'B..
.
:
.
.
.
-
-- -
-
--
-
--
--
-
-...--
."
----
.
..
.
NATIONAL BUREAU OF STANDARDS - 1963
MICROCOPY RESOLUTION TEST CHART
| 1.25 1.4 11.6
9!
om -
50
1753
ORNL P
I OF 2
-
.
.
..
.
.
-
-
-
---
-
--
I
1
-
-
--
.
..
-
-
-
-
.......
-
CRNL P-17srz
Conf. 451017-2
18 165€
DEFORMATION, CREEP, AND FRACTURE IN ALPHA-ZIRCONIUM ALLOYS*
D
ORNL - AEC - OFFICIAL....

.
.
-.
-
M. L. Picklesimer
Metals and Ceramics Division
Oak Ridge National Laborator
Oak Ridge, Tennessee
RALLASAD FOR ANNOUNCEMEN:
IN NUCLEAR SCIENCE ABSTRACTS
ABSTRACT
·
The anisotropy of mechanical properties observed in a-zirconium alloys can
be related to the preferred orientation of the material. The influence of
preferred orientation on anisotropy of yield strengths, plastic flow, creep,
fracture, their strain rate and temperature sensitivities, and the stability of
such textures to further deformation are discussed in terms of the deformation
systems observed in single crystals. It is concluded that other deformation
systems are forced to operate because of the restraint of neighboring grains
in polycrystalline materials.
INTRODUCTION
The development of zirconium-base alloys is being forced by the demands for
VAN
stronger and lighter sections of more corrosion-resistant materials having low-
neutron absorption cross section for structural use in nuclear reactor systems.
Similarly, the engineering use of such metals as beryllium; magnesium, and
titanium is being forced by demands for higher strength-to-weight ratios and
stiffer sections at operating temperatures in waterials for use in supersonic
aircraft, missiles, 'satellites, and space vehicles. While the theme of this
symposiun concerns zirconium-base alloys, the problem of understanding and
using to advantage the inherent anisotropy is common to all metals having a
close-packed-hexagonal crystal structure.
ORNL - AEC - OFFICIAL
*Research sponsored by the U.S. Atomic Energy Commission under contract
with the Union Carbide Corporation.
.......
...
........
....
.
ORNL-AEC - OFFICIAL
of the 68 metallic elements solid at room temperature, 24 are close-packed
hexagonal (cph), 15 are face-centered-cubic (fcc), 15 are body-centered cubic :
(bcc), 3 are rhombohedral, 3 are orthorhomibc, 2 are diamond cubic, 2 are :
tetragonal, 2 are hexagonal, and 2 are cubic. Of these, 11 fcc., bcc, 10 cph,
and 1 orthorhombic elements are of sufficient importance, as, commercial or special
ORNI - AEC - OFFICIAL
structural materials, that a knowledge of their mechanical properties is required.
Preferred orientation, deformation, creep, and fracture of cubic metals have been
the subject of numerous studies; relatively few studies have been conducted on
the cph metals.
Examination of the literature on deformation, creep, and fracture in
several cph metals shows that such materials differ markedly in their behavior
among themselves, as well as differing in behavior from the better understood
fcc and bcc metals. Examination of the literature on design practice, evaluation
--..
....
tests, theories of the mechanical state, and theories of the anisotropy of
mechanical properties shows that the criteria developed for and from the use of
the cubic metals can not be applied satisfactorily to cph metals. If cph metals
are to be used safely and economically in engineering structures, new criteria,
tests, and theories must be developed, and engineers must be educated in their
use,
ANISOTROPY OF MECHANI CAL PROPERTIES.
-
.
.
.....
Anisotropy of yield and flow strengths has long caused difficulties in
sheet forming operations, causing earing and tearing in many materials,
including the cubic metals. Yet anisotropy, if controlled, can be used to
strengthen sections, improve drawability, stiffen structures, and .
increase maximum allowable design stresses in regions of known conditions of
....
..
.
..
EC - OFFICIAL
.
.
biaxial stress.
ORNE-AEC - OFFICIAL
--.
.
i v..........
.
.
.
This is true of the cubic metals, which are isotropic...
LEGAL NOTICE
This report was prepared as an account of Government sponsored work. Nelther the Unltod'
States, nor the Commission, nor any person acting on bubalf of the Comission:
A. Makes any warranty or representation, expressed or implied, with respect to the accu-
racy, completeness, or usefulness of the information contained in this report, or that the use
of any information, apparatus, mothod, or procosa dioclosed in this roport may not infringe
privately owned righto; or
B. Assumes any liabilities with respect to the use of, or for damages roswting from the
use of any information, apparatus, method, or process disclosed in this report.
As used in the above, "person acting on beball of the Commission” includes any om-
ployeo or contractor of the Commission, or employee of such contractor, to the oxtont that
guch employee or contractor of the Commission, or employeo of much contractor proparos,
diosominatos, or provides access to, any Information pursuant to his employment or contract
with the Commission, or his omploymont with such contractor.
---
c
in....
·
3
It can be done more effectively in the cph metals, which are inherently aniso-
ORAL-AEC - OSPICIAL
tropic. For example, recent investigations indicate that the maximum permissible :
design stress in Zircaloy-2 pressure tubing could be increased by 50% or more,
simply by using the anisotropy of yield strengths due to the preferred orienta-
tion (texture) that is present in the material to fit the application. 1.
· In any discussion of anisotropy, it is necessary to define what property is
being examined, what is meant by anisotropy, and what specific test is used to
evaluate that anisotropy. If a property varies with direction in a material,
the material is said to be anisotropic in that property. The variation may be
with crystallographic direction, or it may be with the reference axes defined for
the body. The electrical industry is using highly textured silicon-iron alloy
sheet (about 3 wt % Si) in transformer cores to increase the electrical efficiency
and decrease the amount of heat generated.
In this case, the anisotropic property
of concern is magnetic permeability, which varies with crystallographic direction.
The commercial fabrication and annealing schedules are deliberately controlled
to produce consistently, in the alloy sheet, a specific crystallographic tex-
ture of a high degree of perfection.
In most discussions of anisotropy, however, it is the variation of yield and
flow stresses with direction in the sheet or plate that is of primary concern.
In most cases, the metals are cubic, and the problems are generally those of
deep-drawing operations. In almost every case, only the anisotropy observed in'
the rolling plane is considered; that is, two-dimensional or planar anisotropy.
Little heed is given to the fact that the properties can also be, and generally
are, different in the tiickness direction of the sheet.
These differences can
greatly affect the planar behavior of the material when it is subjected to
ORNL-AEC - OFFICIAL
biaxial stresses in the plane of the sheet. Determination and conuideration oť
.
ORNL-ALC - OFFICIAL
the three-dimensional anisotropy should result in a better understanding and
wie
control of the properties.
The usual test for determining the anisotropy of plastic flow consists of
the examination of the stress-strain behavior of uniaxially loaded sheet-type
tensile specimens having tensile axes in two or more directions in the plane.
of the sheet. Yield strengths, ultimate strengths, uniform elongations, and
total elongations are compared as functions of the tensile axis directions of
the specimens.
Another evaluation method which is coming into more common usage, particularly
for information on drawability, is a determination of the "R* value. The "R" ,
value is the ratio of the strain in the width direction to the strain in the
thickness direction at some defined point of total axial strain in the sheet-
tensile specimen.
This evaluation does consider to some extent the properties
in the thickness direction of the sheet. Indeed, "R" values can be correlated,
in some materials, to sheet drawability, but it appears that the correlation is
fortuitous rather than as a result of the examination of some basic characteristic
of ühe metal.?
Impact energy absorption as a function of temperature is presently used to
determine the susceptibility of material to crack propagation in a brittle manner
by shock loading. Standard practice requires two specimen orientations, one
.
.
having the major dimension of the specimen in the rolling direction and the
other in the transverse direction but both having notch axes in the normal
direction of the plate. Specimens are broken at several temperatures to determine
the temperature below which the specimen fractures in a brittle manner with low
energy absorption. The impact energy test can be an important measure of
ORNL - AEC - OFFICIAL
anisotropy if notch axes are also located in the rolling plane of the material.”
For example, the impact energy curves for four notch orientations in one lot of
Zircaloy-2 are shown in Fig. 1., If only the conventional notch orientations were
used (curves L and TV) it would be concluded that the material absorbs large
amounts of energy at all. temperatures above -100°C. The data for the notch
axes located in the rolling plane (curves I and In) show that such a conclusion:
ORNI ~ AEC - OFFICIAL
is false, and that all four notch orientations must be studied if the true
Anisotropy of impact behavior is to be determined.
The impact energy curves
for a second lot of Zircaloy-2 (Fig. 2) show that the anisotropy of impact
behavior can vary in the same alloy. The differences in the two sets of curves
are not due to composition differences but are due to differences in the
orientation of the crystallographic texture. The material of Fig. 1. has a con-
centration of basal poles in the normal direction while that of Fig. 2 has a
concentration of basal poles in the transverse direction.
The fracture
appearance for both materials indicated a ductile fracture at all test tempera-
tures. Thus, brittle fracture does not occur in Zircaloy-2, as in some steels;
it simply does not absorb much energy during fracture at certain temperatures.
The reasons for such behavior are not known.
Other measures of anisotropy are fracture appearance in impact-tensile,
tube-burst, and creep specimens; creep rates; stress-rupture tests; and fatigue
tests. These techniques have received' little attention to date but should
become more important as the cph metals are studied more intensively and are
used more generally as structural materials.
Very few investigations of the anisotropy of mechanical properties have
considered compression testing. Almost all have used only the standard round
or sheet tensile specimens. While this may be sufficient information for the
ORNL - AEC - OFFICIAL
OIHL- AEC - OFFICIAL
.
use of the cubic metals, it is not enough for the use of cph metals. For
example, the yield strengths in tension and compression were determined in two
lots of highly, but differently textured, Zircaloy-2 (Ref. 4) with the results.
presented in Tabla 1. For the trareverse direction of Schedule-J material,
the yield strength was 70,000 psi ir tension while it was 101,000 psi in com-
pression. Compression tests on the cther lot indicate that compression yield
strengths may be higher than 122,000 :si in some directions in some lots of
Zircaloy-2. In both cases, the yield data can be correlated with texture, the
direction of high compression yield strength being along a concentration of
Oil
basal poles in the texture. The datą also show that compression yield strengths
vary more with test direction than do the tensile yield strengths. When the
engineering use of the material requires that the structure be in a condition
of biaxial tensile stress, as in pressure vessels, the important property is
not that of the tensile yield strength in either tensile stress direction, but
the compression yield strength in the direction perpendicular to the plane of
stress.
Thus, if Schedule-8 Zircaloy-2 (Table I) were used to construct a pres-
sure vessel, the maximum design stress could be twice that permitted by conven-
tional practice using the tensile cata.
Single crystal specimens of cpt. metals can be used to illustrate and
emphasize some of the anisotropy of train behavior of the polycrystalline
material. A single crystal of beryll: um will shatter without detectable
plastic flow at a stress of about 280,000 ppi if it is compressed accurately
along the basal pole. 5 Even a few degrees of tilt between the compression axis
and the basal pole will allow the specimen to deform extensively by basal slip.
The single crystal of beryllium will defcrm plastically by (2012) twinning if
pulled in tension along the basal pole. Single crystals of magnesium will
crack with very little plastic flow if compressed along the basal pole at low
temperatures, but extensive flow will occur under the same loading conditions at
higher temperatures, 6 Single crystals of zinc will cleave on the basal plane at
Iow temperatures üz pulled in tension along the basal pole, yet will readily
flow by (1012) twinning in compression along the basal pole.? Single crystals
of titanium (Rei, 6) stad zirsonīm (Dei.) will flow extensively in eitzer
tension or compression along the basal' poile, Cleavage has not been positively
identified with any plane in these two metals.
Other examples of anisotropy could be cited.
Those presented show that
anisotropy is of major concern, that its magnitude depends on the material and
the application, and that it can be a significant advantage if properly used.
-
-
-
.
CRYSTAL STRUCTURE MODELS
. .
.
.
.
The picture of anisotropy in cph metals as so far presented would seem to
-
--
make little sense and the possibility of correlation of properties to anything
would seem doubtful. Yet, the diversity of behavior observed can be readily
exp].ained, in the qualitative sense, if one considers the deformation systems
that can operate in each of the materials and assumes the highly textured poly-
crystalline specimens to approximate single crystals fsingle-crystal analogue)
in their behavior. Little can be stated quantitatively at the present time,
for we are just beginning to find out the kinds of studies that must be made
-
-
----
so that the anisotropic stress-strain behavior of noncubic metals can be treated
-
quantitatively.
- -
-
-
It is quite helpful in a discussion of the various deformation systems
-
- - -
-
operating in cph metals to review the atomic arrangements on and near the
-
-
-
various crystallographic planes. While the drawing board can be on aid in
ORNI - AEC - OFFICIAL
trying to visualize the three-dimensional patterns of the crystallographic planes
permitting the excellent drawings of Rosenbaum, 10 ball models containing 3000
ORNI ~AIC - OFFICIAL
or more balls are more instructive. Such models are readily built using steel
balls 3/32 to 1/8 in. in diameter, plastic petri dishes, and "horteshoe"
.
permanent magnets.
The ferromagnetic steel balls stick together when placed in
the field of the magnet. The plastic petri dishes are placed over the pole faces
of the magnets to provide a continuous base and permit sume containment of the
balls when spills occur. A device for adding balls to the model is easily con-
structed from copper tubing. One end of the tubing is flared to make loading
easier and is closed with a rubber stopper. The other end of the tubing has two
slots cut from opposite sides, a bal}. diameter apart, and about 3/4 in. from the
end. A brass spring strip is shaped with tabs to enter the slots so that one
.
ball at a time is fed into the region between the slots and out the end of , the
tubing as the tabs are moved in and out of the slots. Balls can then be added
to the model as fast as the finger can be flicked. It is useful to color some
of the balls (heat tinting in a glass beaker over a Bunsen burner) so that
selected features of the crystallographic plane or direction can be emphasized.
With such ball models, the atomic arrangements usually drawn in two dimensions
can be seen in three dimensions, and the point-to-point movements of atoms
during slip in cph metals, as discussed by Rosenbaum, 10 can be more readily
followed.
A ball model of the fcc crystal structure is shown in Fig. 3. The fcc.
unit cell protrudes from the top of the model, showing (100) faces. Colored
bulls have been used in some places so that the shape of the model is more
readily seen in the photograph.
Ball models of the cph crystal structure are shown in Figs 4, 5, and 6,
delineating the atomic arrangements on several crystallographic planes. The
model shown in Fig. 4 points out that neighboring parallel (1010) planes do
ORNL - AEC - OFFICIAL
not have equivalent surroundings, and that iü is the corrugated sheet of atoms
that is reproduced in space in the cph structure. The pattern of arrangement
on the plane on the front face of the model above the fifth row from the bottom
18 that of a bent diamond shape while that below 18 of offset squeres. Viewed
from the side, it is quite apparent that the depth of corrugation of the two
ORXL - AEC - OSSICIAL
arrangements differs by a factor of two.
Yet the only difference in the model
is that the next layer of atoms was added in the bottom rows.
Colored balls are used to define some of the (1123) directions on (1011),
(1121), ani (1122) planes in Figs. 5 and 6. These figures show two views of the
same model and emphasize different features. The atomic arrangement in the
close-packed-hexagonal prism is shown on the left side of the (2011.) plane, and
the six-atom hexagon is defined in the basal plane. Also shown is the arrange-
ment of atoms on the (1072) twin plane, the atoms which truly constitute that
plane being colored.
· Twin models are easily constructed using the ball method. Two planes of
plastic are arranged to meet at the angle formed by the basal planes of the
respective components of the twin. Permanent magnets are taped to each of the
bases to provide a magnetic flux to hold the balls. The north poles of the
magnets should face each other, or else the balls will tend to bridge the
centerline of the twin and make model construction difficult.
Construction
of such models shows immediately the amounts of local strain existing in and
near the twin interfaces, and that "shuffling" of atoms must take place in
the region of the twin interface. The model also shows graphically that it
is difficult to define a true twin interface, the twin plane actually serving
only as a mirror plane for atom positions well away from the apparent interface.
The fact that incompressible spheres are used in constructing the models does
not constitute an appreciable error for the atoms chemselves are only slightly
RNL-AEC - OFFICIAL
Compressible.
--
OXML-AEC - OFFICIAL
ce
Defects such as vacancies, dislocations, and stacking faults are readily
built into the model; in fact, it is at times difficult to be certain that
faults have not been included. If the model is placed between the pole faces
of a larger magnet, it can be "deformed," allowing observation of the movement
of such defects. The model can only present the behavj.or of a "hard sphere"
solid, but it does constitute an observable and closer approximation of the
true behavior than has been realized before. The ball models are a good
arproximation of the structures of Zr, Ti, Hf, Mg, and Be, since their c/a
ratios differ from the value for spheres (1.633) by less than 4%.
Another kind of crystal model found to be quite useful is constructed from
hexagonal bar stock of an aluminum alloy. The pieces are cut to give the c/a
ratio of cph metal concerned, using the width of one face of the hexagonal
bar for the "a" dimension. Iwin planes are machined at the proper angle on
two pieces, and the pieces are joined by gluing. A set of twin models for
zirconium are shown in Fig. 7. Using several of each model in succession, the
changes occurring in the texture during consecutive twinning operations can
be readily worked out by stacking the faces of one twin model on another. In
addition, such twin models aid visualization of the deformation modes under
: various stress configurations.
DEFORMATION SYSTEMS
In cubic metals, studies have shown that slip is the important mode of
deformation in both tension and compression while twinning plays a small, but
sometimes vital, role. The multitude of slip planes and directions available,
ensures that one or more of the slip systems will be not too unfavorably
oriented for slip to occur. This does not say that slip will occur, for the
ORNL-AEC - OFFICIAL
.
.
.
. .
.
.
.
. .
11
stress required for slip in the material may be higher than that required for
ORNL - AEC - OFFICIAL
fracture. Since the principal mode of deformation is slip, the sense of the
stress (tensile or compressive) is of no irportance, and the yield strengths
in tension and compression differ but littl.e.
Such is not the case in the cph metals. While slip is observed, the slip
directions that have been positively identified are limited to the (1120) type.
The slip planes that have usually been observed in cph metals (depending on
the metal, among other things) are (1070), (0001), and (1071). Other planes
containing (1120) have been reported in some metals. There is a temperature
dependency for the critical resolved shear stress at each of the slip systems,
which is either ill-defined or has not been studied for most of the cph metais.
For example, slip in a (1120) direction on the (0001) plane is not observed
during the room temperature deformation of zirconium and Zircaloy-2, but it is
observedll and may be the primary slip system at 500°C. There are no known and
ven
proven cases of slip systems that will produce strain in a direction not con-
tained in the basal plane, except for one report of slip on (1122) in zinc (Ref. 12)
and one in zirconium. (Ref. 13) Isolated observations of complex slip markings
in individual grains have been made but could not be explained using the known
slip systems. Reed-Hill13 has reported as many as five sets of slip markings in
single grains of deformed polycrystalline zirconium.
Iwinning plays a much more important role in the deformation of cph metals
than it does in the cubic metals. The operation of twinning systems will produce
strain components in directions out of the basal plane. The twin planes reported
have been (1072), (1071), (1121), (1122), (1.173), and (1124). The twin that
operates in a given strain condition is dependent on the metal, the crystallo-
graphic c/a ratio, the temperature, and the restraint. If the c/a ratio is less
than that for spheres (1.633), (1072), and (1171) twins are usually formed in
ORNL - AEC - OFFI
.
12
ORHLAEC - OFFICIAL
tension along the basal pole and (1122) twins in compression. If the c/a ratio
is greater than 1.633, (1012) twins are formed in compression along the basal.
· pole. The behavior of the (1011), (1123), and (1124) twins are less well-
defineş, although (1124) has been reported to be a major twin mode in titanium.
(Ref. 14) However, no deformation system, slip or twin, operates to produce
strain in single-crystal beryllium compressed along the basal pole, 5 although
(1122) twinning would be expected because the c/a ratio is less than 1.633.'
Most studies of the deformation systems in cph metals have been conducted
by uniaxial tension and compression tests. There is no literature, giving data
on deformation systems operating in single crystal specimens of cph metals
tested under biaxial stress or in restraint (to enforce plane strain in pre-
determined directions). One would suspect that the systems would be different
under such circumstances, and that systems not observed during uniaxial testing
-
-
-
-
.
- - -
...
.
.. ..
..
.
..
-.---... .
testing of polycrystalline materials. · Reed-Hill21 has reported that basal
..
slip becomes a primary mode of deformation in polycrystalline zirconium at
.
temperatures near 500°C, producing kinking. He has also reported accommodation
..
strains produced by basal slip, at lower temperatures, at the ends of twins or
.
at the intersections of twins and of twins with grain boundaries. Such obser-
..
vations indicate that deformation systems requiring high stresses can be forced
- operate when other lower stress systems are unfavorably oriented to permit
the strain required. For example, consider a zirconium single crystal in the
shape of a cube, having as faces the (2002), (1070), and (1120) planes. The
cube is placed in an open-ended die with the (1120) planes against the walls,
the basal planes on top and bottom, and the (2010) planes facing the open ends
of the die. The specimen is then compressed along thé basal pole. What .'
- AEC - OFFICIAL
nimmt
man inte
min
· ...
... ... ..meron
13
deformation system can operate to produce the strain required? Twinning on the
(1122) plane is prevented, since it would have a strain component into the die
ORNL - AEC - OFFICIAL
walls. None of the known slip systenis can operate for all of the (1120) slip
directions are perpendicular to the compression axis.
One answer is postulated
by the argument below.
It is thought that there should exist in cph metals a perfect dislocation
having a (1123) Burger's vector. 15 Such a dislocation has not been experimentally
observed or reported, except possibly on (1171} planes in zirconiun. (Ref. 16)
However, slip on the (1122) plane in zinc (Ref. 12) and zirconium (Ref. 13) has
been observed. Rosenbaumł0 analyzed the atom movements that would be required
for slip on the (1122) plane and concluded that a two-layer fault is created by
what is in essence a twinning dislocation. These faults are glissile, form a
two-layer slip zone, and can be considered as a total zonal dislocation. The
resulting slip direction is (1123). A similar argument is applied for slip in
the (1123) on the (1071} plane. The same arguments would seem to apply to slip
in the (1123) direction on the (1171) plane as well. The atom arrangements for
some of the (1123) directions on the (1011), (1121), and (1122).planes were
shown in Figs. 5 and 6.
If slip in the (1123) direction on (1011), (11āl), or (1122) requires a .
critical resolved shear stress appreciably higher than either prism or basal
slip in zirconium and titanium, it is unlikely that such slip systems could be.
observed in single crystals tested in uniaxial tension or compression. The high
stresses required could only be reached if the plastic flow were restrained such
that no deformation mode known from uniaxial studies is permitted to operate,
Such a condition would hold for compression along the basal pole of the cube
specimen discussed above,
ORNI - AEC - OFF
· 14
OXHL - AEC - OFFICIAL
A further argument for the possibility of slip in the (1123) direction in
zirconium and titanium is that the rolling textures developed and their stability
to further rolling can not be explained using the known deformation systems. They
can be explained, however, if slip in the (1123) direction can occur. The
argument is presented in more detail later in the paper.
TEXTURES DEVELOPED IN POLYCRYSTALLINE CLOSE-PACKED-
HEXAGONAL METALS
Any practical engineering metal is polycrystalline and has some degree of
preferred orientation. ' If the part was made by casting, it has a freezing tex-
.
ture which depends on the freezing rate, the crystal structure, the mold design,
.
.
.
.
.
and numerous other factors.
If it is a fabricated shape, the deformation re-
...
.....
quired to share it has altered the freezing texture of the original casting,
--.
---
but a deformation and/or annealing texture remains. . If the metal has a
cubic crystal structure, a random texture after fabrication can be approached
E
for most practical purposes, so that the metal behaves in an essentially iso-
tropic fashion. If the metal has a noncubic crystal structure, near-random
textures are much more difficult, if not impossible, to obtain. In cubic
metals, the annealing texture can be, and really is, appreciably different
from the deformation texture. The annealing texture itself can be changed, in
many cases, by allowing secondary recrystallization to take place. In cph
metals, the annealir.g and deformation textures are usually related by about a
30° rotation of the prism poles about the basal pole. So far as the aniso-
tropy of mechanical properties are concerned, there is little difference in
· ORNL - AEC - OFFICIAL
the effects of the deformation and annealing textures.
It is only through a knowledge of the texture of the material that the
anisotropies of properties can be understood and correlated with fabrication
practice, stress-strain behavior, etc. It behooves every investigator studying
ORNI - AEC - OFFICIAL
imeinen
--. .
...
15
any property of polycrystalline metals to determine the texture of the specimens
he is using, especially when the crystal structure is not cubic. The value of
many otherwise excellent investigations has been appreciably lessened because
the textures of the specimen macorials were not determined and correlated with
the results.
Many investigators have reported the textures developed in various metals
and alloys by extrusion, rolling, swaging, drawing, and tube reducing. In only
a few instances have the properties been correlated with textures in the same
specimens. In even fewer were the fabrication procedures reported in sufficient
S
.
detail to allow estimation of the effects of fabrication variables. Seemingly
minor changes in fabrication procedure and equipment can cause appreciable
differences in the textures observed. In one 'study, 17 it was found that the ingot
axis orientation during ingot breakdown could determine the final texture in
1/4-in.-thick Zircaloy-2 plate, when the ingot sizes were 12. in. in diameter by
38 in. long and the fabrication procedures were otherwise identical. If the
ingot axis was, always in the rolling direction, the basal poles of the grains
When the ingot axis was turned to the transverse direction after an extension
from 38 to 52 in. in length, the basal poles were concentrated in the transverse
direction to an equal intensity. It is quite difficult to understand the dif-
ference in texture, for both lots were rolled from 12-in. diam ingot to l-in.-
thick plate at temperatures where the Zircaloy-2 was entirely bcc in crystal
structure (beta phase). Fabrication of new lots of 1000-1b ingots confirmed
the results.
In general, the textures developed in polycrystalline cph metals during
rolling are such that basal poles are concentrated near the normal to the
rolling plane. The principal difference between the textures for Zr, Ti, Be,
..
.
....
...
............
.
..
...,
.
.m
oremom. ......................
16
.
17
OANI AEC - OFFICIAL
and Mg is the spread of basal poles about the normal direction. In magnesium, i ja
(Ref. 18) the spread tends to be greatest in the rolling direction while it is
greatest in the transverse direction in beryllium,' (Ref. 19) titanium, (Ref. 20)
and zirconium. (Ref. 17) If the total reductions are of the order of 98% or ..
more, the basal pole concentrations approach a position intermediate between
the normal and rolling directions in magnesium (Ref. 21) and between the
normal and transverse directions in beryllium,' (Ref. 19) zirconium, (Ref. 22)
and titanium. (Ref. 23) A typical pole figure for Zircaloy-2 is shown in Fig. 8.
The textures observed in magnesium can be explained in terms of basal slip but
those observed in beryllium, zirconium, and titanium can not.
. The textures developed in cply metals during the extrusion and subsequent
fabrication of tubing have been studied only in zirconium-base alloys. Several
reports24-26 present data that show that texture can be one with basal poles
concentrated in the radial direction, in the tangential direction, in both, or
in the 45° position between the radial and tangential directions.
Cheadle and
Ells24 have shown that basal poles are found in the axial direction as well as
the tangential direction in heat-treated Zr-2.5% Nb alloy tubing but not in
i
heat-treated Zircaloy-2.
The textures produced in tubing during fabrication are thought to be
dependent on the details of the fabrication procedures used, including types of
· equipment, extrusion ratios, die angles, mandrel dimensions and angles, recluc-
tion per pass, number of passes between anneals, changes in wall thickness vs
circumferential reduction in each operation.
Though of great importance, Jittle
useful data have been obtained on these effects, for the commercial manufacturers
are to utilize texture to advantage during mill fabrication and in the final
NL-AEC - OFFICIAL
application.
.
..-.--. .
. .
. .
.
.
...
..
.
..
-...
.
..
.
...
.--'
-
17,
-
.
-
•-
--
ORNL - AEC - OFFICIAL
-
-
-----
One effect on texture development in cph metals that has received essen-
tially no specific study is that of alloying additions. If an alloying
addition can change the stress level' at which a deformation system operates
under given conditions or temperature and stress configuration, then the tex-
tures observed in the final products should be different in the various alloys
for the same fabrication procedure. Several experimental studies27,23 have
shown such behavior, though the experimental programs were not designed to
ms were
examine this effect, and the investigators did not comment on it.
The textures developed in cph metals by fabrication methods other than
rolling can be rationalized and estimated from a knowledge of the rolling
textures if an approximate plane strain condition exists during fabrication, and
the plane of strain can be defined. In rolling, a compressive strain is produced
in the normal direction and a tensile strain in the rolling direction. No
.
...
strain occurs in the transverse direction if the width of the plate or sheet
.-
.
-
is many times greater than its thickness.
The plane of strain is defined by the
.-
-
rolling and normal directions and is perpendicular to the transverse direction.
In swaging, two planes of strain can be defined. The first contains the radial
(compressive) and axial (tensile) directions; the second is formed by the tan-
gential (compressive) and axial (tensile) directions. The rolling texture
usually observed in moderately deformed cph metals is one with a concentration
of basal poles around the compressive strain axis and a mixture of prism
directions, (1010) and (1120), in the tensile strain axis direction. For the
first "plane of strain" case in swaging, the prism directions should be concen-
trated in the axial (tensile) direction, and the basal poles concentrated in
the radial (compressive) direction. For the second "plane of 'strain," the
prism poles should be concentrated in the axial (tensile) direction and the
basal poles in the tangential (compressive) direction. A mixture of the two
morno porno
18
"planes of strain," as would actually occur in swaging, should produce only a
ORAL-AEC - OISICIAL
concentration of prism poles in the axial direction. The experimentally
observed textures are "fiber textures," laving (1070) directions concentrated
ORNI - AC - OFFICIAL
in the axial direction. Pole concentrations in the radial and tangential
directions are mixed to nearly random.
Similar analyses can be made of textures in tubing made by other methods
if the "planes of strain" can be defined. In tubing being reduced in outside
ūiameter without an appreciable decrease in wall thickness, the "plane of
strain" is defined by the tangential (compressive) and axial (tensile) directions.
From the rolling textures observed, the final texture should consist of a concen-
tration of basal poles in the tangential directions and prism poles in the axial
directions. This texture is commonly observed in Zircaloy-2. If the wall
thickness is being reduced at the siame time, or immediately after part of the
circumferential reduction, there should also be some concentration of basal
poles in the radial direction. This texture is frequently observed in tube-
reduced Zircaloy-2. It is common practice in tube drawing, swaging, or tube
reducing for the tube to receive a circumferential reduction before the mandrel
is encountered and wall thinning beisins.
STABILITY OF THE TEX! URE TO FURTHER DEFORMATION
One of the goals of present research or iexture and anisotropy is the con-
trol of the development of texture curing mill fabrication so that mill products
having specified textures can be reproducibly and economically produced. To do
this, we must understand how textures are developed in terms of the strain paths
and fabrication sequences undergone by the material.
-
-
-
-
-
- AEC - OFFICIAL
To
ORNI NEC - OFFICIAL
A problem inherent in such understanding is that of how and why certain
textures are stable to further deformation of the kind that produced them.
The rolling textures observed in the cph metals, once formed, are not screciably
changed by further rolling. Any changes that occur are usually only a sharpening
of the texture, rather than a change in orientation relative to the fabrication
directions. Annealing also produces a sharpening of the texture and usually: .
interchanges the (1010) and (1120) directions relative to the rolling direction.
In terms of the deformation systems observed in single crystals and in poly-
crystalline specimens tested in uniaxial tension and compression, the only
deformation systems that can operate for a compression axis along or near the
basal pole are basal slip and/or (1122) twinning. Basal slip can be used to
explain the stability of the rolling textures in magnesium but it can not in
• beryllium, zirconium and titanium. For basal slip to be the primary mode of
deformation during rolling, a texture must develop in which the basal poles are
concentrated near the normal direction and are spread toward the rolling direc-
tion.
This does not occur in beryllium, zirconium, and titanium, the basal
poles moving toward the transverse directions instead.
A long sequence of successive twinning operations can be developed that will
UCCE
allow the final texture to be essentially the same as the starting texture in
Zircaloy-2.
This sequential twinning involves all of the known modes in various
orders and requires more than six successive twinning operations to reach the
starting point on the pole figure. It is difficult to believe that such succes-
sive twinning is actually the case, although specific sequences can be guessed
to explain the stability of most of the texture observed in Zircaloy-2.
www.mivim prostore
.co
20
A much simpler explanation of both the development and the stability of the
ORAL-IC-OFFICIAL
ONN-AEC - OFFICIAL
rolling texture in Zircaloy-2 can be made if slip can occur in (1123) directions.
As previously discussed, this direction occurs on (1010), (1011), (1171), and
(1122) planes, and cross slip should be possible between the slip systems. ...
These slip systems would produce a strain component out of the bacal plane. No
large changes in basal pole orientations are required when these slip systems
operate, and the developed texture should remain essentially the same on further
rolling, as is observed.
Such slip systems can explain the development and stability of texture even
in the most severely rolled zirconium foils. In these foils, the basal poles
are concentrated at about 50° from the sheet normal toward the transverse direc-
tion with little spread in either the transverse or rolling directions, 22 In
this case, either the (1071) or (1171), or both, are the slip planes. The
(1122) can not be a slip plane here since the basal poles would then be located
toward the rolling direction rather than the transverse direction.
In polycrystalline materials, the strain produced in one grain by an
applied load must be accommodated by strain in its neighbors. Similarly, the
strain in the neighboring grains must be accommodated by the strain in the first
grain. With such accommodation required, it would be relatively simple for local
stresses to rise high enough to force the operation of high-stress deformation
wystems in preference to the lower stress systems observed in uniaxially loaded
single crystals. Similar restraints are placed on the polycrystalline bodies
-,.
.
0
when they are deformed in commercial mill equipment. There appears to be no
.
valid reason for requiring polycrystalline bodies subjected to restraint to
deform by the systems observed in uniaxial tests. Of course, the explanation,
understanding, and control of the development of texture could be much simpler
and easier if they did.
41 - AEC - OFFICIAL
in
21
It is quite tempting, and frequently done, to attempt to analyze the
development and stability of textures from a determination or estimation of the
principal stresses in the body. Attempts have been made to analyze the behavior
OKHL - AEC - OFFICIAL
of a material during rolling, using stress analysis and the mechanical properties
as determined by uniaxial tensile testing. 29 Such attempts may be reasonably
successful för materials that are cubic in crystal structure, when the primary :
mode of deformation is slip and when the sign of the stress does not determine
the operating mode of deformation. Such analyses must invariably fail when the
material is a cph metal, for no known theory of the relationships between stress
and strain can make allowances for a change in deformation mode with a change
in the sign of the stress.
An analysis of the strain path during deformation will at least qualitatively
explain the behavior observed. In rolling, the material is deformed in plane
strain only, provided that the thickness is much less than the width. . The
amount of compressive strain at any point in the material between the rolls,
plus its transport in space, is approximately proportional to the distance from
the centerline. Assuming that the volume remains constant during plastic flow,
the tensile strain in the rolling direction of a unit cube of the material must
be equal to the compressive strain in the normal direction, there being no
transverse strain in this example. Material is forced or drawn into the center-
line as the plate is rolled, for otherwise there could be no lengthening of the
sheet at the center of the thickness. This complicates the analysis and, at the
present time, prevents a quantitative explanation. Nevertheless, the change in
orientation and the probable deformation system operating can be qualitatively
determined for any grain located away from the centerline under the conditions
of rolling. The change in shape of any small unit of volume can be approxi-
mately determined by the defined strain path. If the grain orientation is
ORNL - AEC - OFFICIAL
Kaar
.22
then stated in reference to the strain axes, the possible deformation modes can
be examined and the possible changes in crystallographic orientations of the
OEHL-AIC-OISICI:?
body determined.
Such approximate analyses can be made for material deformed in any commercial
mill equipment if the strain path forced by the die design is known. Conversely,
it seems feasible to design dies for mill equipment to produce specified strain
paths that, in turn, would produce specific textures in the material. If the
deformation modes operating under various conditions of restraint are known for
the materia.., he problem of producing specific textures in the particular material
shape becomes one of determining the strain path and the sequences of strain paths
that must be undergone.
Another problem involved in the understanding of texture development and
stability is that of the effects of a change in the strain path. For instance,
what are the effects of cross-rolling on the final texture? The strain paths are
of the same form but are rotated about the normal direction by 90°. One answer
is that the texture must also be rotated by 90° if there is enough cross-rolling.
How much is enough? What is the change in texture when the amount of cross-
rolling is small? We must obtain detailed quantitative answers to such
questions if we are to control texture during fabrication.
ANISOTROPY OF YIELDING AND PLASTIC FLOW
The anisotropies of yield strengths and plastic flow are of principal
interest in the use of materials in most structures. If we are to use anic stropy
to advantage in the design of structures, we must have a thorough knowledge of
.
the way that anisotropy is influenced by various stress configurations.
',.
.'nin 4., sinni....
At the present time, there is no simple and generally accepted test by
which the three-dimensional anisotropy of a material can be characterized, and,
ORNI-AEC-OISICIAL
23
thereby, its behavior under any conditions of application be predictable. All
presently i:sed tests have been designed to examine variables that have, at one
time or another, been shown to be correlatable to service failures. Through
experience, design engineers have learned to use the proportional limit, the
0.2% offset yield strength, or the tensile strength to set maximum permissible
design stresses for a material in any part of a structure.
The design stress is
usually obtained by dividing the evaluation stress by a factor of two to five,
depending on the evaluation stress used, the service life expected, the dangers
likely to be encountered in case of a service failure, and the ductility
evaluation used.
An important material requirement, particularly for steels, is the amount
of energy absorbed in an impact energy test at some specified temperature. The
value is not used in the design of the structure, but it is used as a measure
of "toughness" or "notch sensitivity" to evaluate the tendency for brittle
crack propagation.
One can not expect design criteria established for cubic metals to be safely
.
and properly usable with noncubic metals, The symmetry which permits the extra-
polation of a set of properties obtained in one test direction to any other
direction in cubic metals does not exist in noncubic metals. There is no basis
for expecting the conventional laboratory tests and design criteria to be usable
for cph metals. We must determine what tests and criteria are valid.
The ASTM tensile tests specify round or rectangular cross-section bars
having a reduced gage section between end grips.30 The specimens are pulled in
uniaxial tension at a constant heaä rate and a continuous load-elongation or
load-head movement curve is recorded throughout the test. Yield, maximum, and
fracture ioads are determined, and yield, tensile (or ultimate), and fracture
OUNT - AEC - OFFICIAL
T
.
...
..
.
.
.
...
...
.....
..
ORNI ACC - OFFICIAL .
strengths are calculated. Uniform and total elongations between gage marks
ORHL AEC - OFFIC!,!.
and the reduction of area at the fracture are determined. In some cases, par-
ticularly for characterizing the anisotropic behavior in deep-drawing, the
wiath and thickness strains are measured to permit calculation of an "R" value.
In most cases, both rolling- and transverse-direction specimens are used,
although specimens having tensile axes at 45° to both the rolling and transverse
directions are sometimes included.
A very important factor usually ignored in tensile testing is the influence
of specimen size and configuration on the tensile properties measured.
The
SS
investigation by Miklowitz31on the stress and strain behavior of flat tensile
bars of one heat of a killed steel resulted in the following conclusions:
(1) a minimum cross section is produced in the specimen at the beginning of uni-
form strain (i.e., just after yielding) and this section remained the minimum
throughout the test to fracture; (2) the average width and thickness strains,
and their ratio (the "R" value) were strongly influenced by the width-to-
thickness ratio (W/T) and the cross-sectional area of the unstrained specimen,
the amount of total strain that had occurred, and the position on the gage sec-
tion at which they were measured; (3) the local ratio of width-to-thickness
strains varied across the width at all axial positions along the specimen at
all axial strains; (4) the yield, tensile, and fracture strengths and the
elongations varied appreciably with W/T; and (5) head restraint influenced the
local and average strain behavior during uniform straining and during neck for-
mation. He also showed the progressive development of the "necking cross"
which Nadai 32 has shown theoretically should happen during neck formation in a
sheet-tensi.e bar of an isotropic material. The necking cross is formed by:
highly localized increases in the thickness strain due to the configuration of
stresses in the necked region. If the complete cross is not formed, the
.'..'
-.
..;
-
..
......
-
ORNI - AEC - OFFICIAL
......
ORNI - AEC - OFFICIAL
.
.
.
..
ini.
25
ORNL - AEC-OPFICIAL
specimen usually fails by shear along one arm of the cross, 31 This careful
investigation clearly shows that even in the isotropic cubic materials, the
specimen configuration strongly influences the value of the tensile property
measured and, consequer.tly, such specimens and property values can not provide
information characteristics of the true three-dimensional anisotropy of the
material, free of all testing effects,
Similar effects were found in sheet tensile specimens vf magnesium alloys
by Avery, Hosford, and Backofen.27 They, however, hela specimen size and con-
figuration essentially constant and varied composition and fabrication history
in the materials to vary the preferred orientation present in the test specimens.
They found that the anisotropy parameter of width-to-thickness strain ratio ("R")
was variable with test direction, strain in testing, processing history, and
alloy composition.
They also showed that the elliptical yield locus curves
for isotropic materials in biaxial stress are distorted and usually elongated in
anisotropic magnesium alloy sheet, and that the yield strength in biaxial tension
could be 50% greater than would be predicted by present ſtheory from the uniaxial
tensile tests.
Recent work on Zircaloy-2 sheet tensile specimens“ has shown that the "R"
values are strongly influenced by W/T and by the total axial strain at which
the "R" values are measured. The contractile strain was primarily in the width
direction, very little occurring in
thickness direction. A schematic
drawing of the test is presented in Fig. 9a along with the single-crystal
analogue of the texture present in the specimen.
Thickness strains in such a
U
specimen require compressive strains 7ong the basal pole of the cph prism. This
strain can be accomplished by (1122) twinning (of the snown deformation systems)
Dii -
and compression tests of highly textured specimens have shown that the compression
...vidiu
26
OXNL-AEC - OFFICIAL
yield strengths are 120,000 psi or more. 1 The tensile yield strength in the
transverse direction in a sheet-type specimen of the material is 48,500 psi and
that in the rolling direction is 45,400 psi. The specimen shown in Fig. 9a
deforms, primarily by (2010) slip, causing the specimen to narrow in the width
1330 - SIVKO
direction. Despite work hardening, the true stress in the necked region did
not rise high enough to cause flow in the thickness direction except at the
fracture.
If such a material ,were to be used to construct a pressure vessel by conven-
tional steel design practice, the 0.2% offset yield strength or the tensile
strength as determined with a sheet-tensile specimen would be used to determine
the maximum permissible design stress. The material in actual use would be
subjected to biaxial tensile stresses in the surface of the vessel. Yielding
in biaxial tension would require that the material thin in the thickness direc-
tion. Thus, the important yield strength for this application is the compression
yield strength in the thickness direction, ~120,000 psi, not the uniaxial ten-
sile yield strength of 45,400 psi. The maximum permissible design stress could
be increased by more than 50%, lowering the thickness requirement for the
material to be used in the vessel. Tube burst tests of Zircaloy-2 tubing of
somewhat poorer texture for resisting the applied load confirm the conclusion, 33 :
In the above case, conventional design practice allows a large and unknown
safety factor to become incorporated into the design. If the texture were
Pre
different (as it can be), the vessel might be dangerously underdesigned. con-
sider a Zircaloy-2 sheet tensile specimen having a texture with basal poles
.
..
.
....
.
........
concentrated in the axial direction, as shown in Fig. 9b. Tensile straining
will cause plastic flow in both width and thickness directions at a yield
strength of about 70,000 psi (see Table I). Deformation must first occur by
.
..
O&ML-TEC - OFFICIAL
171):sic - Dit-inju
(1012) twinning since the slip systems are not favorably oriented. Continued
deformation in the twinned material is probably by prism and/or basal slip
since the flow stresses do not approach 100.000 psi.? The compression yield .
strength in the thickness direction of such a specimen is 72,500 psi (Schedule-
J material, Table I).' Here, it seems the designer is safe in using the experi-
mentally determined tensile yield strength for setting the design stress. Yet,
if the material were used to construct a cylindrical pressure vessel such that
the basal poles were concentrated in the circumferential direction of the cylin-
der (a commonly observed texture in Zircaloy-2 pressure tubing), the vessel
could be quite unsafe, as will be shown. The biaxial stress condition would be
that shown in Fig. 10c at point G (a = 2). The yield'loci in Fig. 10 are based
on the known tensile and compression yield strengths of highly textured
Zircaloy-2 and the assumption that the yield loci are elliptical (if the octa-
hedral shear stress theory can be extended to apply). Now consider that a
blister is forming in the wall away from the ends of the cylinder. As the
blister forms, the stress condition in the wall of the blister approaches that
for a spherical vessel under internal pressure, that is, at point H (Q = .) in
Fig. 10c. The yield locus for this material texture under such conditions shows
SS
.
.
that, from the stress configuration alone, the stress required for yielding
decreases as the blister forms. The blister will form catastrophically unless
work hardening rapidly increases the stress required for flow. Sheet-tensile
tests for material of this texture sirow, however, that the work-hardening
coefficient is small.
Blister formation has been a problem in Zirca.loy-2 fuel;
sheathing and may become a problem in pressure tubes of Zircaloy-2 if the design
.: stre's ses are increased
........omne ou asem
..
28
-
ORHL- AEC - OFFICIAL
SU
ORNLÖAEC - OFFICIAL
A similar analysis for pressure tubing having basal poles concentrated in
the radial direction, Fig. 10d, shows that the yielding of flow stresses required
increases rapidly. (G to H) as the blister forms. This flow stress increase is
due to the biaxial stress configuration and does not consider the added bonus
of work hardening. Thus, blister formation is resisted by such a texture. In
this case, the maximum permissible design stress could be safely increased since
the compressive yield strength is 122,000 psi (Schedule-8 material, Table I). . .
A recently developed analysis of anisotropic strain behavior in Zircaloy-2
may provide a better measure of the anisotropy characteristic of the material., 4,17 :
requiring data from only two round tensile specimens and three cylindrical com-' ;
pression specimens. The study has shown that the contractile strains in round
tensile specimens are‘linearly proportional (proportionality constant = kx ;) to
COU
the axial tensile strain at all levels of axial strain from the yield point to
the fracture.
The proportionality constants for the compression specimens (k_3x)
were found to fit the simple equation
.
.
.
where the first subscript of each k denotes the axial strain axis and direction
(minus denotes compressive strain along the cylinder axis) and the second denotes
the direction in which the diametral strain is measured. The rolling, trans-
verse, and normal fabrication directions are taken as the x, y, and z. coordinate
axes, respectively. It can be shown that Hill's. theory of anisotropy34 yields
the same equation for the tensile specimens, and it has been assumed that such
an equation can be applied to the Zircaloy-2 round tensile specimens as well.
Thus, using this equation and assuming constancy of volume during plastic flow,
measurements of the contractile strains and yield strengths of two tensile
Le
DRNE - AEC - OFFICIAL
Specimens and three compression specimens is believed sufficient to provide a
29
OCHI-MIC-OFFICIAL
complete characterization of the three-dimensional anisotropy in the material.
The strain and yield stress measurements can be correlated with texture in more
than 20 lots of Zircaloy-2, having a wide variety of orientations and perfections
of texture. The reader is referred to the two papers referenced for complete
details of the method and the data, as both are too long for summarization
here.
It is not yet known how well such uniaxial data can be used to quantitatively
describe the biaxial stress behavior, but the qualitative description is excellent.
The necking cross discussed previously has not been observed in unirradiated
sheet-tensile specimens of d-zirconium and a-titanium alloys, regardless of
the width-to-thickness ratio.
It has been observed in some ring-type tensile
specimens cut from Zircaloy-2 tubing.35 and in Zircaloy-2 sheet-tensile bars 36
irradiated to about 1021 neutrons/cm2. When the necking cross occurs, a plane
strain condition exists in the region of the cross. The tensile strain occurs
at about 50° to the applied tensile load axis and the contractile strain occurs
in the thickness direction. All of the strain in the specimen is confined to
the arms and center of the necking cross. Generally, a plane strain condition
also exists in sheet-tensile specimens of d-zirconium and Q-titanium alloys
(and many textures in magnesium alloys as well) but the plane of strain is formed
by the tensile strain in the axial direction and the contractile strain in the
width direction.
There is essentially no thickness strain outside the fracture
region.
Thus, the plane of strain during necking in isotropic materials is 90°
from the plane of strain during recking in the anisotropic materials. Since
the strain behavior in uniaxial tension is very different in the two cases, the
test criteria used in setting design limits will have to be appreciably different.
Sanderson 35 has studied the development of necking crosses with increased
width-to-thickness ratios (w/T) in ring-type specimens of Zircaloy-2 and has
ORNI - AEC - OFFICIAL
Tie w model
wanneer
ermord:
.quor
um promem mais
em
30
ORHI - AC - OFFICIAL
observed well-developed crosses for w/T = 6 or more. The ring-type specimen,
used to determine the stress-strain properties of tubing in the tangential
direction, is produced by making transverse cuts in tubing, and is tested by
pulling the r:ing apart with two semicircular heads in a tensile machine.
ORNL-AEC - OFFIC!!!
Since the separation of the two sericircular heads is the same at the start of
each test, the "effective length" of each test specimen is the same for a given
diameter of tubing, independent of the w/T used as a variable in the study. 'In
effect, the width-to-effective length (w/L) ratio between the grip ends of the
specimens was varied at the same time that the W/T was being varied. It appears
that the necking crosses observed in the ring-type specimens were more likely
to have been caused by head restraint than by „W/T variations, especially since
necking crosses have not been observadt in unirradiated sheet tensile specimens
having w/T = 10. .
The observation of necking crosses in Zircaloy-2 sheet-tensile specimens
irradiated to 1021 neutrons/cm², but not in the unirradiated controls, 36 indi-
cates that the modes of deformation cperating in the two cases may be different.
no
RSS
Radiation damage could harden the normal deformation systems sufficiently to
force other systems into operation. In Zircaloy-2 sheet specimens having the
usual texture, prism slip coula ba so barieved that (1122) tvirning could be
2orcsi is scout to grciüce strains in the winess cracion anti e time
irradiated specimens woüiù behave as "isotropic" material in the necking .
behavior. Another possibility is that once some local strain has occurred,
further strain is confined to that region by "work softening"; that is, the
radiation damage is at least partially removed by the lattice rearrangements
occurring during plastic flow.
RNL-NEC-OFFICIAL
ORAL - AEC - OFFICIAL
..................
.
..
.
.
31
CRCEP
OKNL - AEC - OFFICIAL
Most of the studies of creep in cph metals have been concerned with
Q-titanium (Ref. 37) alloys and with the Zircaloys. (Ref. 38) These studies
have usually employed uniaxially loaded specimens of sheet-type design. The
strain behavior observed in the specimens has been much like that of the tensile
specimens, contractile strains occurring in the wiath direction but little strain
occurring in the thickness direction.
The investigators have seemed, for the
most part, to be unaware of the preferred orientation existing in the specimens
and of its importance. Specimen axes have been rimarily in the rolling
direction of sheet and strip and in the axial direction in tubing,
Typical observations at temperatures of 300°C and lower are that there is
a large elongation during loading and little or no further extension until the
tensile yield strength is exceeded. At higher temperatures, second- and third-
stage creep occur at stress levels below the tensile yield strength but well
above the conventional design stresses.
Reed-Hill39 has reported that creep has been observed at room temperature
in transverse-direction specimens dead-ioaded to stress levels below the tensile
proportional limit. The creep deformation occurred by both prism slip and
(1072) twinning. The texture of the material was one with basal poles concen-
trated in a plane containing the normal and transverse directions, forming a
ORNL - AEC - OFFICIAL
fiber texture around the rolling direction.
Only one investigation of creep in biaxial stress has been reported to my
knowledge. Aungst40 has conducted accelerated creep and stress-rupture tests
with biaxial stress at room temperature and at 300°C in Zircaloy-2 tubing in
the following conditions: (1) as-received, usually with 15 to 20% cold work;
(2) autoclaved; and (3) autoclaved and irradiated to various levels of neutron
doses, He found that the creep rates in biaxial tension were much lower than
would be predicted from uniaxial tests, that the autoclaving treatment decreased
32
creep rates by factors of four to five, and that irradiation further decreased
OXN1-AEC - Oificial.
the creep rates in each lot of Zircaloy-2. He found that 50,000 psi was
required to obtain the same creep rate in biaxial stress that 30,000 psi pro-
duced in the uniaxially loaded specimens of the same material. The tightly
adhering oxide films produced on the surface of the specimens by the auto-
claving treatment probably interfered with the generation of dislocations at
the surface, while the radiation damage very likely increased the number of
pinning sites for and decreased the mobility of moving dislocations. The much
higher creep rates observed in the uniaxially loaded specimens illustrate the
very great importance in the design of the experimental study of the anisotropy
of strain behavior and its relationship to the texture present in the material.
This study also illustrates the danger of applying criteria and tests developed
for cubic metals to the engineering use of cph metals.
FRACTURE BEHAVIOR AND FRACTURE MECHANISMS
The problems of fracture behavior and fracture mechanisms in cph metais have
not been studied to the extent such phenomena have in the cubic metals. There
are several types of fracture behavior to be discussed. Among these are:
(1) fracture by cleavage; (2) tensile fracture and its variation with test
temperature and strain rate; (3) impact fracture, as determined by the several
types of impact and shock loading tests; (4) crack propagation as a function of
stress configuration, loading rate, and test temperature; (5) fatigue; (6) stress-
rupture, by which a dead-loaded specimen fails with little to no ductile strain;
and (7) stress-corrosion cracking, in which the corrosive attack by the environ-
ment is localized and accelerated by stress concentration.
ORNI - AEC - OFFICIAL
Fracture by cleavage at small total elongations has not been observed in
0-zirconium and Q-titanium alloys. It occurs readily. in zinc but not cadmium
0:11 - AEC - OFFICIAL
when stressed in tension along the basal pole. ilowever, cleavaga, resulting
in microcrack formation but not fracture, has been observed in antitanium in
local regions of high strain and high stress concentration+1 at temperatures
below -250°E.
The microcracks were observed to form only in and around
(1122) twins that had undergone second-order twinning. These microcracks were
of three types: (1) a primary twin - matrix boundary crack; (2) a primary
twin – second-order twin boundary crack; and (3) a second-order twin - matrix
boundary crack. A ductile-to-brittle fracture transition was coincident with
the observation of second-order twinning in the (1122} twins in coarse-grained
material but was not observed in fine-grained material in which second-order
twinning did not occur. Microcracks have been observed in a-zirconium alloys42
only in hydride particles or at the interface between hydride particles and
the matrix,
The tensile fracture behavior of a-zirconium and a-titanium has invariably
been of the "ductile" type, even in irradiated materials that have low total
elongations.43 Typical "brittle fracture" behavior has been observed only in
alloys containing large amounts of hydrogen as precipitated hydrides or in high
supersaturation (such as produced by electrolytic charging) or in alloys of
lower hydrogen content in which a large proportion of the hydride plates have
been oriented normal to the tensile stress. 44
Both metals show very large
:
,
ductilities at all temperature
materials in which twinning is a major
deformation mode. . . .
Tensile fracture at low total elongation produced in magnesium and several
of its alloys has been shown to be the esults of confinement of strain, pro-
duced by slip, to within twins formed in the very early stages of plastic
ORNI-AEC - OFFICIAL
metro mimi sinabi
....
......
.
..
34
flow.45 Such behavior is quite sensitive to the orientation of the tensile
stress relative to the texture present in the material, and specimens
oriented so that all plastic flow can be accomplished by basal slip. show large
0581 - ACC-OSSICI:!
elongations and reductions of area.
The fracture behavior and energy absorbed during impact testing of notched
bars, such as the Charpy V-notch, have been used for determining crack or notch
sensitivity in bcc metals, particularly the structural steels. Typically, a
transition from low to high energy absorption is onserved to occur over a smali
temperature range. Usually there is simultaneously a transition in fracture
appearance from that of brittle cleavage to ductile tearing. Below the transi-
tion temperature, cracks will propagate rapidly at a low stress level and with
low-energy absorption from "notches" and stress concentrators such as arc
strikes at the start of an arc wela, deep machining marks, and unfilleted roots..
Above the transition temperature, the crack is stopped by ductile deformation
of the matrix at the tip of the crack with large energy absorption.
Charpy and subsize Izod impact tests, using "y-notch" specimens, have been
conducted on Q-zirconium and Q-titanium alloys. 46 While the transition in enersy
absorption may occur over a narrow temperature range, the transition in fracture
appearance is observed only in alloys containing quite large amounts of hydrogen.
Regardless of the amount of energy absorbed, the fracture has the rough torn sur-
face characteristic of a "ductile fracture." Studies on the effects of notch
orientation relative to the texture present in several lots of Zircaloy-2 have
shown that the "transition temperature" for a given lot depends on the orienta-
tion of the notch relative to the texture. 3 Notches oriented so as to permit
prism slip to occur at the root of the notch absorb much more energy and have
lower "transition temperatures" than do specinens with notches oriented so that
twinning must occur. The same material can have a low or high transition tempera-
ture depending entirely on ühię noich orientacion usei :.. wie test. This
Oftit - hit-01116iXL
A
V
35
C
behavior is shown in Fig. I for a lot of Zircaloy-2 in which the basal poles
'! - HTC - OFFICIAL
are concentrated around the normal direction of the plate.
Drop-weight tests have been used in steels to determine another criterion
of notch and impact sensitivity, the "nil-ductility temperature."
In a typical
test, a brittle welament is made on the back side of a beam-l.oaded specimen, a
thin groove or notch is cut into the brittle wela material, and a known weight
is dropped from a measured height onto the face of the specimen. The specimen
easi
is allowed to bend only through an angle of a few degrees, and the temperature
at which the crack in the weld metal iust propagates throughout the specimen is
determined. Such tests have been conducted on a number of commercial a-titanium
illoys and on Zircaloy-2. (Ref. 47) Il a "nil-ductility teriperature" exists for
tese materials, it is beiow -100°C. În no case was the crack observed to
2.
propiate outside the heat-aitected zone oỉ the embrittling weldment.
Another type of crack propagation test cas been studied in Zircaloy-2
tubing. 48
It is the determination of the minimum crack length, due to an inten-
tional defect, which will propagate the entire length of the tube during burst
tiesting.
Ingenious methods have been devised to ensure sharp cracks and to
f
wa
permit testing oí irradiated tubing. One of the methods involves firing a
sharp-edged projectile into a pressurized tube. Derects less than 0.4 in. long
had no effect on the ultimate strength, and "brittle fracture" was observed only
with defects longer than 1 in. Increased hydrogen content increased the sensi-
tivity to defect length, cold-work levels up to 50% had only a small effect, and
irradiation about 1020 neutrons/cm2 (highest testea) had no effect on the fracture
behavior at room temperature. In the projectile tests, brittle failure, (both by
fracture appearance and propagation length) was observed only in those specimens *
pressurized to the yield point (98% or ultimate) at the moment of impact.
DENI-if-Errn
36
Few studies of fatigue failure have been periormed for cph metals.
Those
Gini - 45C - OFFICIAL
conducted indicate that fatigue failure is not the serious problem in cph
Ome
.
metals that it is in some of the bcc metals. Large stress and strain amplitudes
are required to produce fatigue failure at room temperature and above in a-
zirconium (Ref. 49) and Q-titanium (Refs. 50,51) alloys. Reed-Hill et al. 52
have shown that the presence of large numbers of (1121) twins in d-zirconium
(produced by deformation at -196°C) will permit a large energy absorption during
cyclic tensile loading by the growth and shrinkage of the (1121) twins. This
may be a nixed blessing since Roberts and Partridge 53 have shown fatigue failure
in magnesium at room temperature by the formation and rapid linkage of small
holes in the vicinity of a (1072) twin boundary on the surface of the specimen.
They ascribed the hole formation to selective surface corrosion in regions of
.....
.
....
.
high dislocation density caused by a complex interaction between basal slip dis-
locations and the moving twin boundaries. There is also the possibility that
jogs produced by interaction of the dislocations with the twin interface could
give rise to a large concentration of point defects, which could then condense
--
-,
into a crack.
-
-
Fracture by stress-rupture is characterized by the sudden, low-ductility ·
- -
-
.
..
failure of dead-loaded specimens stressed below the yield point. Stress-
..
.
.
.
rupture tests have been coniucted on Zircaloy-2. tubing specimens40 with no
..
..
4
- -...
CA
indication that the material is susceptible to this type of failure. Specimens
loaded above the yield point fail by rapid creep processes wi in considerable
plastic strain before failure. Stress-rupture failure has not been observed in
Q-titanium alloys but has been observed in two-phase (a + B) titanium alloys. 54
Stress-rupture failure generally appears to occur primarily by cleavage, a small.
crack being initiated at some point of stress concentration and propagating.
slowly until it reaches a critical size for rapid growth. .
ORNE-MEC - OFFICIAL
Stress-corrosion cracking has not been observed in zirconium-base alloys in
the dilute aqueous solutions usually encountered in reactor systems nor in the
...........
.........
--------------------
317
02'1-TEC - OFFICIAL
more concentrated solutions encountered in the chemical industry in which cor-
rosion has been too rapid to permit use. Quite recent research has indicated,
however, that a failure process observed in c-titanium alloys in hot saline
solutions may be related to or accelerated by a type of stress corrosion.55
Ino fracture bclavior of Q-zirconiun alloys secins to be the result of the
nucleation and growth of small cracks (possibly for jeä by cleavage) into voids
at points of high stress concentration and the coalescence of some of the small
voids to form the fracture surface.42,56 The cracks or voids are observed only
in regions of high local strain which may or may not be related to the average
strain away from the fracture surface. Both temperature and strain rate
influence the modes of deformation that will operate in any given condition. It
is probable that the stress configuration is also important. Thus, the amount
of energy absorbed in the fracture process can be small or large, depending on
the deformation modes operating and their strain-hardening coefficients.
Fracture at low tensile elongation or low energy absorption seems to be primarily
a problem of susceptibility to highly localizeð necking, rather than being due
to the processes thought to produce brittle failure in other materials.
STUDIES, CRITERIA, AND THEORIES NEEDED
We are at the point in the understanding of the influence of anisotropy on
the mechanical properties of cph metals that we can begin to define the studies
necessary to conduct in the near future, define the need for design criteria
that incorporate and specify the anisotropy of properties and the evaluation
tests to be used, and define some of the boundary conditions and physical
behavior the stress-strain theories must incorporate. .
We badly need basic studies on the types of dislocations that can exist in
ORNI - ACC - OFFICIAL
a cph metal; how these are affected by c/a ratio, alloying element, and bond
character; and how such dislocations operate to produce the observed strains,
38
0:31-ACC-os;
The studies must include transmission electron microscopy and x-ray diffraction
techniques as well as determinations of the body distortions in both uniaxial ::
and restrained flow testing.
Studies of the yield and flow loci in biaxially stressed specimens of
quantitatively known textures must be conducteå to supply both engineering design,
data for immediate use and stress-strain data for use in developing and evaluating
new theories of anisotropy of mechanical properties.
We need more, and more carefully designed, studies of the development of
texture during fabrication, and of how such development is influenced by the
details of the fabrication process. For example, how and why do the textures
observed in tubing fabricated by tube reducing differ from those in tubing pro-
duced by mandrel drawing? How important in the control of texture development
is the ratio of circumferential reduction to wall thickness reduction in each
operation? How important is that ratio in the production of tubing of the same
dimensions but of greatly different textures? Can textures be controlled to
ease manufacture, decrease the number of defects, increase mill yields, and
yet be moved in a final operation to that specified by the user?
We need more rapid and less expensive methods for determining the textures
in the experimental specimens and in commercial materials. Computer programs
have been developed to permit almost direct coupling of computers and automated
-
-
-
COS
...
-
x-ray diffraction equipment, greatly decreasing the time required to determine
a pole figure. 57 The problem of specimen preparation, however, remains. A new
technique, using a polarized light microscope, has been ieveloped for rapidly
(2 hr) determining basal pole figures of cph metals. 58 This technique requires
the metallographic preparation oî only three specimens per material.' It has the
advantage that a layer-by-layer texture analysis from surface to center is
possible. Another new technique uses Knoop microhardness measurements made in
------•. -- -
-
ONL-ACC-OISICIAL
gotenciana
e
mi
39
OSHI - Afr-OFICIAL
a specific pattern to determine an approximate basal pole figure in Zircaloy-2.
(Ref. 59) Other new methods are still needed to permit the investigator to
easily and rapidly determine the texture in each of his many experimental
specimens.
We need studies by design engineers and stress analysts to tell us how to
analyze engineering structures to determine the type of texture best suited to
resist the applied loads. Design engineers need information from trose .con-
ducting mechanical property tests on the anisotropy of stress-strain properties
as functions of texture so that they may safely set design limits and design
structures to utilize the anisotropy present.
More studies of the influence of texture and biaxial stress on the anisotropy
of strain and strain rates observed in creep must be conducted. For those
materials that will be used in reactor systems, the additional effects of radia-
tion damage must be examined during irradiation. Texture must be incorporated
as an important variable of the studies.
New test, evaluation, and failure criteria must be developed that are
те
applicable to the cph metals and they must be in forms. acceptable to both the
producer and the user.
It is quite apparent that criteria developed for and by
the use of steels can not be made to apply to structures built of cph metals.
Finally, we must develop new and adequate theories of the stress-strain
behavior of anisotropic metals in multiaxial stress systems, for both elastic
and plastic strains.
In such theories, the general case must be for metals with
no symmetry of properties so that mathematically, the behavior of cubic metals
would become essentially a special case, and the behavior of cph metals an
intermediate case. The theories must allow the stress-strain behavior to be
dependent on the sign of the stress (tensile or compressive), the crystallo-
graphic texture, the change in deformation systems with temperature and strain
IV1330 - VINO
rate, ani cu
multiaxial stress configuration,
0:31 - AL - OFICIAL
ODNL-KEC -0!8101:1
CONCLUSIONS
The examinations made of the deformation, creep, and fracture in cph metal.s
have led to the following conclusions :
1. Anisotropy of mechanical properties is a major feature of cph metals.
It can be used to advantage if we learn to control and use it, and it can be an
unrecognized hazard if we do not.
2. The anisotropy of mechanical properties in cph metals is directly
related to the perfection and orientation of the crystallographic texture
existing in the material, but the same texture can produce differing behavior
in different metals.
3.
The detailed stress-strain behavior of polycrystalline, highly textured
cph metals in uniaxial tests can be qualitatively expiained in terms of the
texture and the deformation systems observed to operate in uniaxial tests of
single crystals of the metal concerned.
4.
The development of texture during deformation and its stability to
further deformation of the same kind can be qualitatively explained in terms of
the strain path undergone by the material. In some cph metals, the deformation
modes observed in uniaxial tests of single crystals can be used to explain the
development and stability of texture in the polycrystalline material; in others,
they can not and other, presently unproved, deformation systems must be invoked.
5. The "brittle fracture" behavior of Q-zirconium and Q-titanium alloys
in some test conditions is actually that oľ a ductile fracture. The low-energy
DOA
absorption and low apparent ductilities observed are the result of highly
localized necking and low strain harāening.
6. Radiation damage in highly textured Zircaloy-2 tends to make the
08:11.- ALL-OFFICIAL
necking behavior somewhat more isotropic, al though the anisotropy oi yield and
flow stresses are still present...
ORNI - AEC - OFFICIAL
7. Some of the studies needed in the near future to improve our under-
standing and use of anisotropy have been outlined. Among these are a basic
study of the types and movements of dislocations possible in cph metals, more
particularly, a search for an experimental observation of dislocations permit-
ting slip with a strain component out of the basal plane. Others concern (1) the
.
determination of the stress-strein behavior in biaxial stress 'with texture as an.
important variable of the study; (2) the determination of yield loci in biaxial
stress for known textures in particular materials to provide engineering data
for immediate use in the design of structures; (3) new test, evaluation, and failure
criteria for cph metals used in structures; (4) studies of the influences of tex-
ture and biaxial stress on creep rates, both with and without irradiation during
test; and (5) new and adequate theories of anisotropy and stress-strain behavior
in cph metals in multiaxial stress configurations which will allow the stress-
strain properties to vary in space and with the sign of the stress.
ORNI -AS-DISCU
-
42
-...
FEFERENCES
ORX-dic-OFFICIAL
.
4
z 1. M. L. Picklesimer, A Preliminary Examination of the Formation and
Utilization of Texture and Anisotropy in Zi:ccaloy-2, Proceedings of the UJAFC
Symposium on Zirconium Alloy Development, QLAP-4089, Vol. II, Paper No. 13;
also ORNL-TM-460 (Feb. 28, 1963).
2. I. L. Dillamore, Trans. Amer. Soc. Metals, 58, 150–154 (1965).
3. P. L. Rittenhouse and M. L. Picklesimer, Metallurgy of Zircaloy-2
Part I: The Effects of Fabrication Variables on the Anisotropy of Mechanical
Properties, ORNL-2944 (Oct. 13, 1960),
4. P. L. Rittenhouse and M. L. Picklesimer, "Research on the Mechanica!
Anisotropy of Zircaloy-2," this sympo:sium.
5. D. F. Kaufman, J. J. Picket;, and L. R. Aronin, Fundamentals of Single
Crystal Deformation in Zone-Refined Beryllium, NMI-1266 (June 30, 1965).
. 6. P. W. Bakarian and C. H. Mat;thewson, Trans. AIME, 152, 226 (1943).
7. E. Schmid and W. Boas, Kristallplastizitat, Springer, Berlin (1935).
8. F. D. Rosi, C. A. Dube, and B. H. Alexander, Trans. AIME, 197, 257
(1953).
.. 9. E. J. Rapperport and C. S. Fartley, Trans. AIME 218, 869-876 (1960).
10. H. S. Rosenbaum, "Nonbasal slip in cph Metals and its Relation to
Mechanical Twinning," Deformation Twirning, Ed. by R. E. Reed-Hill, J. P. Hirth,
and H. C. Rogers, Metallurgical Society Conferences, Vol. 25 (1964).
11. R. E. Reed-Hill, "Role of Delormation Twinning in the Plastic Deformation
of a Polycrystalline Anisotropic Metal," Deformation Twinning, Ed. by R. E. Reed-
Hill, J. P. Hirth, and H. C. Rogers, Metallurgical Society Conferences, Vol. 25,
(1964).
12. R. L. Bell and R. W. Cahn, Proc. Roy. Soc. (London), (A) 239, 494-521
(1957).
TE0 -
ONI -A
"Oi VTLWte:.amour MC MV
'I,
"ii",
"mam
43
13. R. E. Reed-Hill, An Evaluation of the Role of Deformation Twinning
نه ده. - : زز - ندد از
1
in the Plastic Deformation of Zirconium, Ninth Quarterly Report, AEC Contract
No. AT (38-1)-252 (November 1963).
14. F. D. Roei, F. C. Perkins, and S. S. Seigle, Trans. AIME, 206, 115
(1956).
15. J. Weertman and J. R. Weertman, Elementary Dislocation Theory, The
MacMillan Co., New York, (1964) 113.
16. L. M. Howe, J. L. Whitten, and J. F. McGurn, Acta Met, 10(9), 773-787
(September 1962).
:17. P. L. Rittenhouse and M. L. Picklesimer, Metallurgy of Zircaloy-2
Part II. The Effects of Fabrication Variables on the Preferred Orientation and
Anisotropy of Strain Behavior, ORNL-2948 (Jan. 11, 1961).
18. J. C. McDonald and P. W. Bakarian, Trans. AIME 233, 95–103 (1965).
19. J. L. Klein, V. G. Macres, D. H. Woodard, and J. Greenspan, Chap. VII,
450–54, The Metal Beryllium, Ed. by D. W. White, Jr. and J. E. Burke, ASM, Metals
Park, Ohio (1955).
20. H. T. Clark, Trans AIME 188, 1154 (1950).
21. A. Hargreaves, J. Inst. Metals 71, 73 (1945).
22. J. H. Keeler, W. R. Hibbard, Jr., and B. F. Decker, Trans. AIME 197,
932 (1953).
23. C. J. McHargue and J. P. Hammond, Trans. AIME 197, 57 (1953). . ::
24. B. A. Cheadle and C. E. Ells, Trans. AIME 233, 1044–1052 (1965).
25. J. J. Laidler, Preferred Crientation in Extruded Zircaloy-2 Tubing,
HW-64815 (April 1960).
26. E. F. Sturcken and W. G. Duke, Measurement of Preferred Orientation in
Thin-Walled Zircaloy-2 Tubing, DP-607 (November 1961). '
27. D. H. Avery, W. F. Hosford, Jr., and W. A. Brckofen, Trans. AIME 233,
71-78 (1965).
ORNL - AEC - OFFICIAL
112111o-
51330-5v-RIO.
:
Inen
·
S.
28. H. H. Klepfer, Specific Zirconium Alloy Design Program. Surmary Report,
GEAP-4504 (April 15, 1964).
29. R. Hill, Proc. Roy. Soc. (London), (A) 198, 428-437 (1949).
30. ASIM A370-54T (issued 1.953, revised 1954).
i 31. J. Miklowitz, J. Appl. Mech. 15, 274–287 (1948),
.. 32. A. Nadai, Theory of Flow and Fracture of Solids, McGraw-Hill Book Co.,
New York (1950) 3225-327.
33. R. L. Mehan, Trans. ASME, Series D, J Basic Eng. 83, 449 (1961).
. 34. R. Hill, The Mathematical Theory of Plasticity, Oxford University Press,
London (1950).
35. C. C. Sanderson, The Effect of Specimen Geometry on the Tensile
Properties of Some Zirconium Alloys, AECL-2207 (March 1965). ..
36. A. L. Bement, "Radiation Damage in Hexagonal-Close-Packed Metals and :
Alloys," presented at the AIME Conference on Radiation Effects, Jack Tar Grove
Park Inn, North Carolina, September 8-10, 1965.
37. W. R. Kiessel and M. J. Sinnott, Trans, AIME 197, 331 (1953).
:. 38. L. G. Bell, Some Creep Properties of Zircaloy-2, CRGM-1017 (June 13, 1961)
.: 39. R. E. Reed-Hill, An Evaluation of the Role of Deformation Twinning in
the Plastic Deformation of Zirconium, Third Quarterly Progress Report, AEC Contract
NO. AT (38-1)-252 (April 1902).
40. R. C. Aungst, Stress Rupture Tests of Zircaloy-2 Pressure Tubes, BNWL-8
(January 1965).
• 41. H. I. Burrier, Jr., M. F. Amateau, and E. A. Steigerwald, The
**
V
:
-
Relationship Between Plastic Deformation and Fracture in Alpha Titanium, AFML-TR-65-239
(July 1965).
42. C. Jo Beevers, Trans. AIME 233 780-86 (April 1965..
43. A. L. Bement, J. C. Tobin, and R. .G. Hoagland, Effects of Neutron
Irradiation on the Flow and Fracture Behavior of Zircaloy-2, HW-SA-3194 (September
O.NL - AEC - OFFICIAL
**
* n
ar
1963). '.
.
.
.,
ORXL – AEC - OFFICIAL .
44. M. R. Louthan, Jr. and R. P. Marshall, J. Nucl. Mater. 9, 170 (1963).
45. R. E. Reed-Hill and W. D. Robertson, Acta Met. 5, 728 (1957).
46. J. J. Prislinger, Evaluation of Subsize Izod Specimer. Design for
Determining the Notch Toughness of Zircaloy-2, ORNL-M-336 (October 1962).
47. G. M. Adamson et al., HRP Quart. Progr. Rept. Jan 31, 1957, ORNL-2272,
124-126.
48. R. C. Aungst and L. J. Defferding, Crack Propagation Tests on Normal ..
and Hydrided Zircaloy-2 Reactor Pressure Tubing, HW-80567 (August 1964).
49. K. E. Horton and R. S. Stewart, Thermal Stress Fatigue Behavior oi
Zirconium and Zirconium Alloys, ATL-A-127 (October 1961).
50. J. W. Spretnak, M. G. Fontana, and H. E. Brooks, Trans. ASM 43, 547
(1951).
51. S. M. Bishop, J. W. Spretnak, and M. Go Fontana, Trans. ASM 45, 993
(1953).
52. R. E. Reed-Hill, E. P. Dahlberg, and W. A. Slippy, Jr., "Some Anelastic
Effects in Zirconium at Room Temperature Resulting from Prestrain at 770K," to be
published in Transactions of AIME, 1965.
53. E. Roberts and P. G. Partridge, "The Formation of Fatigue Cracks in
Magnesium at (1072} (1011) Twin Boundaries," Deformation Twinning, Ed. by :
R. E. Reed-Hill, J. P. Hirth, and H. C. Rogers, Metallurgical Society Conferences,
Vol 25 (1964).
54. D. N. Williams et al., Hydrogen Contamination in Titanium and Htanium
Alloys Part IV: The Effects of Hydrogen on the Mechanical Properties and Control
of Hydrogen in Titanium Alloys, WADC-TR-54-616 Part IV, ASTIA Doc. No. AD 131088
(September 1957).
46
-
55.
E. G. Bohlman and F. A. Posey, "Aluminwn and Titanium Corrosion in
..
:
-.
Saline Waters at Elevated Temperatures," presented at the First International
Symposium on Water Desalination, Washington, D.C., October 3–9, 1.965.
56. G. Ostberg, J. Inst. Metals 93, 223–228 (1964–1965)..
57. E. F. Sturcken, "The Use of Powder Reflection Intensities to Predict
Physical Properties in Oriented Material," Norelco Reporter, XI(1), 24-28, 43
(1964).
.
.
-
-
-
-
-
-
--
58.
M. L. Picklesimer and P. L. Rittenhouse, "Hydride and Basal Pole
Figures by. Quantitative Metallography," paper presented at the 19th Atomic Energy
Commission Metallographic Group Meeting, Oak Ridge National Laboratory,
April 20-22, 1965 (proceedings to be published as ORNL-IM-1161).
59. R. L. Rittenhouse and M. L. Picklesimer, "Comparison of Pole Figure
Data Obtained by X-ray Diffraction and microhardness Measurements on Zircaloy-2,"
presented at the AEC Contractors Meeting on Application of X-Ray Diffraction
to the Atomic Energy Field, Richland, Wash., February 25–26, 1965, proceedings
to be published as a BNWL report.
--.
-
• • •..
.-.-.
• •
•
-.
.
.-
-
.. ..….….….……… ….. ………..…. - • •
.-
..
-------.
...
-
...--
.
-
---
..….....
-
…
.
….
ORNI - AIC - OFFICIAL
LIST OF FIGURES
CORNL-LR-DWG
248302
Fig. 1.
Impact Energy Absorption vs Testing Temperature for
Schedule-8 Zircaloy-2. Subsize Izod ""-notch specimens, four notch
orientations:
L. - longitudinal direction specimen, notch axis in the normal
direction;
- transverse direction specimen, notch axis in the normal direction;
longitudinal direction specimen, notch axis in transverse direction
Tn - transverse direction specimen, notch axis in the rolling direction.
Basal pole concentration is in the normal direction.
ORNL-LR-DWG
48303
.
.
.
. .
• Photo 81032
Photo 81031
Fig. 2. Impact Energy Absorption vs Testing Temperature for Schedule-J
Zircaloy-2. Subsize Izod "y"-notch specimens, four notch orientations as
for Fig. 1.
Fig. 3. Ball Model of Face-Centered-Cubic Crystal Structure. The
fcc unit cell. protrudes from the top plane.
Fig. 4. Ball Model of Close-Packed Hexagonal Crystal Structure.
Atom arrangements on {2010} planes. Pattern of the bottom five rows of the
iront face obtained by adding the next layer of balls to the upper rows.
Fig. 5. Ball Model of Close-l'acked Hexagonal Crystal Structure. Atom
arrangements emphasized on {1012} end {1071} planes and (1123) directions.
Fig. 6. Ball Model of close-l'acked-Hexagonal. Crystal Structure. Atom
arrangements emphasized on {1121} and {1122} planes and (1123). directions.
Fig. 7. Twin Models in Zirconium. Constructed from Aluminum hexagona).
Photo 81033
Photo 81030
Y-47965
TORNL-LR-DWG
49430
bar stock.
Fig. 8. Typical Inverse Pole Figure for Plate Zircaloy-2. Rolled
and annealed Schedule-8.
- OFFICIAL
.....
-
-
ORNI - AEC - OFFICIAL
-
ORNL-LA-DWG
7453.8
Fig. 9. Schematic Drawing 02 Uniaxial Tensile Test of Sheet-Type
..
.
...
Zircaloy-2 Specimens.
(a) Single crystal analogue with basal pole in sheet normal direction,
(b) single crystal analogue with basal' pole in with direction,
(c) Single crystal análogue with basal pole in axial direction.
Fig. 10. Possible Yield boci in Biaxial Stress for Highly Textured
ORNL-DWG
64-11634R
Zircaloy-2.
....
(a) Random texture,
(b) Basal poles concentrated in the axlal direction,
(c) Basal poles concentrated in the tangential direction,
ons we .
(a) Basal poles concentrated in the radial direction.
.
ORNL AEC - OFFICIAL
19101110 - IV-INIO
WIJI110-)V-INIO

ORNL-LR-DWG 48302
SCHEDULE 8
..
'LV
LH
TH
I TV
45 ppm Hz
.
.
IMPACT ENERGY (in.- 1b)
.
L
-400
-300
-200
- 100
0
100 200
TEMPERATURE (°F).
300
400
500
600
ORNI-AEC - OFFICIAL
OANI - AEC - OFFICIAL
.
1
.
.
.
.
-
· ..
ORNL-LR-DWG 48303
SCHEDULE J
L
LH
TH
TV.
27 ppm Ha
IMPACT ENERGY (in.- 1b)
-400
-300
-200
-100
0
: 100 200
TEMPERATURE (OF)
300
400
500
600

,
-
.
.
.
..
.
.
.
.
.
.
“
Т
т.
•
т
.
ар
. 1.,
1
4
.
,,1
са?
..
:.
с
.
т
Х,
.
2
PHoтo 84o32
1.
*": ".:
ORNI - AEC - OFFICIAL


ORNL - AEC - OFFICIAL
-
-
.
.
.-
R
{0101}
:
...
'
.
.
.
:
'
bas mayor":
.
O
.
V
IS,
--
-
:
.
!
Transp
1E018 OLOKO
,.
,
'"
.--.
miwonman
ente
minden
-.
-
----
Widiiio - 337-1480

-
mo
tor --....
I.
TIDISJ0-91Y - INYO
.
ORNI - AEC - OFFICIAL
.. جدهن منهم
. ۱۹۰۰: ..... نه ړ
.
.
*
-
د
.
{101}} ای
*
مبتنفعند منتصف تصنقعته
'
ا
=
=
-
متین نشست انداختن
ب
ا
-
1
ه
..
(123)
(Io0o)
هومن محمد ... ۰
۰
مهدیه محلیط
.
ممممممممممممم.. مه مهم
20 .::.. تم الا... ... ...
و فره... رد:... ... ... ... ...
نه... .
.
:دان:13
...
.
. .
.
.
.
. . . . . ..... . . ... ... ...
... ... ... ... - ممه سم
از دهة - جزر -
PHOTO 81033

PHOTO 81030
..
!
s the
er en mor erosion i
{1012}
.
........ Homilier
no
.
(0001)
....
amo
!
-
.
11
.
'
26
:'
...
(1123)
{1011
(1123)
?
1
10
..
mploi Entrepreneur
{1211}}
er
entr se
:
[2211} |EZI
1122}|
.
* min.
carmine
E
islanda
-
.....m
more, n
.
.
-
-
--
--
-
OINL AEC - OFFICIAL
ORN
-OBICNO

Mi-15-OFFISIE

|
|
اب و
اه
ء
ات
,
۱
. IT
(1012)
(1131)
(112)
سیشن
.
../
ANL - AEC - OFFI
•
-
-
-
-
ORNL-LR-DWG 49430.
all20)
''.
.ORNL - AEC - CIFFICIAL


..
.
..
(1124)
...
3.0,
60
5.6
(0001)
4.4 4.2 1.8
(1015) (1014) (1013)
0.3
(1011)
NORMAL DIRECTION
(101o
.
1.9, 201

1
2.2 (2130)
.
1.8
.
(11222016
(2131)
- 2.25
10.5 i
9.5
9.75
0.4
2.0
(0001)
(1010)
(1010)
TRANSVERSE DIRECTION
2.7
(1120)

3.0
2.5
2.8.(2130)
(1122).
1
line
l
l
:
:
ORNL - AEC - OFFICIAL
(0001)
0.9
. (1011)
ROLLING DIRECTION
1.7
(2021)
mmodo consentan abortionem
...onco
ORNL-LR-DWG 74 518
ORNL - AEC - OFFICIAL

(0) KO
Wq> Wq > Wz
tq = tz = t3
13 ONLY SLIGHTLY
LESS THAN TO
ta
13

(6)
N
· W, = W2 %
to > ta>

W, W2W
.
.
W > W2 > Wz
ta> t3
V
..
i
?
ORNL - AEC - OFFICIAL
Texture Cells and Strain Patterns in Sheet Tensile
Specimens of Zircaloy-2.
..
.
.
.
C.
Random
Tangential
or
. Radial
Tangential
Pure intornol, a
a:2, Internal pressure
grossuro
(closed ond)
El Har1
Ideal
Isotropic
(mi
(+
σ, Axial
Pure, a
compression
Q:0, Pure tension
: -1, Pure torsion
Basal pole il to tube
circumference
Axial
Assumptions:
Assumptions:
Yield strength, YS, icontical in all dir actions
. 2. YS identical for tension and compression
1. On solo
i
.
.
2. 02.02.
3.01,2€ -01,26
Axial -radial stress
ollipse as in d
.
d.
Tangential
or
Radial
. Tangential
- Tangentiel is.
(+)
GH
Basal poie !! to-
tube axis
(+)0, Axial
Axial
Basal pole 11 to tube
radius
Assumptions:
Axial-radial stress
ellips. as in C
'Assumptions:
1901
2.026 020 :
3.01.2¢ .25
LOOT
2026026
3.01,24 125
POSSIBLE BIAXIAL YIELD STRESS ELLIPSES FOR ZIRCAL
Assuming: (1) Plane stress conditions (thin wall tubing)
(2) A modified Octahedral Shear Stress Theory is valid
- ASP - OSSICIAL
ORNI - AEC - OFFICIAL








..
:
té
2 OF 2
ORNL P
1753
'
.
- im
.
..
•
..
.
.
.
1145
SO
5 6
A
'
19
1
Pl1.25 1.1.4 16
MICROCOPY RESOLUTION TEST CHART
NATIONAL BUREAU OF STANDARDS - 1963
Table I. Room-Temperature Yield Strengths of Zircaloy-2
in Tension and Compression
Yield Strength
Specimen
(10 psi)
Schedule
Orientation
Tension Compression
RD
54.9
58.3 :
MD
66.7
62.7
122.0
ND
RD
D
55.8
63.2
52.0.
70.0
RD
TD
65.6
101.3
72.5
ND
'For fabrication history, see Ref. 3.'.
PRD – rolling direction, TD – transverse direction,
and ND - normal direction.
ORNI - AEC - OFFICIAL
END
4
.
DATE FILMED
12/ 13 / 65







SH
SY
1