction to ceramics. John
Wiley
&
Sons, New York,
1976
D. H. Buckley, K. Mikyoshi, Fundamental
07.-09.06.1999
137
tribological properties of ceramics.
Roc.
9"
Int.
Conf., Composites and advanced ceramic
materials, Amerc. Ceramic Society,
20.-
24.01.1985
(1 1) E. A. Almond,
L.
A. Lay, M. G. Gee, Comparison
of sliding and abrasive wear mechanisms in
ceramics and cemented carbides. 2"6 Int. Conf.
Science Hard Materials, Rhodes, 1986
(12) B. G. Koepke, R.
J.
Stokes, Effect of workpiece
properties on grinding forces in polycrystalline
ceramics. Proc. Symp. National bureau of
standards, Gaithersburg, Maryland, 13.-
15.1 1.1987
(13) G. Warnecke, Schleifen von Hochleistungs-
keramik-Werkstoff, Anwendung, Bearbeitung,
Qualittit. Verlag
TW
Rheinland, Koln, 1994
(14)
J.
Muller, R. Heinze, Einsatz von Keramik in
Werkzeugen fur die Blechumformung. Int. Konf.
"Neuere Entwicklungen in der Blechumformung"
Fellbach, 23 .-24.05.2OOO
138
EFFECT OF
GRAIN
BOUNDARY COMPOSITION ON HIGH-
TEMPERATURE MECHANICAL PROPERTIES OF HOT-PRESSED
SILICON CARBIDE SINTERED WITH YTTRIA
Donna Hermanub* and Hagen Klemm
Fraunhofer-Institut Keramische Technologien und Sintenverkstoffe, Winterbergstralje 28,
0
1277
Dresden, Germany
ABSTRACT
In materials with microscopic grain sizes, the grain
boundary composition plays a decisive role in
determining the resulting high-temperature oxidation,
creep and slow-crack-growth behaviour. Therefore, it is
crucial that the relationship between composition and
properties is
known
in order to tailor the material to
meet specific requirements.
In the present study, the chemical composition in the
system silicon carbide with Y2O3
as
a sintering aid was
varied. The resultant powders were hot pressed and
tested in four-point bending at room temperature
as
well
as
at 1500 "C. The room-temperature strength and the
fracture toughness (ICL method) were determined. The
dynamic fatigue strength was measured at a stress rate
of 0.05 MPds and an initial stress of 50 MPa. Hot-
pressed and tempered (100 h in air at 1500 "C) bending
bars with either natural flaws or a single edge notch
were tested. The fracture surfaces were examined and
an attempt was made to correlate the observations with
results of the mechanical tests. Differences were found
between the Y203-rich and the Si02-rich materials,
independent of the flaw type. At 1500 "C, the Si02-rich
material exhibited slow crack growth, whereas the
Y203-rich material exhibited typical fracture features
such
as
a
mirror,
mist and hackle. The Y203-rich
specimens failed at higher stresses than the Si02-rich
specimens. Additionally, for notched specimens, the
dynamic fatigue strengths were higher at 1500 "C than
at room temperature and at both temperatures after
tempering. It is assumed that processes involving
blunting of the crack tip are responsible for the observed
behaviour.
INTRODUCTION
Among silicon-based nonoxide ceramics, liquid-phase-
sintered silicon carbide (LPS-Sic) is a potential material
from which combustion chamber linings and other
stationary structural components in high-temperature
gas turbines can be made. For these applications, the
material must have low weight, long-term stability, high
strength and good fracture toughness at room and high
temperatures, no time-dependent strength degradation
and minimal creep
[I].
Silicon carbide, being a mainly covalently bonded
material, is very dificult to sinter. Additives such as
A1203
(or
Al) and combinations of these with Y2O3 and
such additives as
SO2,
CaC03 and MgO have been used
as
sintering aids [2-111, although the rare-earth oxide
Dy2O3 has also been used [12].
An
intergranular
amorphous phase results from liquid-phase sintering or
impurity segregation; this phase acts as a liquid phase
and adopts an equilibrium thickness at high
temperatures
[
131.
Because oxygen, impurities such
as
Ca" and other cations such as rare-earth cations are
essentially insoluble in Sic, they segregate into the
grain boundaries to form a residual amorphous
intergranular phase [14]. It has been proposed that the
thickness of this film is related to the composition and
not
the
amount of the glass phase present
[
151. Thus it is
assumed that the chemical composition of the grain
boundaries plays a decisive role in determining the
mechanical behaviour at high temperatures.
In
this
study, the effect of altering the grain boundary
chemistry on the dynamic fatigue behaviour of notched
and unnotched LPS-Sic bending specimens at room
temperature and at 1500 "C was examined. Up to now,
there have been relatively few studies published that
relate to the high-temperature mechanical behaviour of
LPS- Sic [16-191 or of LPS-SiC/Si3N4 composites
[20-
231.
It is hoped that through this work, a better
understanding
of
the way in which the grain boundary
composition influences the crack-growth behaviour of
LPS-Sic at room and elevated temperatures will be
gained.
EXPERIMENTAL PROCEDURES
The materials were fabricated by mixing Sic powder
with different amounts of Y203 and Si02 in an attritor in
isopropanol, drylng in a rotary vacuum evaporator, and
calcining at 450 "C. The additive amounts were chosen
such that the grain boundary volume remained constant
at
12
%
and the molar ratio of Y2O3 to Si02 was
0.7
or
2.0.
The resulting powders were hot pressed for one
hour at a temperature of 2000 "C in a graphite-lined
chamber in an argon atmosphere. The densities were
measured by the Archimedes method and were found to
match the theoretical values.
139
Bending specimens were produced from the materials
thus obtained. The room-temperature strength and the
fracture toughness (ICL method) were obtained for both
materials. For each of the
two
materials, half of the
remaining specimens were notched with a saw and 1-
pm-sized diamond paste to a depth of ca.
350
pm. Half
of the unnotched and half of the notched specimens
were tempered for
100
h in an oven in air at 1500 "C.
The resulting specimens underwent dynamic fatigue
tests in four-point (inner span:
20
mm;
outer span:
40
mm) bending at room temperature and at 1500 "C with
a loading ramp of
0.05
MPds and an initial stress of
50
MPa.
X-ray diffraction (Cu
Q
was performed on hot-pressed
and tempered specimens after dynamic fatigue testing at
room temperature and at 1500 "C. XRD was
subsequently performed on notched specimens
tempered or tested at 1500 "C after removing an
additional 100 pm.
The fracture surfaces were examined in a light
microscope and in the SEM. EDS was used to determine
the fracture origins. Polished and CF4 plasma-etched
cross sections were also examined in the SEM.
RESULTS
Room-Temperature Strength and Fracture
Toughness
The room-temperature strength and fracture toughness
(ICL method) were
450
MPa and
3.8
MPa.m'"
respectively for the Si02-rich material and
660
MPa and
3
.O
MPa.m"* respectively for the Y203-rich material.
Both materials exhibited mainly intergranular fracture.
Dynamic Fatigue Testing
of
Specimens with
Natural Flaws
The results of the dynamic fatigue tests for specimens
with natural flaws are shown in Table 1.
rable 1. Dynamic fatigue strengths
(0.05
MPds
starting
at
50
MPa) for snecimens with natural flaws
Three or four results were averaged to obtain each of the
above values. Figures
1
-
4
show the fracture surfaces
of
hot-pressed and tempered SiO2-rich and Y203-rich
specimens after the dynamic fatigue tests.
Figure 1. Fracture surface of a Si02-rich specimen with
natural
flaws after dynamic fatigue testing at 1500 "C.
The arrow points to the approximate location of the
fracture origin.
Figure
2.
Fracture surface of a tempered Si02-rich
specimen with natural flaws after dynamic fatigue
testing at 1500 "C
Figure
3.
Fracture surface of an Y203-rich specimen
with natural flaws after dynamic fatigue testing at
1500
"C
140
Figure
4.
Fracture surface of a tempered Y~Oj-rich
specimen with natural flaws after dynamic fatigue
testing at 1500
"C
Dynamic Fatigue Testing of Specimens with
an
Edge Notch
The results of the dynamic fatigue tests for notched
specimens are shown in Table
2;
the fracture surfaces
for hot-pressed specimens tested at room temperature
and at 1500
"C
are shown in Figures
5
-
8.
Table
2.
Dynamic fatigue strengths
(0.05
MPds starting
at
50
MPa) for specimens with an edge notch
of
ca.
350
material
Si02-rich
pm depth
fracture stress at
room temperature
[MPaI
Hot-
fracture stress at
1500
"C
[MPa]
265 265
Figure
6.
Fracture surface of a SiO2-rich notched
specimen after dynamic fatigue testing at 1500
OC
Figure
7.
Fracture surface of an Y203-rich notched
specimen after dynamic fatigue testing at room
temperature
Three or four results were averaged to obtain each of the
above values.
Figure
8.
Fracture surface of
an
Y203-rich notched
specimen after dynamic fatigue testing at 1500
"C
Figure
5.
Fracture surface of a SiOz-rich notched
specimen after dynamic fatigue testing at room
temperature
The X-ray diffraction results are shown in Table
3
for
notched specimens tested at room temperature and
Table
4
for notched specimens tested at 1500
"C.
141