High Temperature Deformation and Failure in OxideOxide Composites

1
High Temperature Deformation and Failure in
Oxide/Oxide Composites
Howard G. Halverson Materials Response
Group Virginia Tech Bill Curtin Division of
Engineering Brown University May 5, 1999
2
Objectives

• Determine quasi-static and stress-rupture
  properties of an oxide/oxide ceramic composite at
  23C, 950C, and 1050C.
• Evaluate applicability of an existing
  micro-mechanical model to life prediction under
  stress-rupture conditions.

3
Outline

• Material System
• Quasi-Static Testing and Modeling
• Stress Rupture Testing and Modeling
• Conclusions and Future Work

4
Material System

• Nextel 610 Fiber
• 99 Al2O3
• Alumina-Yttria Matrix
• 20 porosity
• Fugitive Carbon Interface
• 80 - 100 nm thick
• Produced via sol-gel infiltration by
  McDermott Technologies, Inc. (Lynchburg, VA)
• Unidirectional and 8-harness satin weave
  specimens

5
Material System

• Unidirectional specimens
• 50 fiber volume fraction
• Cracks with a spacing of approximately 30µ in
  untested materials
• Woven specimens
• 13.5 axial fiber volume fraction
• Cracks in 90 tow regions

6
Unidirectional Stress-Strain Curves
7
Unidirectional Quasi-Static Properties
A slight (lt12) decrease in properties up to
1050C
8
Tensile Strength Modeling
An accurate relationship between applied stress
and the stress on the intact fibers under global
load sharing with randomly distributed fiber
breaks is
Curtin Zhou (1995)
9
Tensile Strength Modeling
Where sc is the characteristic strength
10
Stress-Rupture Lifetime Prediction
Slow Crack Growth Modeling
Assume that crack growth is the mechanism for
fiber failure at elevated temperatures.
Crack growth is dictated by the Paris Law
So with time the strength of an individual fiber
is
Iyengar Curtin (1997)
11
Slow Crack Growth Modeling
When combined with a Weibull distribution of
individual fiber strengths, we can track the
number of fibers which have failed under a
stress, T, after a time, t, in a gage length, L.
Where si satisfies the fiber strength degradation
equation
12
Slow Crack Growth Modeling
Combine this with the tensile strength model and
we have a system of two equations which relate
damage (r), stress on intact fibers (T), and time
(t).
13
Stress Rupture Equations
Stress Rupture Equations
These two expressions depend on only a few
parameters

• b - slow crack growth exponent
• A,Y, Kic - slow crack growth coefficients
• m- Weibull modulus
• sc - strength property
• f - fiber volume fraction

All of which can be determined before running a
composite stress-rupture test
14
Fiber Rupture Behavior
Obtain the fiber stress rupture parameters from
single fiber testing
Yun DiCarlo, 1993
15
Unidirectional Lifetimes (950C)
16
Unidirectional Lifetimes (1050C)
17
Woven Lifetimes (1093C)
18
Steady-State Strain Rate
Steady-state strain rates are approximately 40
of that expected from the fiber data alone.
19
Conclusions

• The stress-rupture lifetimes of these
  materials are longer by orders of magnitude than
  those predicted by a micromechanical model (which
  ignores any matrix contribution to load-carrying
  capability).
• The steady-state strain rate in these composites
  is lower than that predicted by the fiber creep
  rate alone.
• Nextel 610/alumina-yttria composites maintain
  their room temperature properties with little
  degradation to at least 1050C.

20
Future Work

• Interrupted stress rupture testing of
  materials to determine rate of strength
  degradation
• Verification of stress partitioning between
  fiber and matrix
• SEM to determine if fracture mirrors exist on
  Nextel 610 fibers and, if so, to more accurately
  characterize fiber degradation due to processing.
• SEM to verify that matrix cracks do not form
  during tensile testing