T vVb - 6 (
A Paper Entitled
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
Submitted for presentation in the
International Joint Tribology Conference
Sponsored by ASME International and STLE
October 1-4, 2000
Seattle, Washington
And for publication in the ASME Journal of Tribology
By
Nagaraj K. Arakere
Associate Professor
Mechanical Engineering Department
University of Florida
Gainesville, FL 3261 1-6300
(352) 392-0856: Tel; (352) 392-1071: Fax
and
Gregory Swanson
NASA George C. Marshall Space Flight Center
ED22/Strength Analysis Group
MSFC, Alabama-35812
*■£ * Niagara] K . Arakere and Gregory Swanson
K ' Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
ABSTRACT
Single crystal nickel base superalloy turbine blades are being utilized in rocket engine turbopumps
and turbine engines because of their superior creep, stress rupture, melt resistance and
thermomechanical fatigue capabilities over polycrystalline alloys. Currently the most widely used
single crystal nickel base turbine blade superalloys are PWA 1480/1493 and PWA 1484. These
alloys play an important role in commercial, military and space propulsion systems. High Cycle
Fatigue (HCF) induced failures in aircraft gas turbine and rocket engine turbopump blades is a
pervasive problem. Blade attachment regions are prone to fretting fatigue failures. Single crystal
nickel base superalloy turbine blades are especially prone to fretting damage because the subsurface
shear stresses induced by fretting action at the attachment regions can result in crystallographic
initiation and crack growth along octahedral planes. Furthermore, crystallographic crack growth on
octahedral planes under fretting induced mixed mode loading can be an order of magnitude faster
than under pure mode I loading.
This paper presents contact stress evaluation in the attachment region for single crystal turbine
blades used in the NASA alternate Advanced High Pressure Fuel Turbo Pump (HPFTP/AT) for the
Space Shuttle Main Engine (SSME). Single crystal materials have highly orthotropic properties
making the position of the crystal lattice relative to the part geometry a significant factor in the
overall analysis. Blades and the attachment region are modeled using a large-scale 3D finite element
(FE) model capable of accounting for contact friction, material orthotrophy, and variation in
primary and secondary crystal orientation. Contact stress analysis in the blade attachment regions is
presented as a function of coefficient of friction and primary and secondary crystal orientation.
Stress results are used to discuss fretting fatigue failure analysis of SSME blades. Attachment
stresses are seen to reach peak values at locations where fretting cracks have been observed.
Fretting stresses at the attachment region are seen to vary significantly as a function of crystal
orientation. Attempts to adapt techniques used for estimating fatigue life in the airfoil region, for
life calculations in the attachment region, are presented. An effective model for predicting
crystallographic crack initiation under mixed mode loading is required for life prediction under
fretting action.
Key Words
Fretting, Fatigue, High Cycle Fatigue (HCF), Single crystals. Crystal orientation, Nickel-base
superalloys, Face Centered Cubic (FCC), Turbine blades, Blade attachments
Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
Introduction
Fretting fatigue failure of mechanical components has gradually come to be recognized as a failure
mode of major importance. The presence of fretting in conjunction with a mean stress in the body of
a component can lead to a marked reduction in high cycle fatigue (HCF) life, sometimes by a factor
as great as 10 [1]. Fretting occurs when assemblies of components such as blade and disk
attachment surfaces, bolt flanges, snap fit areas, and other clamped members are subjected to
vibration, resulting in contact damage. The combined effects of corrosion, wear, and fatigue
phenomena at the fretting contact facilitates the initiation and subsequent growth of cracks,
ultimately leading to failure. Single crystal nickel base superalloy turbine blades are especially
prone to fretting damage because the subsurface shear stresses induced by fretting action at the
attachment regions can result in crystallographic initiation and crack growth along octahedral
planes.
Single crystal materials have highly orthotropic properties making the position of the crystal lattice
relative to the part geometry a significant factor in the overall analysis. Study of failure modes of
single crystal turbine blades therefore has to account for material orthotrophy and variations in
crystal orientation. HCF induced failures in aircraft gas-turbine engines are a pervasive problem
affecting a wide range of components and materials. HCF accounts for 24% of component failures
in gas turbine aircraft engines, of which blade failures account for 40% [2], Estimation of blade
fatigue life, therefore, represents an important aspect of durability assessment. The modified
Goodman approach currently used for life assessment does not address important factors that affect
High Cycle Fatigue (HCF) such as fretting and/or galling surface damage, and interaction with Low
Cycle Fatigue (LCF) [2], Designing to resist fretting fatigue has proven to be a difficult task. The
fretting fatigue process is affected by a large number of variables, perhaps by as many as fifty [3],
For polycrystalline materials the critical primary variables are thought to be the coefficient of
friction, the slip amplitude, and the contact pressure at the fretting interface [1]. However, for single
crystal nickel superalloys, studying the impact of fretting requires consideration of additional
variables. The crack initiated by fretting action can be crystallographic or noncrystallographic
depending on the interaction between the point-source defect species (carbides, micropores and
eutectics) with the effects of environment, temperature and stress [4]. The impact of fretting on
relationships between macroscopic fracture details and the parameters that are used to describe
fatigue crack growth is essential. Understanding the effects of fretting damage on HCF life of single
crystal nickel superalloy components is critical to lay a foundation for a mechanistic based life
prediction system.
Turbine blades and vanes, used in aircraft and rocket engines, are probably the most demanding
structural applications for high temperature materials [5] due to the combination of high operating
temperature, corrosive environment, high monotonic and cyclic stresses, long expected component
lifetimes, and the enormous consequence of structural failure. Directionally solidified (DS)
columnar-grained and single crystal superalloys have the highest elevated-temperature capabilities
of any superalloys. As the understanding of alloying and microstructural control (i.e., directional
solidification) has improved, the maximum use temperatures of Ni-base superalloys has increased to
levels in excess of 90% of their melting temperatures (T/T m = 0.9). Directional solidification is used
to produce a single crystal [6] with the <00 1> low modulus orientation parallel to the growth
direction. The secondary direction normal to the growth direction, which is typically referenced to
a line parallel to the blade attachment, is random if a grain selector is used to form the single crystal.
Both the primary and secondary crystal orientations, though, can be selected if seeds are used to
3
Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
generate the single crystal. However, in most cases, grain selectors are used to produce the desired
<001> growth direction. In this case, the secondary orientations of the single crystal components
are determined but not controlled. Initially, control of the secondary orientation was not considered
necessary [7]. However, recent reviews of space shuttle main engine (SSME) turbine blade lifetime
data has indicated that secondary orientation has a significant impact on high cycle fatigue
resistance [8,9,10],
Currently the most widely used single crystal nickel base turbine blade superalloys are PWA 1480,
PWA 1484, RENE’ N-5 and CMSX-4. PWA 1480 will be used in the NASA SSME alternate
turbopump. These alloys play an important role in commercial, military and space propulsion
systems [4]. Military gas turbine mission profiles are characterized by multiple throttle excursions
associated with maneuvers such as climb, intercept and air-to-air combat. This mission shifts
attention somewhat to fatigue and fracture considerations associated with areas below the blade
platform, which contain various stress risers in the form of buttresses and attachments [4], The gas
turbine industry is moving toward a design philosophy that stresses damage tolerance, structural
integrity and threshold based designs, which has prompted the study of micromechanics of fretting
fatigue and other events that affect HCF life.
Numerous analytical and experimental studies have been conducted on fretting fatigue damage in
polycrystalline alloys. Some representative examples are by Hills and Newell [1], Giannokopoulos
and Suresh [11], Swolwinsky and Farris [12], Attia and Waterhouse [13], Hoeppner [14], Vingsbo
and Soderberg [15], and Ruiz, et al [16], Studies on mechanics of fretting fatigue in single crystal
superalloys are almost nonexistent. This paper presents an investigation of fretting stresses in single
crystal superalloy turbine blades in the blade-disk attachment regions. Large-scale finite element
models of are used to predict stresses in the airfoil and blade attachment regions. Several fatigue
damage parameters based on shear and normal stresses acting on various slip planes for FCC
crystals are evaluated. Blade stress response is presented as a function of primary and secondary
crystallographic orientation.
Deformation Mechanisms in FCC Single Crystals
Nickel based single-crystal material such as PWA1480 and PWA1484 are precipitation
strengthened cast single grain superalloys based on the Ni-Cr-Al system. The microstructure
consists of approximately 60% to 70% by volume of y’ precipitates in a y matrix. The y’ precipitate,
based on the intermetallic compound Ni 3 Al, is the strengthening phase in nickel-base superalloys
and is a Face Centered Cubic (FCC) structure. The / precipitate is suspended within the y matrix
also has a FCC structure and is comprised of nickel with cobalt, chromium, tungsten and tantalum
in solution.
Deformation mechanisms in single crystals are primarily dependent on microstructure, orientation,
temperature, and crystal structure. The operation of structures at high temperature places additional
materials constraints on the design that are not required for systems that operate at or near room
temperatures. In general, materials become weaker with increasing temperature due to thermally
activated processes, such as multiple slip and cross-slip. At temperatures in excess of approximately
half the homologous temperature (the ratio of the test temperature to the melting point, = T/T m ),
diffusion controlled processes (e.g., recovery, recrystallization, dislocation climb and grain growth)
become important, which results in further reductions in strength. Slip in metal crystals often occurs
4
Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
on planes of high atomic density in closely packed directions. The 4 octahedral planes
corresponding to the high-density planes in the FCC crystal have 3 primary slip directions (easy-
slip) resulting in 12 independent primary <1 10> {111} slip systems. The 4 octahedral slip planes
also have 3 secondary slip directions resulting in 12 independent secondary <11 2> {111} slip
systems. In addition, the 3 cube slip planes have 2 slip directions resulting in 6 independent <1 10>
{100} cube slip systems. Thus there are 12 primary and 12 secondary slip systems associated with
the 4 octahedral planes and 6 cube slip systems with the 3 cube planes, for a total of 30 slip systems
[10]. At high temperatures, slip has been observed in non-close-pack directions on the octahedral
plane, and on the cube plane, in FCC crystals. Table 1 shows the 30 possible slip systems in a FCC
crystal [17]. Elastic response of FCC crystals is obtained by expressing Flooke’s law for materials
with cubic symmetry.
Shear stresses associated with the 30 slip systems are denoted by T 1 , T 2 ,..., t 30 . The shear stresses on
the 24 octahedral slip systems are given by [17],
V'
" 1
0
-1
1
0
-f
T 13 '
2
-1
1
-2
1 ’
T 2
0
-1
1
-1
1
0
T 14
2
-1
-1
1
1
-2
T 3
1
-1
0
0
1
-1
T 15
-1
-1
2
-2
1
1
T 4
-1
0
1
1
0
-1
°xx
T 16
-1
2
-1
-1
- 2
-1
T 5
-1
1
0
0
-1
-1
a yy
T 17
-1
-1
2
2
1
-1
T 6
1
0
1
-1
-1
”1
0
r 18
1
2
-1
-1
-i
1
2
T 7
1
-1
0
0
-1
-1
a xy
s
T 19
3>/2
-1
-1
2
2
-1
1
T 8
0
1
-1
-1
1
0
G zx
T 20
2
-1
-1
-1
-1
-2
T 9
l
0
-1
-1
0
~1
a yz
T 21
-1
2
-1
-1
2
1
T 10
0
-1
1
-1
-I
0
T 22
2
-1
-1
1
-1
2
T U
-l
0
1
-1
0
-1
to
OJ
-1
2
-1
1
2
-1
T 12
_-i
1
0
0
1
-1
t 2 \
-1
-1
2
-2
-1
-1_
And the shear stresses on the 6 cube slip systems are
T 25 '
"0
0
0
1
1
0“
T 26
0
0
0
1
-1
0
a »
T 27
1
0
0
0
1
0
1
°zz
to
OO
0
0
0
1
0
-1
G xy
T 29
0
0
0
0
1
1
k
_o
0
0
0
-1
1 _
° yz .
( 2 )
Shear strains (engineering) on the 30 slip systems are calculated using similar kinematic relations.
The stresses (a xx , cr yy , , a yz ) must be expressed in the material or the crystal coordinate system,
i.e., the x, y, and z coordinate axes are parallel to the edges of the FCC crystal.
Fretting Fatigue Crack Initiation and Crack Growth in Nickel-Base Superalloys
Crack initiation in Ni-base superalloys fall into two broad classifications, crystallographic and
noncrystallographic. The interaction between the effects of environment, temperature and stress
5
'Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
intensity determines which point-source defect species initiates a crystallographic or
noncrystallographic fatigue crack [4]. At low temperature (<427 C) and stress conditions
crystallographic initiation appears to be the most prevalent mode. At moderately high temperature
(above 593C) initiation at TaC carbides predominates over occasional (111) eutectic origins. The
subsurface shear stresses induced by fretting action can result in crystallographic initiation of
failure, as seen in Fig. 1, which shows a subsurface fatigue crack emanating from a carbide (TaC) in
a turbine blade attachment (PWA1422, a columnar grain Ni-base superalloy) and propagating along
octahedral (111) planes [4], Fretting fatigue at low slip amplitudes that induces little or no surface
damage can result in greatly reduced fatigue life with accelerated subsurface crystallographic crack
initiation, akin to subsurface shear stress induced rolling bearing fatigue. An example of this
phenomenon is seen in Fig. 2, which details the underside view of a single crystal Ni (PWA1480)
turbine blade platform tip fretting failure at the contact where a centrifugal damper impinged the
blade platform [18]. The damper (or fret pin) was subjected to an alternating compressive load
generating subsurface shear stresses. The fretting fatigue crack initiation was clearly subsurface,
propagating along intersecting (111) crystallographic planes until breaking out to the surface. The
piece was subsequently ejected. An identical scenario occurred with the impinging damper. The
damper was quite small and cast in a very coarse grained form such that the contact point was a
single large grain. It too exhibited a subsurface crystallographic shear fatigue crack
initiation/propagation on multiple intersecting octahedral planes leading to a pyramidal "hole" in the
component as the "chip" was ejected.
The fatigue crack growth (FCG) behavior of single crystal nickel superalloys is governed by a
complex interaction between the operative deformation mechanism, stress intensity, and
environmental conditions. The FCG behavior is determined by the operative microscopic fracture
mode. As a result of the two phase microstructure present in single crystal nickel alloys a complex
set of fracture mode exists based on the dislocation motion in the matrix ( 7 ) and precipitate phase
(/). Telesman and Ghosn [19] have observed the transition of fracture mode as a function of stress
intensity (K) in PWA1480 at room temperature. They also noted that, K rss , the resolved shear stress
intensity parameter on the 12 primary slip planes predicted the crack zigzag behavior caused by
fracture mode transitions. Deluca and Cowles [20] have observed the fracture mode transition that is
environmentally dependent (presence of high-pressure hydrogen). The resolved shear stresses and
not the maximum principal stresses govern failure on the (111) octahedral slip planes. However
stresses normal to the plane (mode I component) are thought to play some role in the failure
process. Crystallographic crack growth along (111) planes under mixed mode loading can be an
order of magnitude faster than under pure mode I loading [21]. These results have important
implications on fretting fatigue. Since fretting action results in mixed mode loading this can result
in crystallographic initiation along (111) slip planes (Fig. 1) and rapid crack growth under loads
lower than that expected under mode I loading.
Effect of Crystal Orientation on Fatigue Life
Single crystal materials have highly orthotropic properties making the position of the crystal lattice
relative to the part geometry a significant factor in the overall analysis. For example, the elastic or
Young’s modulus (E) is a strong function of crystal orientation over the standard stereographic
triangle, given by Eq. (3) [22].
E' 1 = Sn - [2(Sn - S, 2 ) - S 44] [cos 2 <t>(sin 2 4 - sin 2 0 cos 2 <J> cos 2 0)] (3)
6
Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
Where 0 is the angle between the growth direction and <001> and <|> is the angle between the <001 >
- <110> boundary of the triangle. The terms Sn, S 12 and S 44 are the elastic compliances. In
general, the elastic properties are insensitive to composition and values for pure Ni are a good
approximation of the values for even the most highly alloyed Ni-base superalloys.
Strain controlled LCF tests conducted at 1200 F in air for PWA1480/1493 uniaxial smooth
specimens, for four different orientations, is shown in Table 2 [9]. The four specimen orientations
are <00 1> (5 data points), < 1 1 1 > (7 data points), <213> (4 data points), and <01 1> (3 data points),
for a total of 19 data points. Figure 3 shows the plot of strain range vs. Cycles to failure. A wide
scatter is observed in the data with poor correlation for a power law fit. Table 3 shows the fatigue
parameters such as AT max , Acr max , A£, Ay , etc computed on the 30 slip systems. The procedure for
obtaining values shown in Table 3 is explained in greater detail by Arakere and Swanson [9]. Other
fatigue parameters used in polycrystalline material were also evaluated in reference [9], besides
AT max . Figure 4 shows that the maximum shear stress amplitude AT max plotted against cycles to
failure, N, has a good correlation for a power law fit. The power law curve fit is shown by Eq. (4).
Ax max = 397,758 N' 01598 (4)
The correlation for [AT max ] would be better if some of the high stress data points are corrected for
inelastic effects. It must be noted that Eq. (4) is only valid for PWA1480/1493 material at 1200F.
Since the deformation mechanisms in single crystals are controlled by the propagation of
dislocations driven by shear the [AT max ] might indeed be a good fatigue failure parameter to use for
crystallographic cracking. However, this parameter must be verified for a wider range of R-values
and specimen orientations, and also at other temperatures and environmental conditions. Equation
(4) will be used to calculate fatigue life at a critical blade tip location for the SSME turbine blade.
Fatigue Failure in SSME HPFTP/AT Single Crystal Turbine Blades
Turbine blades used in the alternate high-pressure fuel turbopump (HPFTP/AT) for the NASA
SSME are fabricated from single crystal nickel PWA1480/1493 material. Many of these blades
have failed during testing due to fretting/galling fatigue cracks in the attachment regions and the
initiation and propagation of fatigue cracks from an area of high concentrated stress at the blade tip
leading edge. The subsequent investigation into the blade tip cracking failure provided insight into
areas where the robustness of the design could be improved to reduce the potential for failure.
During the course of the investigation an interesting development was brought to light. When the
size of the fatigue cracks for the population of blades inspected was compared with the secondary
crystallographic orientation (3 a definite relationship was apparent as shown in Fig. 5 [8, 23],
Secondary orientation does appear to have considerable influence over whether a crack will initiate
and arrest or continue to grow until failure of the blade airfoil occurs. Figure 5 reveals that for (3 =
45+/- 15 degrees tip cracks arrested after some growth or did not initiate at all. This suggests that
perhaps there are preferential [3 orientations for which crack growth is minimized at the blade tip.
In an attempt to understand the effect of crystal orientation on blade stress response at the blade tip
and in the attachment regions a detailed three-dimensional finite element (FE) model capable of
accounting for primary and secondary crystal orientation variation was constructed [10, 24],
7
Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
Description of the Blade Finite Element (FE) Model
The blade FE model is described in detail by the present authors in reference [9], which deals with
the effects of crystal orientation on blade fatigue life. A brief description of the model is repeated
here for completeness. The Alternate High Pressure Fuel Turbo Pump (HPFTP/AT) first stage blade
ANSYS FE model was constructed from a large 3D cyclic symmetry model that includes the first
and second stage blades and retainers, interstage spacer, disk and shaft, and the disk covers [10].
The FE models are geometrically nonlinear due to the contact surfaces between the separate
components. The element type used for the blade material is the ANSYS SOLID45, an 8-noded 3D
solid isoparametric element. Anisotropic material properties are allowed with this element type.
The FE model of the first stage turbine blade with the frequently observed blade tip cracking
problem is shown in Fig. 6. Figure 6 also shows the material coordinate system, which is
referenced, to the blade casting coordinate system.
To examine a wide range of primary and secondary variations in crystal orientation, 297 material
coordinate systems were generated for this study. The primary and secondary crystallographic
orientation is shown in Fig. 7 [8]. Figure 8 shows the distribution of the 297 different material
coordinate systems within the allowed 15-degree maximum deviation from the casting axis. The
secondary repeats after 90 degrees, so only 0 to 80 degrees needed to be modeled. The two angles,
A and T, locate the primary material axis relative to the casting axis, and the third angle, [3, is the
clocking of the secondary material axis about the primary material axis.
The loading conditions represent full power mainstage operation of the Space Shuttle Main Engine,
referred to as 109% RPL SL (Rated Power Level Service Life). The shaft speed is 37,355 rpm, the
airfoil temperature is approximately 1200 F, forces representing the blade damper radial sling load
are applied to the blade platform, and aerodynamic pressures are applied to the blade surfaces and
internal core.
The connection between the blade and disk are modeled with ANSYS COMBIN40 elements, for
cases where the coefficient of friction between the blade-disk attachments was ignored. These
elements have one degree of freedom at each node. The nodal motion in that degree of freedom sets
the separation or contact of these elements only. This element does not have the capability for
friction tangent to the contact surface.
To account for coefficient of friction at the blade attachment regions, the COMBIN40 element was
replaced with CONTAC52 type elements. The CONTAC52 element has 3 degrees of freedom at
each node. Coefficient of friction at the contact is allowed, thus developing tangential traction
forces. The relative position and motion of the end nodes are used to define displacement and
reaction force magnitude and direction at the contact. For FE models including friction at the
attachment regions,
Post processing of the 297 FE output files represented a considerable amount of work. Two
FORTRAN programs, developed at MSFC, were employed for this effort. The first strips the
element results from the coded binary output files and places them into ASCII text files. The
second program processes the ASCII files to calculate averaged nodal results, the resolved shear
stresses and strains and the normal stresses and strains in the single crystal material coordinate
system. Fatigue parameters chosen for study are then calculated and sorted based on user set
8
Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
criteria. Postprocessing of the friction data was performed by constructing a local coordinate
system with one axis in the tangential direction of contact slip on the blade attachment (Fig. 14).
An ANSYS batch-postprocessing routine was written to output tabulated results for selected nodes
in the attachment.
Effect of Crystal Orientation on Blade Tip Stress Response
Effect of crystal orientation on stress response at the blade tip critical point is discussed first,
followed by a discussion of stresses in the blade attachment region. The crack location and
orientation at the critical blade tip location is shown in Fig. 5. Stress response at the blade tip as a
function of crystal orientation is discussed in greater detail in reference [9].
Fatigue life estimation of single crystal turbine blade components represents an important aspect of
durability assessment. Towards identifying a fatigue damage parameter for single crystal blade
material, several fatigue failure criteria used for polycrystalline material subjected to multiaxial
states of fatigue stress were evaluated for single crystal material by Arakere and Swanson [9]. They
are outlined here briefly to facilitate comparisons with fatigue in contact locations.
Turbine blade material is subjected to large mean stresses from the centrifugal stress field. High
frequency alternating fatigue stresses are a function of the vibratory characteristics of the blade.
Kandil, et al [25] presented a shear and normal strain based model, shown in Eq. 5, based on the
critical plane approach, which postulates that cracks initiate and grow on certain planes and that the
normal strains to those planes assist in the fatigue crack growth process. In Eq. (5) y nia x is the max
shear strain on the critical plane, e n the normal strain on the same plane, S is a constant, and N is the
cycles to initiation.
7 m ax+5£„=/W (5)
Socie, et al [26] presented a modified version of this theory shown in Eq. (6), to include mean stress
effects. Here the maximum shear strain amplitude (A)) is modified by the normal strain amplitude
(Ae) and the mean stress normal to the maximum shear strain amplitude (c n o)- These observations
are similar to FCG on octahedral planes for single crystals.
A r t A£ n
2 2
= f(N)
( 6 )
Fatemi and Socie [27] have presented an alternate shear based model for multiaxial mean-stress
loadings that exhibit substantial out-of-phase hardening, shown in Eq. (7). This model indicates that
no shear direction crack growth occurs if there is no shear alternation.
Ar <7 max
+ ~) = f(N) (7)
2 Gy
Smith, Watson, and Topper [28] proposed a uniaxial parameter to account for mean stress effects
which was modified for multiaxial loading, shown in Eq. (8), by Bannantine and Socie [29]. Here
9
Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
the maximum principal strain amplitude is modified by the maximum stress in the direction of
maximum principal strain amplitude that occurs over one cycle.
Ac
T (cr max ) = f (N) (8)
Equation 8 has been used successfully for prediction of crack initiation in fretting fatigue, for
aluminum and titanium polycrystalline alloys [12]. The polycrystalline failure parameters based on
equations (5-8) was applied for single crystal uniaxial LCF test data, shown in Table 2, by
evaluating the fatigue damage parameters on all of the 30 slip systems [9, 10]. However, the most
effective fatigue damage parameter was found to be maximum shear stress amplitude AT max shown
in Fig. 4.
Variation of crystal orientation on stress response at the blade tip critical point prone to cracking
(tip point on inside radius) was examined by analyzing the results from the 297 FE model runs. The
FE node at the critical point was isolated and critical failure parameter value (Ax max ) computed on
the 30 slip systems. A contour plot of At max was generated as a function of primary and secondary
orientation, shown in Fig. 9. The contour plot clearly shows a minimum value for Ax max for
secondary orientation of P = 50 deg and primary orientation designated by cases 5 and 20. Case 5
corresponds to a primary orientation of A = 0 deg and T = 7.5 degree. Case 20 corresponds to a
primary orientation of A = 5.74 deg and T = 13.86 degree. Using the fatigue life equation based on
the Atmax curve fit of LCF test data (Eq. 4) we can obtain a contour plot of dimensionless life at the
critical point as a function of primary and secondary orientation, as shown in Fig. 10. The maximum
life is again obtained for p = 50 deg, and A = 0 deg and T = 7.5 deg, and A = 5.74 deg and T = 1 3.86
deg. The optimum value of secondary orientation P = 50 deg. corresponds very closely to the
optimum value of p indicated in Fig. 5. This demonstrates that control of secondary and primary
crystallographic orientation has the potential to significantly increase a component's resistance to
fatigue crack growth without adding additional weight (for the airfoil region).
Stress Response in the Attachment Region (Frictionless Contact)
The 297 FE models representing 33 variations (cases) in primary crystal orientation and 9 variations
in the secondary orientation were run for frictionless contacts at the blade attachments. Blade
loading conditions were kept constant while the crystal orientation was varied. Figure 1 1 shows
representative stress plots for the blade attachment region. The critical point with the most severe
contact stresses is shown in Fig. 12. Fretting damage in the attachment regions is highlighted in
Figs. 13-14. Effects of variation of primary and secondary orientation on stress response are studied
for this critical region. Figures 15-16 show contour plots of At max and t max * Ay/2, as a function of
primary and secondary orientation. The crystal orientation does not appear to have any discemable
correlation at the contact point for these parameters. Figure 17 shows a contour plot of
(G max Ae/2, Eq. 8) as a function of primary and secondary orientation. The secondary orientation
does appear to have an optimum value of about 55 degrees. Szolwinski and Farris [12] have used
Eq. (8) successfully as a fretting damage parameter for polycrystalline alloys. For single crystal
material, however, the fretting damage parameters need to be evaluated in greater detail. A simpler
geometry such as a cylindrical fret pin made of polycrystalline material contacting a flat plate made
of single crystal material with a specific crystal orientation would be amenable to an analytical
10
Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
solution. Surface and subsurface stresses at the contact can then be examined in detail, and fretting
damage parameters evaluated, as a function of crystal orientation, coefficient of friction, slip
amplitude, loads, etc.
Figures 1 and 2 illustrate that subsurface shear stresses from fretting action, at low or even zero slip
amplitude, can induce subsurface crystallographic crack initiation, with stage I-fatigue cracks
propagating outward to the surface. No crack involvement of the fretted surface occurs and
negligible surface damage is noted at the contact surfaces. Clearly, an effective model for predicting
crystallographic crack initiation under mixed mode loading is required for life prediction under
fretting action.
This study shows that stress at the attachment region varied significantly as a function of crystal
orientation alone, from x max = 60 ksi to x max = 130 ksi. This indicates that by selecting the appropriate
crystal orientation, for specified geometry and loading, it may be possible to optimize blade
structural integrity.
Stress Response in the Attachment Region Including Contact Friction
The presence of friction results in the development of traction forces at the fretting contact.
Evaluation of the contact forces and the resulting contact stresses at the interface require careful
analysis of the interaction between the contacting surfaces and applied loads. To study the effect of
variation of friction coefficient at the contact, and variation of primary and secondary crystal
orientations, 18 FE model runs were completed. The 18 FE models comprised of 3 primary
orientations (cases 0, 5 and 30), 2 secondary orientations (P = 0 and 50 deg.), and 3 values for
coefficient of friction ((l = 0, 0.3, 0.7). Figure 12 shows representative contact stress results in the
critical attachment region. Blade loading is kept constant for all 18 cases. Figure 18 shows contour
plots of tangential normal stress, ax, at the contact surface. The X coordinate is in the direction of
slip, as shown in Fig. 12. The tangential normal stress is of practical interest because cracks are
thought to initiate at locations where ax reaches a maximum tensile value. The maximum value of
tangential normal stress, ax, reached in the contact zone is listed in Table 4 for all the 18 cases. It is
seen that <J\ increases with jx. ax is also seen to vary considerably with variations in primary and
secondary orientation. Since the component stiffness varies with crystal orientation the stress
distribution is also expected to vary, under constant loading. Minimum values of G\ are reached for
case 30 and p=0 [(84.3 ksi, (1=0), (89.9 ksi, (1=0.3), (97.7 ksi, (1=0.7)] indicating that perhaps key
design parameters can be optimized for specific blade geometry and loading. Additional fretting
damage parameters need to be examined, for design optimization possibilities.
The frictional contact study showed that both a x and x max stresses at the attachment region varied by
about 20%, as a function of crystal orientation alone. For instance, the ax variations (Table 4) as a
function of crystal orientation alone are 84.3 - 108.6 ksi for (I = 0, 89.9 - 1 14.4 ksi for (i = 0.3, and
97.7 - 122.6 ksi for (t = 0.7. The variations would have been larger if peak variations due to crystal
orientation were realized. It appears that if case 25 with P=30° was compared with case 17 at P=0°
or, P=20° stress variations due to crystal orientation would have been considerably larger than 20%.
Figures 13 and 14 show representative fretting/galling induced cracks in the blade attachment
region for HPFTP/AT first stage blades. Region shown in Fig. 13 is on the suction side, trailing
11
Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
edge of the blade. Several arrest marks are also visible. The blade crystal orientation is A = -6.7°, y
= 11.3°, |3 =4.2°. Figure 14 shows fretting/galling induced cracking showing multiple origins with
stage II cracks. The crystal orientation for this blade is A = -2°, y = 3°, (3 =7°. Many blades in
different engine units exhibited similar blade attachment cracks. In the blade casting process, the
secondary crystal orientation is not controlled, while the primary crystal orientation is controlled to
within 15°. A systematic investigation of severity of fretting/galling induced attachment cracks,
similar to study shown in Fig. 5 for blade tip cracking, has to be done to discern relationships
between fretting damage and crystal orientation. Tolerance in clearance between the blade and disk
attachment surfaces also plays an important role in the contact stress distribution. The analysis
results presented are for nominal clearances between mating parts.
Conclusions
Three dimensional finite element stress analysis of HPFTP/AT SSME single crystal turbine blades
subjected to rotational, aerodynamic, and thermal loads is presented. Stress response at the blade tip
and attachment regions are presented as a function of primary and secondary crystal orientation. 297
FE models are analyzed to study a wide range of variation in crystal orientation. For the blade tip
location, the maximum shear stress amplitude [Ar max ] on the 30 slip systems was found to be an
effective fatigue failure criterion, based on the curve fit between AT max and LCF data at 1200F for
PWA1480/1493. Variation of At max as a function of crystal orientation at the blade tip reveals that
control of secondary and primary crystallographic orientation has the potential to significantly
increase a component's resistance to fatigue crack growth without adding additional weight or cost.
Current seeding techniques used in the single crystal casting process can readily achieve the degrees
of control in primary and secondary crystallographic orientation required. Fretting contact stresses
are presented for the attachment region. Attachment stresses are seen to reach peak values at
locations where fretting cracks have been observed. For frictionless contact, stress at the attachment
region varied significantly as a function of crystal orientation alone, from T max =60 ksi to T max = 130
ksi, under constant loading conditions. Attachment stresses varied significantly for models that
included coefficient of friction at the contact. The tangential normal stress ax increased with
increasing coefficient of friction at the critical contact location. This might be of practical interest
since cracks are thought to initiate at locations where ax reaches a maximum tensile value, for
polycrystalline materials. The parameter Ax max , which showed a definite correlation with crystal
orientation at the blade tip, appears to have no discemable correlation at the critical contact region.
The damage parameter a ma x Ae/2 appears to be correlated with the secondary orientation. Fretting
damage parameters need to be evaluated in greater detail for single crystal materials.
Subsurface shear stresses induced by fretting action, at low or even zero slip amplitude, can induce
subsurface crystallographic crack initiation, with no crack involvement of the fretted surface. An
effective model for predicting crystallographic crack initiation under mixed mode loading is
required for life prediction under fretting action. This will necessitate specialized testing to account
for crystal orientation and mixed mode loading effects. One way to conduct these tests is to use
Brazilian disk specimens as outlined by John, et al [21], to study mixed mode loading in single
crystals. However, to study crystallographic crack initiation behavior, the disks would have to be
tested without starter flaws, requiring high compressive loads and potential for failure at the loading
points. One solution perhaps is to thin the disks at the center which necessitates lower compressive
loads to initiate a crystallographic crack at the disk center.
12
Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
Acknowledgements
The authors would like to thank the NASA/ASEE 1999 Summer Faculty Fellowship Program. The
support from this program, administered by the University of Alabama in Huntsville, enabled the
first author to perform part of this work at the NASA Marshall Space Flight Center, Huntsville, AL.
REFERENCES CITED
[J] D. A. Hills and D. Nowell, Mechanics of Fretting Fatigue, Kluwer, Deventer, 1 994.
[2] B. A. Cowles, High cycle fatigue in aircraft gas turbines-an industry perspective,
International Journal of Fracture, (1996) 1-16.
[3] J. Dombromirski, Variables of Fretting Process: Are There 50 of them? Standardization of
Fretting Fatigue Test Methods and Equipment, ASTM, pp. 60-68, 1990.
[4] Deluca, D., Annis. C, "Fatigue in Single Crystal Nickel Superalloys"; Office of Naval
Research, Department of the Navy FR23800, August 1995.
[5] C.T. Sims, “Superalloys: Genesis and Character”, Superalloys - II, Eds., C.T. Sims, N.S.
Stoloff and W.C. Hagel, Wiley & Sons, New York, New York, (1987), p. 1.
[6] F.L. VerSnyder and R.W. Guard, “Directional Grain Structure for High Temperature
Strength”, Trans. ASM, 52, (1960), p. 485.
[7] M. Gell and D.N. Duhl, “The Development of Single Crystal Superalloy Turbine Blades”,
Processing and Properties of Advanced High-Temperature Materials, Eds., S.M. Allen, R.M.
Pelloux and R. Widmer, ASM, Metals Park, Ohio, (1986), p. 41.
[8] J. Moroso, “Effect of Secondary Orientation on Fatigue Crack Growth in Single Crystal
Turbine Blades”, M.S. Thesis, Mechanical Engineering Department, University of Florida,
Gainesville, FL, May 1999.
[9] N. K. Arakere and G. Swanson, “Effect of Crystal Orientation on Fatigue Failure of Single
Crystal Nickel Base Turbine Blade Superalloys,” accepted for presentation in the ASME
IGTI conference (May 8-1 1, 2000, Munich, Germany) and for publication in the ASME
Journal of Gas Turbines and Power.
[10] N. K. Arakere and G. Swanson, “Fatigue Failure of Single Crystal Nickel Base Turbine
Blade Superalloys,” accepted for publication as a NASA Technical Paper.
[11] Giannokopoulos and S. Suresh, “3D Fretting Stresses,” Acta Materialia, 1999.
[12] M. P. Szolwinski and T. N. Farris, “Mechanics of Fretting Fatigue Crack Formation,” Wear,
v 198, 1996, pp. 93-107.
[13] M. H. Attia and R. B. Waterhouse, Editors, Standardization of Fretting Fatigue Test
Methods and Equipment, ASTM (04-01 1590-30), STP 1 159, (1992)
[14] D. W. Hoeppner, Mechanisms of Fretting Fatigue and Their Impact on Test Methods
Development, Standardization of Fretting Fatigue Test Methods and Equipment, ASTM, pp.
23-32, 1990.
[15] O. Vingsbo and D. Soderberg, On fretting maps. Wear, 126 (1988) 131-147.
[16] C. Ruiz, P. H. B. Boddington and K. C. Chen, An investigation of fatigue and fretting in a
dovetail joint, Experimental mechanics, 24 (3) (1984) 208-217.
[17] Stouffer, D. C., and Dame, L. T., “Inelastic Deformation of Metals,” John Wiley & Sons;
1996.
13
Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
[18] Personal Communication with D. P. DeLuca, Pratt & Whitney, (Government Engines and
Space Propulsion), Mechanics of Materials, West Palm Beach, FL.
[19] Telesman, J.; Ghosn, L, “The unusual near threshold FCG behavior of a single crystal
superalloy and the resolved shear stress as the crack driving force,” Engineering Fracture
Mechanics, Vol. 34, No. 5/6, pp. 1183-1196, 1989.
[20] Deluca, D. P. and Cowles, B. A., “Fatigue and fracture of single crystal nickel in high
pressure hydrogen”, Hydrogen Effects on Material Behavior, Ed. By N. R. Moody and A.
W. Thomson, TMS., Warrendale, PA, 1989.
[21] John, R., DeLuca, D. P., Nicholas, T., and Porter, J., “Near-threshold crack growth behavior
of a single crystal Ni-base superalloy subjected to mixed mode loading,” Mixed-Mode Crack
Behavior, ASTM STP 1359, Editors. K. J. Miller and D. L. McDowell, Paper ID: 5017,
November 1998.
[22] M. McLean, “Mechanical Behavior: Superalloys”, Directionally Solidified Materials for
High Temperature Service, The Metals Society, London, (1983), p. 151.
[23] Pratt and Whitney, “SSME Alternate Turbopump Development Program HPFTP Critical
Design Review,” P&W FR24581-1 December 23, 1996. NASA Contract NAS8-36801.
[24] Sayyah, T., “Alternate Turbopump Development Single Crystal Failure Criterion for High
Pressure Fuel Turbopump First Stage Blades,” Report No.: 621-025-99-001 , NASA
Contract NAS 8-40836, May 27, 1999.
[25] Kandil, F. A., Brown, M. W., and Miller, K. J., “Biaxial low cycle fatigue of 316 stainless
steel at elevated temperatures,” Book, pp. 203-210, Metals Soc., London, 1982.
[26] Socie, D. F., Kurath, P., and Koch,J., “A multiaxial fatigue damage parameter,” presented at
the Second International Symposium on Multiaxial Fatigue, Sheffield, U.K., 1985.
[27] Fatemi, A, Socie, D, “A Critical Plane Approach to Multiaxial Fatigue Damage Including
Out-of-Phase Loading,” Fatigue Fracture in Engineering Materials, Yol. 1 1, No. 3, pp. 149-
165, 1988.
[28] Smith, K. N., Watson, P., and Topper, T. M., “A stress-strain function for the fatigue of
metals,” Journal of Materials, Vol. 5, No. 4, pp. 767-778, 1970.
[29] Banantine, J. A., and Socie, D. F., “Observations of cracking behavior in tension and torsion
low cycle fatigue,” presented at ASTM Symposium on low cycle fatigue - Directions for the
future, Philadelphia, 1985.
14
Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
LIST OF FIGURES
Fig. 1 A subsurface fretting fatigue crack emanating from a carbide in a turbine blade
attachment (PWA1422) and propagating along octahedral (111) shear planes [4].
Fig. 2 Subsurface fretting fatigue crack initiation in a single crystal Ni turbine blade (platform
tip) [18]
Fig. 3 Strain range Vs. Cycles to Failure for LCF test data (PWA1493 at 1200F)
Fig. 4 Shear Stress Amplitude [AT max ] Vs. Cycles to Failure
Fig. 5 Secondary Crystallographic Orientation, p, Vs Crack Depth for the SSME AHPFTP 1 st
Stage Turbine Blade [8, 23]
Fig. 6. First Stage Turbine Blade FE Model and Casting Coordinate System
Fig. 7 Convention for Defining Crystal Orientation in Turbine Blades [8]
Fig. 8 33 primary crystallographic orientations (cases) and 9 secondary crystallographic
orientations ((3 or 0 values), for a total of 297 material orientations.
Fig. 9 Maximum Shear Stress Amplitude (Ax max , ksi) Contour Plot at the Blade Tip Critical
Point
Fig. 10 Normalized HCF life (Contour Plot) at the blade tip Critical Point, as a function of
primary and secondary orientation
Fig. 11 Representative stress plots for the single crystal blade attachment region
Fig. 12 HPFTP/AT first stage blade vonMises stress plot with local zoom in of the suction side
upper contact region at the blade leading edge and the local coordinate system used for
the contact results.
Fig. 13 Fretting/galling induced crack in the contact region (suction side, trailing edge of blade)
Several arrest marks are visible. Crystal orientation: A = -6.7°, y = 1 1 .3°, p =4.2°.
Fig. 14 Fretting/galling induced cracking showing multiple origins and stage II cracks (pressure
side trailing edge location). Crystal orientation: A = -2°, y = 3°, p =7°.
Fig. 15 Contour plot of max shear stress amplitude, Ax ma x (ksi), at the critical contact location,
as a function of primary (Case number) and secondary (P or 0) crystallographic
orientation.
Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
Fig. 16 Contour plot of the parameter, T max *(Ay/2), at the critical contact location, as a function
of primary (Case number) and secondary (P or 0) crystallographic orientation.
Fig. 17 Contour plot of the parameter, a max *(Ae/2), at the critical contact location, as a function
of primary (Case number) and secondary (p or 0) crystallographic orientation.
Fig. 18 Case = 0, P = 0, p = 0, Contour plot of tangential surface stress, G x , in upper lobe,
suction side, near leading edge.
LIST OF TABLES
Table 1 Slip Planes and Slip Directions in a FCC Crystal [17]
Table 2 Strain controlled LCF test data at 1200 F for 4 specimen orientations
Table 3 Maximum values of shear stress and shear strain on the slip systems and normal stress
and strain values on the same planes.
Table 4 Values of crystal orientation, friction coefficient, and max tangential normal stress ax
at the critical contact location, for the 1 8 FE model runs
Slip Plane
Slip Direction
Octahedral Slip a/2<l 10>f 1 1 1]
12 Primary Slip Directions
[111]
[1 0 -1]
[111]
[0 -i 1] !
[111]
[1 -1 0]
[-1 1 -1]
[1 0 -1]
M 1 -1]
[110] I
[-1 1 -1]
[0 1 1]
[1 -1 -1]
[1 1 0]
[1 -1 -1]
[0 -1 1]
[1 -1 -1]
[1 0 1]
[-1 -1 1]
[0 1 1]
M -1 1]
[1 0 1]
[-1 -1 1]
[1 -1 0]
Octahedral Slip a/2<l 12>f 1 1 1
[111]
[111]
[111]
[-1 1 - 1 ]
[-1 1 - 1 ]
[-1 1 - 1 ]
[1 -1 - 1 ]
[1 -1 - 1 ]
[1 -1 - 1 ]
[-1 -1 1 ]
[-1 -1 1 ]
[-1 -1 1 ]
Cube Slip a/2<l 10>{ 100
[1 0 0]
[0 1 0]
[0 1 0]
[0 0 1]
[0 0 1]
12 Secondary Slip Directions
[-1 2 - 1 ]
[2 -1 - 1 ]
[-1 -1 2 ]
LLAJJ
[i -i -2]
[-2 -i n
[-1 1 - 2 ]
[2 1 1 ]
[-1 ~2 1 ]
[-2 1 - 1 ]
[1 -2 -1]
[1 1 2 ]
6 Cube Slip Directions
[0 1 1]
[0 1 -i]
[1 o n
[1 0 -i]
[1 1 0]
[-1 1 0 ]
Table 1 Slip Planes and Slip Directions in a FCC Crystal [17]
Specimen
Orientation
Max
Test Strain
Min
Test Strain
R
Ratio
Strain
Range
Cycles
to
Failure
<001 >
.01509
.00014
0.01
.01495
1326
<00 1>
.0174
.0027
0.16
0.0147
1593
<001 >
.0002
0.02
0.011
4414
<00 1>
.01202
.00008
0.01
0.0119
5673 1
<001 >
.00891
.00018
0.02
.00873
29516
<1 1 1>
.01219
-0.006
-0.49
.01819
26
<11 1>
.0096
.0015
0.16
0.0081
843
<1 1 1>
.00809
.00008
0.01
.00801
1016
<1 1 1 >
.006
0.0
0.0
0.006
3410
<1 1 1 >
.00291
-0.00284
-0.98
.00575
7101
<11 1>
.00591
.00015
0.03
.00576
7356
<1 1 1>
.01205
0.00625
0.52
0.0058
7904
■ESEaBi
.01212
0.0
0.0
.01212
79
■
.00013
0.02
.00782
4175
.00005
0.01
.00596
34676
BiH
.006
0.0
0.0
0.006
114789
<01 1>
.0092
.0004
0.04
0.0088
2672
<01 1 >
.00896
0.01
.00883
7532
<01 1 >
.00695
0.03
.00676
30220
Table 2 Strain controlled LCF test data at 1200 F for 4 specimen orientations
g <D
j=2 3
> 1
,g .c
c3
ft CA
c c
2 S
S £
« ‘3
o o
Cu Cl
«a So
5 .5
6 £
3 §
c e
2 ’3
b j3
V5 V
ft to
<U <D
w +- »
1 §
e e
<r> *3
O (D
Ch ex
«« CO
X C
«j -5
£ £
0,0 0,0
lb <— I 4— I 4-4
00 CO CO CO
£ £ £ £
<D 4> O O
W W ^ 4— 1
{/) W (/5 Cfl
>> ^
CO CO CO CO
CW Oh Oh
CO CO CO CO
O o O O
co m m co
o o
e c
'Si £
b b
CO CO
& £
<D O
■s -s
S 3
s s
II II
1 1
5 s
o o
co co
CO co
<L> P
b b
co CO
£ £
o <d
•s -s
3 3
S S
ii ii
s I
J ^
0 1
*0 O M
£ ^ «
W fe
cn ^ m 'S
oj ck — r- , 7 :
£23 8 §
26
843
1016
3410
7101
7356
7904
un %c
S ^ ^
04 CS ^
ocn ^
on g
04 O- g
3 >5J
< °*
7.68E+04
7.54E+04
5.65E+04
6.17E+04
4.57E+04
237E+05
1.05E+05
1.05E+05
7.84E+04
7.50E+04
7.50E+04
8.20E+04
1.30E+05
8.46E+04
6.50E+04
6.50E+04
n n n
?? ?
IdU Id
m o 0-
'sD O- 04
i *
2 w § o o
r- on 2
m ^
-7.80E+04
1.96E+04
1045
0
-3.70E+04
1959
7.80E+04
O § o o
OO r - 1
co o 00
U-) ^ KT\
r- co
j*
't 't
off f O
w w w w w
w, m vo o* o-
r» ® in
d od voi vd d
1 .59E+05
1.25E+05
1.06E+05
7.84E+04
3.80E+04
7.70E+04
1.60E+05
1.30E+05
8.60E+04
6.50E+04
6.50E+04
in in in
O Q O
+ T +
WWW
m o —
o- r^i
-5 1
1 .08E+05
1 .06E+05
7.98E+04
8.73E+04
6.47E+04
3.35E+05
1.49E+05
1.48E+05
1 , 1 0E+05
1.07B+05
1.06E+05
M5E+05
inn 5 ^
??? f
uimn u
S38 8
— « t— od od
n in Tf
?? ?
W U UJ
oo n n
— 04 O
W W ov
B
*?cs
og? ° =
~ ov
|£5 Tf 3 w-i
?f « ■
W tq ^ o S t w
2^2 ° SS2
7M «? —
<N
0^0 O
CO 04
cn m
co ° in
W-j 04
X
vl *
>0 IO ^ ^ Tt
??? f ?
UJ W W U UJ
© vo m m r-
— o< — tt
W — ’ 00 00 sd
2.25E+05
1.77E+05
1.49E+05
L10E+05
5.40E+04
1.09E+05
2.25E+05
1.60E+05
1.06E+05
8.00E+04
8.00E+04
in m, tj-
C Q ©
+ T +
WWW
mm o
04 04 m
— : w d
<N
1
0.0004804
0.000661
0.0003604
0.000397
0.0002945
0.00153
0.000675
0.0006733
0.0005
0.000485
0.0004825
0.0005
0.001
0.0006395
0.00049
0.00049
^ E 2
§S 8
d O o
■
9.25E-06
1.78E-04
1.32E-05
0
0
-1.01E-03
0.00025
1.34E-05
0
-0.00048
2.50E-05
0.001
un
o
o§o o
<N
^ in
o o
W o w
oo O
P p
W oo
X
I
CO
0.00097
0.0015
7.34E-04
7.94E-04
5.89E-04
2.05E-03
0.0016
0.00136
0.001
0.00049
9.90E-04
0.002
CO
<n 22 O' g\
ml
O O' O'
m m 04
88 8
odd
CS
<T
0.0099075
0.0097
0.007368
0.008
0.006
0.01053
0.00462
0.004703
0.0038
0.00335
0.003362
0.0035
gS| 1
Sgl 8
° P o d
o
m m
o> in ^
— r- m
o o m
8 § 8
o o
_c
A
m
OC vO o
22% ° °
Oo«
o 04
-7.06E-03
0.00176
9.40E-05
0
-0.0033
1.76E-04
0.007
f
ogo o
o>
s s
W o w
o w o
in — ;
vd m
X
PJ
?s
-K.. m in vo <n
S M - — f
P o o © o
° o b o' b
Tt — w-i vo xj- o\ Tj-
OOO O O b O
ocv P P °. o
00 (S oo oo
- -h co oo
§§s 8
■
<= §
<y < h
e |
’8 S
a-c
*r v
A "V V
o 1! 11
© * g
v 1 1
A *U*U
^ II II
v 1 1
w
A r V'V
2 ii ii
OJ 2 g
v 1 J
A - ^
^ II II
O x x
° 1 J
m
<D
|
"E.
<D
I
00
Primary Orientation
Case Number
Secondary
Orientation
A deg
Friction
Coefficient
V
Max Gx
(Tangential
normal
stress, ksi)
0 (A=0, y^O)
0
0.0
99.4
0
0
0.3 I
106.8
0
0
0.7 1
116.9
0
50
0.0
103.9
0
50
0.3
109.9
0
50
0.7
118.2
5 (A=0, 7=7.5)
0
0.0
105.7
5
0
0.3
113.1
5
0
0.7
123.3
5
50
0.0
108.6
5
50
0.3
114.4
5
50
0.7
122.6
30 (A=5.74, 7*=- 13.86)
0
0.0
84.3 (min)
30
0
0.3
89.9 (min)
30
0
0.7
97.7 (min)
30
50
0.0
94.1
30
50
0.3
98.8
30
50
0.7
105.2
Table 4 Values of crystal orientation, friction coefficient, and max tangential normal
stress Ox at the critical contact location, for the 18 FE model runs
Nagaraj K. Arakere and Gregory Swanson
Fretting Stresses in Single Crystal Superalloy Turbine Blade Attachments
320X
Fig. 1 A subsurface fretting fatigue crack emanating from a carbide in a turbine blade
attachment (PWA1422) and propagating along octahedral (111) shear planes [4],
Strain Amplitude (Uniaxial LCF Data)
Power Law Curve Fit ( R A 2 = 0.469) : Ae = 0.0238 N
• •
1000 10000 100000 1000000
Cycles to Failure
Fig. 3 Strain range Vs. Cycles to Failure for LCF test data (PWA1493 at 1200F)
Max Shear Stress Amplitude on Slip Planes
Power Law Curve Fit (R A 2 = 0.674~) : Ax = 397,758 N
Fig. 4 Shear Stress Amplitude [AT max ] Vs. Cycles to Failure
Fig. 5 Secondary Crystallographic Orientation, (3, Vs Crack Depth for the
SSME AHPFTP 1st Stage Turbine Blade [8, 23]
Fig. 7 Convention for Defining Crystal Orientation in Turbine Blades [8]
a total of 297 material orientations.
Case Number (Primary Orientation)
80 91
Theta, Degrees (Secondary Axis Orientation)
Fig. 9 Maximum Shear Stress Amplitude (Atmax. ksi) Contour Plot at the
Blade Tip Critical Point
Theta, Degees (SeocndaryOriertaticn)
Fig. 10 Normalized HCF life (Contour Plot) at the blade tip Critical Point, as a function
of primary and secondary orientation
Crystal Orientation: Case 2, 0 Primary, 22.5 Secondary, Von Mises, Suction Side
(j!
iUiiiiisi j
fuel pump, assembly with refined MODELS, CASE 2
ANSYS 5.2
FEB 28 1997
12:19:44
NODAL SOLUTION
STEP=3
SUB =1
TIME=3
SEQV (AVG)
DMX =.035574
SMN =426.469
SMX =369035
SMXB=513816
I , -12500
[ — i 0
md 12500
HI 25000
37500
HI 50000
62500
75000
87500
100000
CHI
Crystal Orientation: Case 2, 0 Primary, 22.5 Secondary, Von Mises, Pressure Side
ANSYS 5.2
FEB 28 1997
12:44:01
NODAL SOLUTION
STEP=3
SUB =1
TIME=3
SEQV (AVG)
DMX =.035574
SMN =426.469
SMX =369035
SMXB=513816
, , -12500
rH 0
mm ^2500
25000
pm 37500
s iss
™ 87500
1 1 100000
FUEL PUMP, ASSEMBLY WITH REFINED MODELS, CASE 2
Fig. 11 Representative stress plots for the single crystal blade attachment region
1
Fig. 12 HPFTP/AT first stage blade vonMises stress plot with local zoom in of the suction side upper
contact region at the blade leading edge and the local coordinate system used for the contact results.
Fig. 13 Fretting/galling induced crack in the contact region (suction side, trailing edge of blade)
Several arrest marks are visible. Crystal orientation: A = -6.7°, y= 1 1.3°, p =4.2°.
r
, 3
<o
TD
•
C/3
<D
P
C/3
on
C/3
o
O
c/3 rn
.5 ii
60 ^
*C
° «
<D i
E ii
1-
^ o
bo ■-
P cj
ii
“ 13
C 4— >
1 &
2 U
U ^
T3 c
<U o
O • -
*2 °
.S o
bO iu
C bO
P5 t 3
13 ^
bD W)
’bb .5
a ~~
tJ
£
Pu
rr
bi
•
ta
Case Number (Primary Orientation)
Contour Plot of Max Shear Stress
Theta, Degrees (Secondary Orientation)
Fig. 15 Contour plot of max shear stress amplitude, AT max (ksi), at the critical contact
location, as a function of primary (Case number) and secondary (P or 0)
crystallographic orientation.
Case Number (Primary Orientation)
Contour Plot of t *yt 2
max
Theta, Degrees (Secondary Orientation)
Fig. 16 Contour plot of the parameter, t max *(Ay/2), at the critical contact location, as a
function of primary (Case number) and secondary (P or 0)
crystallographic orientation.
Case Number (Primary Orientation)
Contour Plot of Max normal * Normal Strain/2
1 1 1 1 1 i i i i
0 10 20 30 40 50 60 70 80 90
Theta, Degrees (Secondary Orientation)
Fig. 17 Contour plot of the parameter, G ma x*(A£/2), at the critical contact location, as a
function of primary (Case number) and secondary ((3 or 0)
crystallographic orientation.
o
o o o o o o
mvoooooooooo
NOtOinOLOOOOOQ
\rin^-oiMinr^oLnoino
in ■Hcn-Hivxr'dorv.inC'JiM
CM CO I C*> I I I I rif'l'iJO'rl
II II II II I II II I II II II I II
Dashed line indicates tl
boundary of the contac
between the blade and
disc attachments, at the
critical region.
