nasa cr- 1433 nitinol characterization study and …€¦ · nasa cr- 1433 nitinol characterization...
TRANSCRIPT
NASA CR- 1433
NITINOL CHARACTERIZATION STUDY
By William B. Cross, Anthony H. Kariotis,and Frederick J. Stimler
Distribution of this report is provided in the interest of
information exchange. Responsibility for the contents
resides in the author or organization that prepared it.
Issued by Originator as Report No. GER 14188
Prepared under Contract No. NAS 1-7522 byGOODYEAR AEROSPACE CORPORATION
Akron, Ohio
for Langley Research Center
NATIONAL AERONAUTICS AND SPACE ADMINISTRATION
For sale by the Clearinghouse for Federal Scientific and Technical Information
Springfield, Virginia 22151 - CFSTI price $3.00
https://ntrs.nasa.gov/search.jsp?R=19690026989 2020-06-22T06:20:24+00:00Z
FOREWORD
Goodyear Aerospace Corporation (GAC) of Akron, Ohio, conducted the Nitinol Charac-terization Study for NASA-Langley Research Center under Contract NAS 1-7522 to define the
structural and recovery properties of several Nitinol compositions possessing different tran-sition temperatures. Mr. Frederick J. Stimler of the Space Systems and Analytics Divisionwas the GAC project engineer, Mr. William B. Cross the material specialist, and Mr. AnthonyH. Kariotis the test engineer. The study was essentially a 19-month program conducted from
July 1967 through January 1969.
The work was administered by the Applied Materials and Physics Division of LangleyResearch Center (LRC), with Mr. John E. Cooper of the Spacecraft Applications Section/SpaceEnvironment Branch acting as project engineer. Dr. David E. Bowker and Mr. Howard L.Price of AMPD also served as valuable technical monitors on the program. Mr. Kenneth D.Albert of the LRC Procurement Division was the contract administrator.
Nitinol rod and all foil materials were supplied by Mr. William J. Buehler of the Mag-netism and Metallurgy Division of the U.S. Naval Ordnance Laboratory (NOL), Silver Spring,Maryland. Mr. Buehler's enthusiastic support and constructive suggestions did much to pro-vide understanding of the behavior of this unique material and guidance during the test and data
analysis phases of the program.
As subcontractor to GAC, Battelle Memorial Institute (BMI), Columbus, Ohio, processedthe basic Nitinol rod into diameters of 100, 20, 15, and 10 mils under the direction of Dr.Curtis M. Jackson and Mr. David C. Drennen.
The contractor's number for this report is GER 14188.
iii
SUMMARY
Goodyear Aerospace Corporation has determined the physical, mechanical, and electri-cal characteristics of nominal 55-Nitinol, a nickel-titanium alloy, to enable evaluation of itssuitability for potential applications in space. Materials from three different melts were fab-
ricated into rods, wire, and foil for the investigation, andthe tests summarized in table I
were accomplished.
The three different compositions had martensite transformation starting temperatures
(Ms)* as follows: Type A, -60°F; Type B, 65°F; and Type C, 105°F. The starting materialsfor this program are defined in table 1I. A rod diameter of 100 mils, nominal wire diametersof 20, 15, and 10 mils, and foil thicknesses of 6 and 3 mils were evaluated during the study.
Complete process records were kept to ensure detection of the critical variables and re-
peatability of the material fabrication and performance characteristics.
As indicated in table I, several tests were repeated during the program, with differentsamples of the same type Nitinol resulting in consistent or explainable data. In addition, ef-fects of such parameters as sample width, sample length, temperature, loading, and blockingwere investigated sufficiently to define design criteria. Cooling and heating effects on thestress-strain behavior of Nitinol were investigated in detail. Sufficient comparison tests were
made to substantiate the expected Nitinol behavior based on the tension data.
Substantial design information is now available for utilization of 55-Nitinol in many varied
applications. The processing techniques necessary to preserve NiTi "memory" character-istics are better understood and seem somewhal controllable. Ways have been developed to
predict and improve this unique memory phenomenon of Nitinol. Program results showed thatthe material can be processed into rod, wire, and foil shapes and still maintain the desired
memory, physical, and structural properties.
This program represents a big coordinated step. Earlier test data could be consideredexploratory - limited for general use and applied with caution for many applications. Manytests and cross checks were possible with the present program, since the raw data was thor-oughly analyzed by cognizant specialists from government and industry to verify the data and
direct further testing.
The characterization tests conducted during this program provide adequate informationto facilitate the design of expandable or erectable structures in particular, and many other
applications in general.
*M s values shown are for hot-swaged bar stock in the "as-received" condition.
2
I_1 I
i 1__
L _
vi
I
i
_o
oz _
_uuu_
0Z
00
ft.
0
z121
.<N
r..)
o
0
U
.<
.<
0
r._
i
c.oo
o .o
_or._
0
0
d
•_,oo"_
0
.,l-a
0o
0 _0 Lea_0 ['- 0
f.O f,_ 0
0 00 0 0
0 0
! _,1 0
I I
0
0
o _ ,_r._ LO L_
N uN oN! I !
• 0
• 0
I
0
_ _ o _"_, o_ _ -_ ._
g
N o • _ l_o _N
._ _ o oh ._ _ .o
..0 t.-,
Z "_ o_ ._ o
• !
_ _. _ .r._ •
CONTENTS
Page
FOREWORD ............................... • iii
SUMMARY , ............................... vSYMBOLS AND ABBREVIATIONS ...................... xii
INTRODUCTION .............................. 1
REVIEW OF NITINOL PROPERTIES AND UNIQUE BEHAVIOR ......... 2
MATERIAL INVESTIGATION ...................... 4
General ................................ 4
Swaged Rods .............................. 4
Rolled Thin Sheet (Foil) ....................... 4Drawn Wire .............................. 5
EXPE RIMENTAL PROC EDURE ....................... 10
EXPERIMENTAL RESULTS AND DISCUSSION ................. 11
Direct Observation of Transformation ............ . ...... Ii
General .............................. 11
Electrical resistance versus temperature ............... 11
Unisxial dimensional change .................... 20Mech__nic al Behavior .......................... 23
Stress versus elongation curves ................... 23
Degree of shape recovery ...................... 27
Recovery stress ........................ 40Mechanical work ........................ 42
Shape recovery fatigue ......... ............ 51
Thermal cycling effects ..................... 53
CONCLUSIONS .............................. 5759REFERENCES ..............................
BIBLIOGRAPHY .............................. 59
Figure
1
2
3
4
5
ILLUSTRATIONS
Page
Block diagram of wire drawing schedule ............... 6
Microstructures of longitudinal sections of as-received hot-swagedNitinol alloys ........................... 7
Microstructures of longitudinal sections of as-drawn Nitinol alloys . 8
Microstructures of longitudinal sections of annealed Nitinol alloys 9
Typical electrical resistivity versus temperature curves for drawnNiTi wire annealed at 932°F (500°C) ............... 14
Effect of post-drawing anneal temperatures on electrical resistanceand bend recovery behavior ..................... 15
Electrical resistance versus temperature curves for swaged rods 16
ix
Figure
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Page
Electrical resistance versus temperature curves forComposition A materials ..................... 17
Electrical resistance versus temperature curves forComposition B materials ..................... 18
Electrical resistance versus temperature curves forComposition C materials ..................... 19
Typical uniaxial dimensional change'behavior for Composition B andC drawn wire annealed at 932°F .................. 21
Uniaxial dimensional change versus temperature curves for100-mil diameter Composition B and C rods ............. 22
Uniaxial dimensional change versus temperature curves for100-rail diameter Composition A rods ................. 23
Stress versus elongation curves for 100-rail diameter Composition B
rods during cooling and heating cycles ............... 24
Stress versus elongation curves for 20-mil diameter Composition Bwire during cooling and heating cycles ............... 25
Yield stress and modulus versus temperature curves for 100-mildiameter Composition B rods during cooling and heating cycles .... 26
Tensile shape recovery versus initial strain curves forComposition A materials ..................... 28
Tensile shape recovery versus initial strain curves for
Composition B materials ..................... 29
Tensile shape recovery versus initial strain curves forComposition C materials .... ................. 30
Compression shape recovery versus initial strain curves forComposition B and C rods ..................... 31
Strain in outer fiber as a function of bend diameter and specimen thickness 32
Bend recovery versus temperature curves for 100-mil diameterComposition A rods ........................ 33
Bend recovery versus temperature curves forComposition B materials ..................... 34
Bend recovery versus temperature curves forComposition C materials ..................... 37
Bend shape recovery versus initial strain curves forComposition A materials ..................... 40
Bend shape recovery versus initial strain curves forComposition B materials ..................... 41
Bend shape recovery versus initial strain curves forComposition C materials ..................... 41
Tensile recovery stress versus temperature curves for 100-raildiameter Composition B rods at various strain levels ........ 42
Maximum recovery stress versus initial strain curves forComposition A materials ..................... 43
x
Figure
30
31
32
33
34
35
36
37
38
39
40
41
Page
Maximum recovery stress versus initial strain curves for
Composition B materials ..................... 44
Maximum recovery stress versus initial strain curves forComposition C materials ..................... 45
Mechanical work versus temperature curves for 100-mil diameter
Composition A rods ........................ 46
Mechanical work versus temperature curves for 100-mil diameter
Composition B rods ........................ 47
Mechanical work versus temperature curves for 100-mil diameter
Composition C rods ........................ 47
Maximum mechanical work versus initial strain curves for
composition A materials ..................... 48
Maximum mechanical work versus initial strain curves for
Composition B materials ..................... 49
Maximum mechanical work versus initial strain curves for
Composition C materials ..................... 50
Tensile recovery versus number of deformation cycles for 20-mildiameter Composition B wire. 6 and 8 percent strain ........ 51
Load-elongation curves following shape recovery cycling for20-mil diameter Composition B wire ............... 52
Electrical resistance versus temperature curves following repeated
deformation cycles for 20-mil diameter Composition B wire ...... 54
Thermal cycling effects on bend recovery for 20-mil diameter56
Composition B and C wire .....................
TABLES
Table
I
II
III
IV
V
VI
Page
Summary of Test Program for Nitinol Characterization Study ..... vi
Starting Materials for Nitinol Characterization Study Program .... vii
Summary of Physical and Mechanical Properties of Nominal55-Nitinol ............................ 3
Summary of Types of Tests, Specimen Sizes, Gage Lengths,and Load Rates .......................... 12
Approximate Md, Ms, Mr', and A s Critical Temperature Points forMaterials Investigated ....................... 20
Degree of Strain Maintained in Thermal Cycling Specimens ...... 55
xi
SYMBOLS
A S
Md
Mf'
M s
Start of reverse.transformation, NiTi (III)--,--NiTi (II).NiTi (II) and NiTi (III) designate structure variations(phases) produced by diffusionless (martensitic) trans-formation.
Maximum temperature for martensite formation bydeformation. NiTi (II) structure thermodynamically
stable at this point.
Temperature where macroscopic atomic shear c eases.(An arrest point, since transformation can be made to
proceed through mechanical deformation. }
Martensite transformation starting point in cooling.
ABBREVIATIONS
BMI
GAC
LRC
NOL
SMC
Battelle Memorial Institute
Goodyear Aerospace Corporation
Langley Research Center
Naval Ordnance Laboratory
Special Metals Corporation
xii
NITINOL CHARACTERIZATION STUDY
By W.B. Cross, A.H. Ksriotis, and F.J. StimlerGoodyear Aerospace Corporation
INTRODUCTION
Nickel-titanium alloys (Nitinol, TiNi, or NiTi) of proper composition exhibit unique
mechanical "memory" or restoration force characteristics. 55-Nitinol is an NOL designationfor this series of near-stoichiometric intermetaliic compound alloys. This designation is used
to typify the generic class of shape-recoverable alloys ranging in composition from about 52 to56 weight percent nickel, the balance titanium. The name is derived from Ni-Ti-NOL. Theprefix numerical value indicates the nominal nickel content in weight percent.
The shape recovery performance of this material is phenomenal. The material can be
plastically deformed, such as in the simulation of model packaging conditions, and then re-stored to the original configuration or shape by heating it above the characteristic transitiontemperatures. Mr. Buehler and his colleagues at NOL have shown qualitatively, using crystal-lographic techniques, that a martensitic (diffusionless) atomic structure change is responsiblefor this unique mechanical memory, aswell as other anomalous properties (ref. 1).
This unusual behavior is limited to NiTi alloys having near-equiatomic composition. Apure stoichiometric (50 atomic percent) NiTi alloy will have a nickel content of approximately55 percent by weight. Increasing the nickel concentration lowers the characteristic transfor-mation temperature of the alloy. Nickel concentration increases are limited to about 56.5percent by weight due to the formation of a detrimental second phase. Further reduction intransformation temperature, with shape memory retention, however, is possible through par-tial substitution of cobalt for nickel. Using this substitution method, it is possible that asingle-phase alloy can be made which will have a transformation temperature approaching-321°F. Without cobalt substitution, the transformation range can be lowered from about200 ° to about 15°F.
The present effort was directed toward obtaining large quantities of physical behaviordata with three different types of compositions and many specimen forms processed from them.Hundreds of curves of these data have been analyzed and thus provide background for quantita-tive understanding of the shape recovery performance of NiTi alloys not previously available.
REVIEW OF NITINOL PROPERTIES AND UNIQUE BEHAVIOR
Since the recent discovery, at the U.S. Naval Ordnance Laboratory, of the extraordinaryshape recovery or mechanical "memory" property of the equiatomic NiTi intermetallic com-
pound designated 55-Nitinol, considerable interest has been generated in the basic and engin-eering features of this material. Some of the physical and mechanical properties of nominal55-Nitinol are listed in table III. (Detailed data may be found in references 1 and 2. )
The mechanical memory of this material is easily demonstrated by plastically deformingan annealed wire or sheet of NiTi at room temperature and then heating it, whereupon it willrevert to its original shape. This unusual behavior is attributed to a reversible low tempera-ture diffusionless (martensitic type) transformation which is not complete at room temperature.The transformation, however, can be made to proceed at room temperature by straining thematerial. Thus it is assumed, when demonstrating mechanical memory, that plastic defor-
mation occurs through transformation and that reverse transformation takes place upon heating.By this process plastic deformation is erased, and the material recovers its original shape.
In order to preset an original shape, the alloy must first (i.e., prior to plastic deforma-tion) have been annealed while constrained in the shape that it is to "remember. " These an-neals are usually carried out at temperatures from 750 ° to 1 ll0°F. However, the amount ofplastic deformation that the alloy can tolerate, and still revert to its original shape when re-
heated, depends on the temperature at which it was annealed. Thus there appears to be anoptimum annealing temperature, which is a function of the alloy composition and prior process-ing history.
To understand the cause of the shape memory effect, one must consider the phases thatare present in the 55-Nitinols at various temperatures. It is generally agreed that at tem-peratures above about 1200 ° - 1300°F these alloys consist of a single-phase body-centered-
cubic random solid solution. This phase is not believed to enter into the shape memory phe-nomenon.
There is much disagreement in the technical literature (ref. 3, 4, 5, 6, and 7) in re-gard to the phases that exist in the Nitinol alloys from 1200 - 1300°F down to their transition
temperatures (Ms). Some investigators believe that the alloy in this temperature r_nge con-sists of an ordered single-phase NiTi matrix that has a complex structure with a 9 A repeatdistance; others state othat it is a single-phase (NiTi) ordered CsCl-type phase with a latticep.arameter of about 3 A; while still others believe that two phases are present. One of the two
phases is said to be NiTi with an ordered CsCl-type structure, and the other is either Ni3Ti(for alloys that are slightly rich in nickel) or Ti2Ni (in the case of alloys that are slightly richin titanium).
For the present, it will suffice to say that a phase that is stable or metastable above the
M s temperature can transform to a martensitic-type phase at temperatures below M s. Trans-formation occurs below M s both athermally (on cooling) and by plastic deformation. It appearsthat the grains that transform athermally do not contribute to the shape memory effect.
Shape memory apparently comes about as a result of the formation of the acicular phaseby deformation of the metastable retained higher-temperature phase. When 55-Nitinol is
plastically deformed at a temperature near or below the M s point, the retained phase is trans-formed into the acicular product. The orientation of the transformed phase depends on thestate of stress that caused the transformation. The "memory" arises from the fact that, uponheating (above the A s point), the oriented grains of acicular phase transform in the reversedirection and assume their original shapes; thus the mechanical plastic deformation is annihilatedand the sample returns to the shape it had prior to deformation.
TABLE HI. - SUMMARY OF PHYSICAL AND MECHANICALPROPERTIES OF NOMINAL 55-NITINOL
Density, lb/in. 3 ....................... 0. 234
Melting point, °F ...................... 2390
Magnetic permeability .................... < 1. 002
Electrical resistivity, microhm-cmAt 68°F .......................... ~80At 1652°F ........................... 132
Mean coefficient of thermal expansion(75 ° - 1652°F) per OF x 10 -6 .................. 5.7
Hardness a
1742°F, furnace-cooled ................. 89 R B1742°F, quenched (R.T. water) .............. 89 R B
Tensile aUltimate tensile strength, ksi ................ 125
Yield strength, ksi .................. See fig. 16
Elongation, percent .................. 60
Young's modulus, millions of psi ............ See fig. 16
Shear modulus, millions of psi ............... 3.6Poisson's ratio ..................... 0.33
Impact (Charpy) aUnnotched (R. T.), ft-lb .................. 155Unnotched (-l12°F), ft-lb .................. 160Notched (R. T.), ft-lb .................... 24Notched (-ll2°F), ft-lb ................... 17
Fatigue (Std R.R. Moore Test) aStress (107 cycles-runout), ksi ............... 7O
NOTES: 1. Tests were performed at room temperature except wherenoted. Room temperature was below the transition tempera-
ture range.
2. Detailed data may be found in references 1 and 2.
aData obtained on hot wrought alloys.
MATERIALINVESTIGATION
General
Threedifferent compositionsdesignatedA, B, andC were evaluatedduring the courseoftheprogram. In eachcasethebasic material wassuppliedby the U.S. NavalOrdnanceLab-oratory (NOL), Silver Spring, Maryland, in theform of 0.2- or 0.3-inch diameter swagedrods and0.003- and0.006-inchrolled sheet. Prior to testing, therods were processedintowires of 100, 20, 15, and10mils in diameterby Battelle Memorial Institute (BMI) of Colum-bus, Ohio. The thin sheet (foil) materials, with the exception of final annealing, did not re-
quire any additional processing. The chemical composition and martensite start temperatures(Ms) of the materials are given in table II.
Details of the processing steps involved in producing the swaged rod and thia sheet aredescribed in the following paragraphs.
Swaged Rods
Composition A rod was processed by NOL from a sound cast ingot hot-rolled to a thick-
ness of about one inch and then cut into square bars. Three bars were swaged at 1562°F toapproximately 0.3 inch in diameter.
Composition B was hot-swaged by NOL at 1562°F to 0.2-inch diameter rod from hot-
rolled 1-inch square bars supplied to NOL by Special Metals Corporation (SMC), New Hartford,New York.
Composition C was processed by NOL in a manner similar to Composition B. However,original ingot sections obtained from SMC were first remelted by nonconsumable arc (under apartial atmosphere of argon) into bar stock approximately 5/8 inch in diameter.
Rolled Thin Sheet (Foil)
Compositions A, B, and C were rolled by NOL into 0.006- and 0.003-inch thick sheetsas described in the following paragraphs.
Hot-rolled. - All heats were hot-rolled at a temperature of 1562°F for the initial passes.This temperature was progressively lowered, as the sheet became thinner, to a final tempera-ture of 1292°F. Final hot-rolled thickness was 0. 040 inch. All hot-rolling was done in a 2-high rolling mill.
Warm-rolled. - Following appropriate shearing, the 0. 040-inch strips measuring ap-proximately 3 inches wide were warm-rolled in a 4-high rolling mill. The warm-rolling pro-cess consisted of heating the strips in a tube furnace mounted ahead of the rolls and maintained
at 1200°F. Heating time varied with the strip thickness and ranged from three minutes for the0. 040-inch thick strip to one minute for the 0. 006-inch thick strip. The strip was reheated
after eachpass through the rolling mill. Up to 5 mils of "screw-down" was used for each passon the approximately 0.040-inch thick strip until a smooth surface was attained. As many asfour or five passes were made at the same setting when approaching the 0. 006-inch thickness.
A portion of each heat was processed to 0. 006-inch thickness. The remainder were slit Ion-gitudinally in 3/4-inch wide strips. The edges, which were of variable thicknesses, were dis-carded. These strips were then rolled, by repeated passes and heating (1200°F), to a thick-ness of slightly greater than 0. 003 inch.
Cold-rolled. - Following the hot-rolling sequence, strips of Composition B and C mate-rial slightly thicker than the desired 0. 006 and 0. 003 inch were given a reduction by cold-rolling. This resulted in a strip approximately 5 percent longer and desired thicknesses of0. 006 and 0.003 inch. Composition A strips were not cold-rolled since the material is well
above its M s point at room temperatures and therefore difficult to plastically deform.
Studies were conducted on the 0.006- and 0. 003-inch thick sheets to determine the final
anneal temperature required for good shape recovery performance. The results of this workindicated that 930°F (500°C) was satisfactory for 0. 006-inch thick Composition A, 0. 006- and0. 003-inch thick Composition B, and 0. 006-inch thick Composition C. A slight performance
improvement, however, was gained by annealing the 0. 003-inch thick Composition A sheet atll12°F (600°C) and the 0. 003-inch thick Composition C sheet at 752°F (400°C).
Drawn Wire
The Composition B and C materials were successfully processed by BMI from 0.2-inchdiameter swaged rod into wires of 100, 20, 15, and 10 mils in diameter by utilizing cold-
drawing techniques. Drawing was accomplished below the upper range of each alloy's transi-tion temperature. Above the transition temperature, fabrication difficulties arise, in thatthese materials are stable at a particular phase. The procedures developed by Buehler (ref.
1) were used as a general guide, but were modified in several important respects.
For Composition A, special drawing techniques were developed whereby it could bedrawn readily despite its low transition temperature range. Drawing was accomplished at atemperature of about -10°F. The wires were held at the drawing die in an insulated tank con-raining a mixture of Dow polyglycol P-400 and dry ice. This mixture was maintained at atemperature between -4 ° and -22°F. Furthermore, the polyglycol served as an effective lub-ricant; thus no additional lubricant was required.
Studies were conducted on alloy samples of 100, 20, 15, and 10 mils to determine thebest annealing temperature required to produce complete shape recovery. The results of these
tests indicated that 930°F (500°C) was the optimum temperature for Compositions B and C ateach diameter. The optimum annealing temperature for Composition A was found to be il 10°F
(600°C). Final "straight" annealing was accomplished by passing the wire, under tension,through a 19-inch long tube furnace. Tension and rate of travel through the furnace were ad-
justed to the requirements of each particular wire size.
The actual processing schedule employed in converting the alloys into wire is summarizedin figure 1.
Figures 2 through 4 show typical microstructures of Compositions A, B, and C in theas-received hot-swaged (fig. 2), as-drawn (fig. 3), and annealed (fig. 4) conditions. In gen-eral, the microstructures consist of a single acicular phase containing a random dispersion
(as well as some stringers) of nonmetallic inclusions. The inclusions are presumed to be
i COMPOSITION A
0.3-1NCH DIAMETER
HOT-SWAGED ROD
CENTERLESS GRIND JTO 273 MILS
HOT SWAGE AT1475°FTO 140 MILS
+I
BELT GRIND TO J
I130 TO 136 MILS
ANNEAL AT1475°F FOR
5 MIN - - AC
DRAW TO 100 MILS,10% RA/DI E.DEGREASE ANDANNEAL AT 1290°F
AFTER 10% RA DRAWAT-4 ° TO-22°F.
J FINAL "STRAIGHT" IANNEAL AT 1110°F
f
COMPOSITIONSB AND C
0.2-1NCH DIAMETERHOT-SWAGED ROD
CENTERLESS GRINDTO 175 MILS
ANNEAL AT1290°F FOR6 MIN - AC
DRAW TO 100 MILS,10% RA,/DI E.DEGREASE ANDANNEAL AT 1290°FAFTER 10% RA.
DRAW TO 32 MILS,Ira/0 RIb/DIE. _DEGREASE ANDANNEAL AT 1290°F
AFTER 10°/0RA.
DRAW TO 20 MILS,2m/0 RA/DI E.DEGREASE ANDANNEAL AT 1290°F
AFTER 20% RA.
DRAW TO 15 MILS,2m/0 RA/DI E.DEGREASE ANDANNEAL AT 1290°F
AFTER 20% RA.
DRAW TO 10 MILS,20% RA,/DI E.DEGREASE AND
ANNEAL AT 1290°FOR 1110°F AFTER
200/0 RA.
TEST MATERIALS
II00-MIL DIAl
l _IFINAL "STRAIGHT" l
I
20-MIL DIA J
15-MIL DIA J
10-MIL DIA ]
LEGEND
RA = REDUCTION IN AREA
AC= AIR-COOLED
Figure 1. - Block diagram of wire drawing schedule.
6
(A) COMPOSITION A NITINOL ROD, 273-MIL DIAMETER
(B) COMPOSITION B NITINOL ROD, 205-MI L DIAMETER
(C) COMPOSITION C NITINOL ROD, 205-MIL DIAMETER
!?' ,
6· . \)
.. , <'),0
.....,.
I
o
-' \0'
. -.. :, .... _,..
500X
:.J (~' 7;.,
, ,....
, ..
.' ,-<> -
o
I. - , 0_ .. . •
,. . ';t'........ . , '-1 _ \
(') ~79 'r
" .. ~
o -
o
,-'
< ..
/ ':":"-'
•
9C799
'0
/ -
-....:=- ...
'--- 0' .,,-/,'
500X 82H20-14HN03-4HF 9C803
.. .
0', •
.- - .... -
500X
. . ....
c" " " "
" .
..... ..... I . .. :'~" ... -~ . ~.. - '.. ..
" .. ~ .. '
"., ", ' , '-I <
<>
__ : .. ~ r: .... . ~ . . . .
. . "-..,' ..
...-- - . / . 9C802
Figure 2. - Microstructures of longitudinal sections of as-received hot-swaged Nitinol alloys. Alloys were hot-swaged at NOL.
7
If .J
" .....
n 0 '" .
~ .. . .. fJ 0
4 .. -~
~~ . ~ . . - . '"' .
o
' . .. -..' - . ~.
-- '0-'"
.. ~ .-" .......
.:" . :-.
500X 9C798
. ,
-,
, .. o .-.. -0-. _
.' c.... o '. ~
' -
500X 9C794
- :' ...... -~- -- ~ ... "": -: - . - o ~ . _ -: ,. ---
- • ~ 6Q - ~.-'JII-'~
.. _.-.. -......
t",· • ..
-a---"-
.":'-c -'. • .' - ..;; -- ..
-- . ~ 11, _It-
_ ... '~ _ .. .c.~ ..
~--- .:"'D_ • ........ ". ", _ ,0 .
.' .... .: .. ~ . . .. . . '-.:q~- 0 _. ,-
.• 0
.-....-. ,. .. -- ... .. ..,::. -
-.
-" .. : . ~~ ... o~~ .. . : ....... : .... 0 _ ...
- ~: ... - '--' ... ~·"'=Aj",.-. . .. .. ... -~. -~ ... - .!.: - '. ~ .--'. -:- . - - . -
':;"'.- .. . _0,':,-' --
500X 9C797
(A) COMPOSITION A NITI NOL WI RE, 100-MIL DIAMETER, 10% REDUCTION
(B) COMPOSITION B NITINOL WIRE, 20-MI L DIAMETER, 20% REDUCTION
(C) COMPOSITION C NITINOL WIRE, 100-MIL DIAMETER, 10% REDUCTION
Figure 3. - Microstructures of longitudinal sections of as-drawn Nitinol alloys.
8
.- -- - - -- -- --------~
(A) COMPOSITION A NITINOL WIRE, 100-MIL DIAMETER, 10"10 REDUCTION + STRAND ANNEALED AT 11100F (AIR-COOLED)
(B) COMPOSITION B NITINOL WIRE, 20-MIL DIAMETER, 20% REDUCTION + STRAND ANNEALED AT 9300F (AI R-COOLED)
(C) COMPOSITION C NIT! NOL WI RE, 1 OO-MI L DIAMETER, 10"10 REDUCTION + STRAND ANNEALED AT 930°F (AIR-COOLED)
-..
500X
• 0 •
500X
500X
..... - ."
-, .
o
--.. -
. - >
-.J
; .. \.
.. (;
V' . . ..... . ..: '"".
-t t> 00 . 0 • () o-~-::.
D o .:~ '.,
..... - ... ... ~. ~ ,~
o ..... ..:
...... . - . ...:
..
9C796
- 0
-a
9C795
9C805
Figure 4. - Microstructures of longitudinal sections of annealed Nitinol alloys.
9
primarily titanium carbonitrides, with a few nickel-titanium oxides. The stringers of inclusions are aligned with the axis of the rod or wire, as would be expected.
A comparison of figures 2A and 2B shows that the microstructures of as-received Compositions A and B are very similar. By comparing these compositions with that of Composition C in figure 2 C, it may be seen that they contain larger, but fewer, inclusions.
A comparison of the as-drawn and the annealed microstructures of each alloy in figures 3 and 4 shows that there is a tendency for· the acicular appearance of the matrix to be partially eliminated. No other changes were apparent during examination under the light microscope at magnifications up to 800X.
EXPERIMENTAL PROCEDURE
Evaluation of the unique properties of Nitinol required the development of special testing techniques. In general, the type of apparatus utilized consisted of commercially available equipment. Preliminary tests were conducted to establish the most acceptable methods that would produce meaningful and repeatable data.
As shown in table I, the test program consisted of evaluation of the following mechanical, thermal, and electrical properties:
(1) Stress-strain characteristics in tension and compression as a function of temperature
(2) Shape recovery from tensile, compressive, and bending deformation as a function of initial strain level
(3) Recovery stress (or force) from tensile deformation as a function of initial strain level
(4) Mechanical work potential from tensile deformation as a function of initial strain level
(5) Shape recovery fatigue, or degree of recovery from tensile deformation as a function of percent strain and number of cycles
(6) Thermal cycling effects, or degree of recovery from exposure to temperature extremes as a function of percent strain, number of cycles, and degree of restraint
(7) Uniaxial dimensional change as a function of temperature
(8) Electrical resistance as a function of temperature
Specimens were tested in wire and foil forms. Wires were 100, 20, 15, and 10 mils in diameter. The foil specimens were 3- and 6-mil thick and were tested in O. 12- and 0.25-inch wide strips, respectively. Three Nitinol compositions having different transformation temperatures were tested. Wire was received in coils, and when uncoiled retained a curvature. After the test specimens were cut to length, they were strain-relieved by submerging them in boiling water.
10
Exceptfor bendingtests, all testingwasconductedin an Instronuniversal testingmach-ine equippedwith anX-Y chart recorder systemcapableof recordingload versus time orstrain. Interchangeableloadcells permitted selectionof proper cells for maximumaccuracy.Theload-measuringsystemwascalibrated usingdeadweights. Strains at elevatedor reducedtemperatureswere measuredwith a linear variable differential transformer (LVDT)typeex-tensometer_BaldwinModelPS-6M, modifiedfor useat temperaturesrangingfrom -320° to+500°F. TheLVDT extensometerandanInstron strain gageextensometer,ModelG-51-13,were usedfor roomtemperaturetests. Bothextensometersare accurateto within +0.5 per-cent of full-scale range. Tests at temperatures other than ambient were conducted in acontrolled-temperature chamber capable of providing temperatures from -300 ° to +1000°F
with an accuracy of ±l°F.
Wherever possible, testing was conducted in accordance with the following standards:
(1) ASTM E8-61T, "Tension Testing of Metallic Materials"
(2) ASTM E9-61, "Compression Testing of Metals"
(3) ASTM D1906-62T, "Test for Gage Length in Single Fiber Testing"
Testing for properties such as shape recovery, recovery stress, mechanical work poten-
tial, shape recovery fatigue, and thermal cyclic fatigue necessitated the supplementation of theabove standards with special test methods. Extensometers used to measure initial strain and
degree of recovery were mounted on test specimens and insulated electrically to allow heatingof the specimen while passing electrical current through it. Mechanical work was measured by
applying a constant force with either a pneumatic actuator or dead weights, measuring theamount of displacement with an extensometer, and recording both force and displacement on theInstron machine recorder. Thermal cycling effects required that strained specimens be held
rigidly during cyclic exposure to temperature extremes. This was accomplished by winding thewire specimens over a mandrel of the necessary diameter and spot welding the ends of the
specimen to the mandrel. Degree of recovery after thermal cycling was calculated using themeasured radius of curvature of the tested specimen after transformation recovery. Table IV
summarizes the specimen lengths, gage lengths, load rates, and percent of initial strain where
applicable to each of the types of test.
EXPERIMENTAL RESULTS AND DISCUSSION
Direct Observation of Transformation
General. - A number of experimental methods can be applied to the quantitative observa-tion of transformation. Direct observation during transformation obtained by monitoring a
change in some physical property as a function of temperature has been used successfully byinvestigators (ref. 8). Properties often monitored include electrical resistance, uniaxial di-mensional change (controlled dilation), and mechanical damping. In the present study, quanti-tative data was obtained by applying the first two methods. The third method, although not used
in this program, has been successfully applied by NOL to establish the approximate M s and A s
points.
Electrical resistance versus temperature. - Equipment was set up to measure electricalresistance as a function of sample temperature so that an accurate determination of the
11
TABLE IV. - SUMMARYOF TYPESOF TESTS,SPECIMENSIZES,GAGELENGTHS,ANDLOADRATES
o=
_ Stress-strain
Total Gage Load
length length rate
100-rail 6.0 2.0 0.05dia
20-rail 8.0 5.0 0.5
dia
._ 15-mil 8.0 5.0 0.5
dia
lO-mil 8.0 5.0 0.5
dia
6-mil x 6.0 3.0 0.2
-_ 1/4
3-rail x 5.0 2.0 0.2
1/8
Tensile Tests
Shape recovery
Total Gage Load
length length rate
6.0 2.0 0.05
7.0 2.0 0.2
7.0 2.0 0.2
7.0 2.0 0.2
6.o 2.o" o.2
Recovery-stress
Total T Gage ]Load
length _h_ rate
6.0 12.0_2.0 0.05
7.0 2.0 0.5
7.0 2.0 0.2
7.0 2.0 0.2
0.5
6.0 2.0 0.2
5.0 2.0 0.2
Mech_tcal work
Total Gage Load
le_h le_h rate
6.0 2.0 0.2
7.0 2.O O.2
7.0 2.0 0.2
11.0 8.0 0.2
6.0 2.0 0.2
5.0 2.0 0.2
Shape recovery
fatigue
Total Gage Load
length length rate
6.0 2.0 0.05
10.0 6.0 1.0
B5
100-rail
dia
20-rail
dia
15-rail
dia
10-mil
dia
6-rail x
.N 1/4
3-mil x
i/8
Compression
Stress-strain,
shape recovery,
recovery stress
Tot_ Gage Load
length length r_e
3.22 2.0 0,5
Bending
Shape recovery
Total %
length strain
5.0 to 8.0 4.0 to 14.0
Thermal
cycling effects
Total Gage % Total
length length strain length
5.0
6.0
4.5
5.5
5.0
4.0 to 18.0
3.0 to 18.0
2.0 to 16.0
1.0 to 10.0
3.0 to 12,0
4.0 3.0 6.08.0
1.0
Dimensional
change
Gage
length
1.0
Notes: !. All lengths are in inches.
2, All load rates are in inches/minute crosshead speed.
Electrical
resistance
BeforeAfter final
optimum annealanneal
Total Total
length length
6.0 6.0
6.0 12.0
6.0" 10.0
6.0 14.0
4.0 6.0
6.0 5.0
12
transformation start temperature in cooling (Ms ) and heating (As ) could be established for each
composition.
The general interpretation of the electrical resistance versus temperature curves isbased in part on the work of Wang, et al (ref. 8). This work, combined with the large quantityof resistance versus temperature data gathered during the present effort, has led to a betterunderstanding of the transformation process. A typical resistivity (resistance) versus tem-perature profile obtained on cold-drawn wire annealed at 932°F (500°C) is presented in figure5. The characteristic changes associated with transformation during cooling and heating areshown. It is important to consider here that the general shape of the resistance versus tem-perature curves can be significantly altered by prior cold work or area reduction (in the caseof drawn wire) and final annealing temperatures.
As the program progressed, the usefulness of the resistance versus temperature pro-files was expanded beyond that of determining M s and A s points. These profiles were ulti-mately used as a reliable indicator of relative shape recovery performance.
Prior to the investigation, it was noted that shape recovery performance quite frequentlyvaried from one wire sample to another even though they were drawn from the same alloy.Deformation and heating would produce rapid and precise recovery in one specimen while show-ing signs of sluggishness and incomplete recovery in another.
Careful investigations indicated [hat post-drawing anneal temperature had a significanteffect on the shape recovery behavior and that a small variation in temperature could have amarked effect on performance. Furthermore, it was established that a relationship exists be-tween the shape of the resistance versus temperature profile and shape recovery performance.Figure 6 depicts this unique relationship. The experimental data for figure 6 were obtained byfirst giving a single wire specimen 10 to 20 percent final area reduction by drawing. Thisdrawn wire was then cut into four sections of about equal length. Each section was annealedfor three minutes under slight tension to yield a specimen with straight configuration. One sec-tion was annealed at each of the four temperatures of 752°F (400°C), 932°F (500°C), ll12°F(600°C), and 1292°F (700°C). After annealing, the resistivity profile of each wire was meas-ured. Each annealed wire specimen was then bent around a mandrel, at room temperature,producing an outer fiber strain of about 9 percent. As would be expected, more elastic spring-back was noted for the lower annealing temperatures. Following bending, each wire was heatedabove the transformation temperature range and allowed to freely recover. The ultimate re-covery behavior was found to correspond to the area encompassed within the resistivity loop.The best recovery was produced by the 932°F anneal, which showed a large area encompassed,while very poor recovery was obtained at the ]292°F anneal temperature. The latter produceda profile where very little area was encompassed within its resistivity loop.
Through the use of an automatic electrical resistance versus temperature plotting sys-tem, highly reliable and repeatable data were obtained on a large number of Composition A,B, andC alloys prior to and during the various drawing stages. Curves were also obtained onthe 6- and 3-rail foils. A summary of the tests conducted is given in table I. A selected groupof curves (reproduced from actual recorded data sheets) is included for reference purposes.The effect of various anneal temperatures (752°F, 932°F, ll12°F, and 1292°F) on 100-mildiameter Composition B alloy is shown in figure 6. These anneal temperatures were also in-vestigated fo_ the wires and foils to determine the final anneal condition for optimum perfor-mance. The resistance versus temperature curves for the unannealed hot-swaged rods of Com-position A, B, and C starting materials are shown in figure 7. Finally, the profiles obtainedon each material type and size following, final processing are shown in figures 8, 9, and 10.Using the technique illustrated in figure 5 for selecting critical temperature points (Md, Ms,Mr', and As) , and applying it to figures 7, 8, 9, and 10, quantitative values were obtained foreach of the materials investigated. A complete summary of these data is given in table V.
13
I--
iii
u
u
LU
!
MF'
SYMBOLS
NiTi (IT) and NiTi (Tl'r) designate structure
variations (phases) produced by diffusionless(martensitic) transformation and are those usedin references 3 and 8.
Md Maximum temperature for martensiteformation by deformation. NiTi (11)structure thermodynamically stableat this point.
!
Mf
Martensite transformation startingpoint in cooling
Temperature where macroscopic atomicshear ceases (An arrest point, sincetransformation can be made to proceed
through mechanical deformation.)
As Start of reverse transformation,NiTi (11T).-_ NiTi (IT)
I
II
METASTABLE ENERGY STATE REGION!
COOLING
II
l>PIC UNIAXIAL SHEAR
MOVEMENT REGION
MACROSCOPIC UNIAXIAL SHEARMOVEMENT REGION
I]1.
HEATI NG .
w
COOLING
Ms As M d
TEMPERATU RE v
Figure 5. - Typical electrical resistivity versus temperature curves fordrawn NiTi wire annealed at 932°F (500°C).
14
I I IWIRE BEND RECOVERY *
752°F (400°C)
I I I.... WIRE BEND RECOVERY * --
_111 llll lllllll IIllJllioo 15o 2oo 250
T[MPERATURE (=_
III I300 350
932°F (500°C)
Z
.,lll_._llll
/
III
I II2°F (600°C) 1292°F (700°C)
L
I
350
WIRES WERE HEATED ABOVE TRANSFORMATION TEMPERATURE AFTER 360-DEGREE BEND.
DEFORMATION REPRESENTS 9% STRAIN IN OUTSIDE FIBERS.
Figure 6. - Effect of post-drawing anneal temperatures on electricalresistance and bend recovery behavior.
15
B
F', I I I
COMPOSITION
HEAT NO. NOL 1015
'i RECESSING ||_[ SW_'_ 850°C
_INAL ANNEAL None
SAMPLE NO. A-2
CROSS SECTION S_ZF 0._05 in. dla,TEST LENGTH 5 L_,
-- -- STARTING RESISTANC_ O,OO_ZI_, OHMS
STARTING TEh_, 8'1
REMARKS
TEMPERATURE (°F]
COMPOSITION A
°Im
_L
_1111 I11t
-1_ °_
STARTING RESISTANCE .,-_,_}..L_QHMS
STARTING TEh4P, 8_)FRE,V_ARKS
IIII IIII IIII lllllllllllllllll I I
50 100 150 _0 250 300 350
TEMPE_ATURE (°F)
COMPOSITION B
0
IIII II
'I o__,_ARE;.?TE,,'..____°,
,till IIII IIII IIIII|IIIII[]J]JLL|
._O 100 150 2C0 2_0 30O 350
TEMPERATURE (OF)
COMPOSITION C
Figure 7. - Electrical resistance versus temperaturecurves for swaged rods.
16
"lIlt lIlt
/
llli Illl III
5o
F_TERIAL TiN10"935 C°0.065
COMPOSITION _,
HEAT NO, NO] 1015
PROCESSING i_ot _wa_ed to 0, ]25 in,
_ia, Cold Dra_._n at -4 co -22°F
EINAL ANNEHL $tr,_d Annealed I 600_C
SAMPLE NO. A-4-2
CROSS SECTION S_ZE O, O98 in. dia.
TEST LENGTH 4,5 in,
STARTING RESISTANCE 0.0184 OHMS
STARTING TEMP. 89 _F
REMARKS _?_led tu - 154°F
I J.
IIII
100 150 2_0
TEMPERATURE pF_
IOO-MIL DIA ROD
t111 I|11 I | I I
2.50 300 350
5
O__.L£_L.._L_LLJso J o I I I I-I 50
MATER_AL TiN10"935 C°0"065
COMPOS_T] ON A
HEAT NO ._OL 1015
PROCESSING 1t9 Rolled at 700uC to .0.040 in Warm R.lled to 0,O06 in,
A S00oc 5 rain A C 'FIN L ANNEAL ..... _:'_ .....
SAMPLE NO A- 5COC
CROSS SECTION StZE O.25 X O.O06
TEST LENGTH _ in.
STARTING RESISTANCE 0.0716 OHMS
STARTING TEMP _ _FeEr.,.,a RK S
t
III ILL[ till I I ! 12_ol [ lJ-_oo-J-J-J. L_oo }5c 2oo
TEMRERATURE _°F)
6-MIL THICK FOIL
'ii0
MATERIAL TIN10" 935 C°0" 065
COMJ_OSI1] ON
H_L-T NO, , HI? L 1015
.... PROCESSING Hqt l_olled at TS0°C to0+030 in. Warm'R_]]_:l 1o 0.003 in.
FINALANNEA[ 600°Ct S nlLq+lA, C.
SAMPLE NO. A-600C*
CROSS SECTION SiZE O. IZ3 X 0.003 in,
TESTI.ENGTH 1.'_ U li_- STARTING RESISTANCE ..... OHMS
STARTING TEMP. 85 o F
I R_M_.I_rs C_led to -IS4°FF_nl _tOUD fh3al processed for testl/_
ii15o Ill II|I100 1:4J 200 250 300 350
TEMPERATURE {o_
3-MIL THICK FOIL
Figure 8. - Electrical resistance versus temperature curvesfor Composition A materials.
17
IZ-- I
6
a
_7
_AATEDJ AL TINJ
CO_tPOSIh ON B
HEAT NO. 0-4028
P_CCESSING old Dr_wn 10% Area
F_NAL ANNEAL tied Anlwaled I fi0o_CSAMREE NO B-_
/
STARTrNG RESISTANCE 0,022_ OHMS
STARTING TEMP. 83
REM._RKS --
L
_, _
IOO-MIL DIA ROD
-ltil{llll
I/
f
IlJl II11 IIII
0
I I I
MATERIAL TINI
COA_POSIT_ON B
HEAT NO. D-40_8
PROCESSING Cold Drawn, 20'r_ AreaReduction Between 70Ooc Anneals
FINAL ANNEAL _Rra_Id AnneaTed, S00oC
_lut4PtE NO. B-9
CROSS SECTION SIZE 0.0145 In, alia.
TEST LENGTH 5.25 ua.
STARTING RESISTANCE I. 169 OHMS
STARTING TEMP. 89 _E
REMARKS F --
II11 Illl IItt ttl I
I00 1_0 _ 250 300 .150
TEMPERATURE {oF)
15-MIL DIA WIRE
lli 1,,i_. MATERIAL TiNL
I COMPOt TION t3
HEAT NC D-402B
_ PROCESS +G ttol R_lled at 700°C ta0,040 { • W_rm E l|edlo 0.0_ in,
_) I_NAL ANNEAL.._OoC, S mln,, A.E.
SAh_LE NO. I}- 500CCROSS SEC_ON SiZE 0.2S_X O, 00_ _n.
TEST LENGTH 3 _;
-- ---- STARTING RESISTANCE ___OHMSO - STARTING TEMP. 97 _F
_ _ _ _" ......
TEMP ERAWjRE _
6-MIL THICK FOIL
8
r
L .
¸ ! :
-I
- I_" F I I I
- ]_____ _ MATERIAL TINI
- _ [ HEAT NO D-4028
_ PROCESSINIG C!)IH Drawg_ 20 r Area..... }] I ]Reduction Between 7GO('C Antlt';lls
F
t \ 5J_haPtE NO n-2• CROSS SECTION SIZE 0,020 In. dia,i
- ,L _ rEST L_NGT_ I-I_/_ In.- t StArtinG RESISTANCE _ 2 _O OHMS
• STARTING TEMP. 8_ I F
8 ...... ;......... _L .... _ ......
tr ....... - --
,/ 20-MIL DIA WIRE
M_TERtAL TINI
COMPOSITION B
_f_T NO D-4028
....... _ROCESS_4G C_Ld Dlawq_ 20_ Area
Reduction Betwccn 700eC Anneal_
FINAL ANNEAL Stralld Annealed t _O0OC
_?LE NO B-BCROS_ _ECTION SIZE 0.0|] in. di,_.
TEST LENGTH 2.75 ill.
STARTING RESISXANCE J_?d}J_OH_S
_ STARTING TEMP _._____F
I _EM_RKS --
......__ ....
. ,.,.L_LL_;TEMPErAn.IRE _F_
tO-MIL DIA WIRE
I
...... -f
L
.,_/Z', ___;
t_'AT ERIAL TIY_
COMPOSITION __/_ HEATN O p-4O2B
¢ t P_OCESS_NG Hot Rolled at 750_C to
/--_--- -- , (2_(_30 bl ............... 0 ......
FINAl ANNEAL 5DO°CI S m_n. A,C+SAMPLE NO, _3- S00C
CROSS 5ECnON SIZE O. 124 _( 0.0O3 In.TEST LENGTH 1.75
- , _ -- STARTING R_SISTANCE 0_22g OHMS
STARTING TEMP 83_F
REMARKS _nal reducUon In 0003 In,
_ Processed for Testing
,/
._LLL/J_LI llllltllllillll Ill
TEMPERATURE _F)
3-MIL THICK FOIL
Figure 9. - Electrical resistance versus temperature curvesfor Composition B materials.
18
_
8 ->
I I I
I _TERIAL TiNt
/I co.os,T,ONcHEAT NO, D-4006
/%I .... OCESSING (_ol ..... 'n. E0?L AleaI_vducLion Between 700nC Anneals
E NAL At" NEAL Strand Anneal._..500°C
CROSSSFCTION SIZE 0.096 in, d_a.
- STARTING TEMP. 79 OF
_ i ..... s
TEMPEr_J RE (OF)
IO0-MIL OiA ROD
I I |
MATERIAL TiNi
COMPOSITION C
HEAT NO. D-_006
.... PROCESSING Cold Draw 20', Area
z FINAL ANNEAL _r_nd Am_ealed# 000°C- r_ _.LENO,c-u--_ _ CROSS SECTION SIZE 0.0140 in. dla.
_ ----i- TEST LENGTH. 5,25 in.
STARTING RESISTANCE L]_..___OHMS
STARTING TEMP, _3 °F
i REMARKS
8/
% /P_\
_LLkJJJXoU._J-I 50 I_ 150 200 250 3_ 3_
TE_PERA_JRE (°F)
15-MIL DIA WIRE
i _ T
MATERIAL TiNl
COMFOS;TION .-._--__ ____
HEAT NO. D-400_
____ __ PROCESSING _','_ T_oHedal 700oc to ....
FINAL A_NEAL_ $. rain. A.C.
SAMPLE NO. C-500C
CROSS SECTION SIZE 0.25 _ 0.000 Ln,TEST LENGTH _ i_,
........ STARTING RESISTANC_ _OHMS
STARTING TEMP, _R
Fin.3] re_l_ to 0.006111._EA4ARKS
_-- ....... _..*.- ........
'. .......I __-Ioo -_0 0 5O 100 150 2oO zSO 3OO 350
TEMPEkATU_E (OF)
6-MIL THICK FOIL
I
I . \8
II
"-- , I I I
k,V_TERIAE TiNi
COMPOSITION C
HEAT NO. D-40_6
PROCESSINO_Z_'awm 2Q' Are_
Rgdu_th,n Bvtween T0,3oC Ann_al_
FTt4AL ANNEAL Stra_J AnneaIed I 50(Y'C
F_MPL E NO _'-Jg
CROSS SECTION SIZE " J_
L-- 'EST_ENGT"'-__i_,,,,,-_."STARTING RESISTANCE 0.2_0 OHMS
STARTtNG TEMP. 78 °F
" -7 o, w,RE-- -- t I I
• MATERIAL TiNi
i COMPOSITION C
HEAT NO D-400_
' _OCESgNG__' 0W Ar_a
_ FINAL A NNEA L _d._-_q0°C._--
--LENOLI ' CROSSSEC_,_> I TEST LENGTH 2,"IS in.
-- _ ..... 4 -4-- _ STARTING REST_TANCE |, ET_...--_._--OHMS
[ [ ", STARTING TEMP. 9_ "F
.2TEMPERATURE (°F)
IO-MIL DIA WIRE
8
'?,
9
- MATERIAL Till
- COMPOSITION
- HEAT NO. p'4096
PROCESSING gg_ l_oi|ed at 750"C to --
.... 0,030 in, Warm Roiled to0,003 In.
_ FINAL ANNEAL 400°___._n_ A.___
_ SAh_PLE NO _-4C_C-CROSS SECTION SIZE O. 128 X 0.003 i__
TEST LENGTH |l'f_
STARTING RESISTANCE ..,_._OHMS
- , STArrING TEMP. _ _FREMARKS jvb_al t(_I_ct[ol_ Io 0,003 _n.
- accomplished b}' cold rollln[.
- • From grip processed (or _esting.
_ _-"
................. =, -,,-,
"."Y7 3-MIL THICK F01L
t/ !t
Figure I0. - Electrical resistance versus temperature curves
for Composition C materials.
lg
T
TABLE V. - APPROXIMATE Md, Ms, Mf, AND A s CRITICAL TEMPERATUREPOINTS a FOR MATERIALS INVESTIGATED
Critic al
points Hot-swagedbar
COMPOSITION
Md
Ms
Mf'
A_ c)
COMPOSITION
Md
Ms
Mf'
A s
Temperature, OF, for indicated critical point
Cold-drawn wire Rolled sheet
COMPOSITION
Md
Ms
Mf'
As
A
90
-60
(b)
-40/-10
B
160
7O
0
140
C
190
110
60
160
100-mil
dia
I00
-I00
(b)
-60/-10
160
65
-30
110
170
90
0
130
20-mil
dia.:
160
35
-40
95
170
45
-25
115
15-mildia
150
50
-50
110
160
70
-40
130
10-mildia
160
30
-50
100
160
70
-I0
130
6-milthick
100
-80
{b}
-30/-5
150
78
25
125
160
105
40
170
3-milthick
100
-85
(b)
-50/0
160
80
-40
II0
180
100
-10
150
aApproximate Md, Ms, Mf', and A s critical temperatures determined from electrical re-sistance versus temperature measurements.
bMf' below range plotted; therefore could not be determined.
¢ Two values are given since a very large increase in the slope of the heating curve is ob-
tained prior to the decrease point (As). This point is shown first and may representmore accurately the start point for reverse transformation in Composition A material.
Uniaxial dimensional change. - Uniaxial dimensional change (controlled dilation) experi-ments were conducted to gain a better understanding of the transformation process and its effecton the dimensional characteristics of drawn 100-rail rods. In addition, in certain cases thedata provided an accurate measurement of Ms, As, Md, ' and Mf' points that agreed very closely
with those obtained using the electrical resistance versus temperature measurement technique.Figure 11 illustrates a profile of a typical uniaxial dimensional change of drawn wire that hadbeen post-annealed at 932°F. Examination of the heating and cooling cycles showed that as
the wire is cooled from a temperature above the M d point, it contracts in a normal fashionuntil the martensite start temperature (Ms) is reached. The slope of this line agrees favorably
20
z0
z
,x
z
fJ
t
J
Mr' M A_ Ma
TEMPERATURE
Figure 11. - Typical uniaxial dimensional change behavior forComposition B and C drawn wire annealed at 932°F.
with published thermal expansion data. Further cooling below this point results in large uni-axial expansion that terminates at approximately the Mr' point. (Mr' is defined as the limitingpoint associated with spontaneous atomic shear.) Cooling below the Mr' point produces normalcontraction in an amount equal to that predicted by thermal expansion and contraction.
Heating the sample from a point below Mr' produces normal expansion behavior until the
approximate A s point is reached. At this point reverse transformation is initiated and atomicshear (contraction) begins. Shearing continues until about the Md point•
The typical heating and cooling curves shown in figure 11 are those obtained when the
sample is cooled from a temp, erature above the Md point to a point at or below the Mr' point orheated from a point below Mf to above the Md point. ,It should be acknowledged that, in cool-ing. the path M d _ M s is reversible, but M s _ Mf is not. Likewise, in heating, the pathMf _---_ A s is reversible, but As _ Md is not. ReversibIe behavior is obtained when thesample is heated or cooled in the region above the Md point and below the Mr' point. If duringcooling along the path Ms ---_Mf' or heating along path A s -----Md, temperature change ceases
• . , ! 0
and the dlrectiorL of change is reversed, a new path between lines M s - Mf and A s - Md will beestablished.
The previously described behavior was found to be characteristic of the Composition Band C alloys that were cold-drawn and post:annealed at 752 °, 932 °, and ll12°F. Furthermore,it was established that the magnitude of spontaneous dimensional change diminished as annealtemperatures were increased. However, it was found that at 1292°F, the highest post-drawinganneal temperature, little or no change occurred in the sample. This relationship is clearlyseen in figure 12.
21
Lu
=uu.so.
z"O
z
_<
z
i.u
LUO-
z"O
z<
LU
.<z
0.8
0.6
0.4
0.2
0
-0.2
-0.4 ---
i
i
i I._ I'_ o ._._- o _
i
\
Specimen Preparation
l O. 1-inch dia w_re cold-drawn
{'ram O. 2-1nch dla hot-swagedrod. Final die reduction 10%
(area). Final anneal 752°F
(400°C), 5 rain, alr-cooled.
2 Same as No. 1 except annealedat 932°F (500°C).
...... 3 Same as No. 1 except annealed
at 1112aF (600oc).
4 Same as No. ] except annealedat 1292aF (700°C).
COMPOSITION B
J 1
0.8 _-_----- 1 ............
o.o-l_r i \ ..... i.°-" _,---"""1
\ J_ _ -'\ I_
-0,4-I00
\:
II 'I
II f COMPOSITION C1 _!__ ......... _____i____
150 200 250 300 350
i\. I,,,
4._ _'
- 50 0 50 100
TEMPERATURE, OF
Figure 12. - Uniaxial dimensional change versus temperature curves for100-mil diameter Composition B and C rods.
22
O
z-J
CURVE NO. FINAL ANNEAL, OF TEST LENGTH, IN.
1 752 1.0012 932 0. 994
3 1112 1.001
4 1292 1.0015
5 REFERENCE CURVE, NITINOL COEFFICENT OF THERMAL EXPANSION
- 4 _-_
- 5
-150 -I00 -50 0 50 100 150 200
TEMPERATURE, °F
Figure 13. - Uniaxial dimensional change versus temperature curves for100-mil diameter Composition A rods.
Uniaxial dimensional tests were also performed on Composition A rods post-annealed attemperatures of 752 °, 932 °, 1112 °, and 1292°F. The results of these tests are shown in fig-ure 13. It can be seen that little, if any, dimensional change occurred in any of the samples
tested. Although the reason for this is yet unknown, it is possible that it is linked with thetemperature level at which wire drawing was performed. This temperature, althoughmaintained at -4 ° to -22°F during drawing, was well above the M s point of -60°F deter-mined for hOtlswaged bar. Since it appears that this material was at a higher energy state
during drawing thm_ Compositions B or C, a slightly different structure may have developed.This is confirmed somewhat by the fact that some distinctive variation between drawn Compo-sition A material and drawn Composition B and C materials can be noted in the resistance
versus temperature profiles. It is also possible that a much lower post-drawing anneal rangewould have produced a different behavior.
Mechanic al Behavior
Stress versus elongation curves. - Stress versus elongation curves were obtained at var-ious temperatures, above and below the transformation zone, during heating a_d cooling cycles
for each type of material as indicated in table I.
Each set of data provides important and useful quantitative information required for thesuccessful utilization of each type of wire and sheet in a structural design. Detailed examina-tion of these curves with regard to how modulus and yield point properties change with tempera-ture adds to the overall understanding of the transformation in Nitinol alloys.
23
Twotypical sets of stress versuselongationcurvesobtainedon 100-mil and20-mil dia-meter CompositionB material whencoolingfrom 300° to -100°F (coolingcycle) andwhenheat-ing from -100° to 300°F Cheatingcycle) are presentedin figures 14and15. Themodulusand0.2 percentoffset yield stress obtainedfrom the datain figure 14andpresentedin figure 16showgraphically howthesetwo properties are affectedby temperature.
With referenceto the stress-strain data, the modulusof elasticity (E) andyield strength(Fy) were obtainedfrom the 100-mil diametermaterial, whereasthe ultimate strength (Fult)andultimate elongation(eult)were determinedfrom thewire sizesof 20-mil diameter andun-der as a functionof temperature.
Referring to figure 16A, it canbe seenthat as thetemperatureis loweredbelowabout180°F(the approximateMd point), yield stress is progressivelylowereduntil aminimum isreachedat the Ms point (60°F). The60°FMs point wasverified from the electrical resistanceversus temperaturecurves shownin figure 9. Continuedcoolingbelowthe Ms point producesa small continuousincreasein yield stress.
120
110
,5z_ 1oo
90
N 70
N 50
z 40
b-.-
N_ 20
lO
i i i I
NOTE: STRAIN RATE
- = 0.05 IN./MIN. -
B
$OO°F
/20001.
160 °F
/l _ 150°F
/L _ 140°F
IA _ 130°F
-- _ 120°F
/7--,,00,y ,OOOF
'oo I/A/- _ 7soF
' (;)CO0 ,NG1 2 3 4 5
ELONGATION, PERCENT
I I
160°F
- 150 °
40OF
32°F1
120o F
-100°F
I1 ._.,,o°F -I
-I
0 1 2 3 4 5
ELONGATION, PERCENT
Figure 14. - Stress versus elongation curves for 100-mil diameter Composition Brods during cooling and heating cycles.
24
200 I I I I I I I I I I I I I I I I l Ilog i00_ F
(A) COOLING
I I J I I I J I I I I I l i
I I I I I I I I I I I ,oo I I l I I120
-30
40
_,t_O t_O
I
3OO
NOTES:
1. CORRECTED GAGE LENGTH=6.65 IN.
2. LOAD RATE=0.5 IN./MIN.
,oo (B) HEATING
I I I I I I [ I I I I I I I [ I I
1 2 3 4 5 6 7 8 9 ]0 ]1 12 13 14 15 ]6 17 18 19
ELONGATION, PERCENT OF CORRECTED GAGE LENGTH
I
300
2O
Figure 15. - Stress versus elongation curves for 20-mil diameter Composition Bwire during cooling and heating cycles.
25
=
13
12
T_
1o
9
b 6
3 3
1
13-i
121-I
k)
Z II -
I0-
9 -
Z
O 7-
b 6
G 5 -
-I,N 4
-_1
O 2 -
1
0
130 I I I I I I I I I
!-_ II0 -
I00 -
90 1- _ MODULUS
_ 80J- o 0.2PERCENTYIELDSTRESS / _
/ -.o, -
- o I_ i _ 'l J t I _ I J JMs M d
130[ 1 I I I I I I I
MODULUS
O 0.2 PERCENT YIELD STRESS
Md
(B) HEATING
I I
3OO
TEMPERATURE, OF
Figure 16. - Yield stress and modulus versus temperature curves for 100-mil diameterComposition B rods during cooling and heating cycles.
26
Increasing tile temperature from a level considerably below the M s point produces a
small gradual decrease in the yield stress until about 100°F (the approximate A s point). Be-yond this point, yield stress sharply increases until a plateau is reached at about 180°F (the
M d point).
Since it is known that decreasing temperature leads to thermodynamic instability, it would
be expected that elastic modulus would also decrease as the M s point is approached. The pos-tulated behavior is confirmed in figure 16A, which shows that modulus declines as temperature
approaches the M s point. However, an anomalous rise in modulus, starting at approximatelythe M d point and continuing to 120°F, precedes a sharp decline to its minimum level at the M spoint. Continued cooling produces only a slight increase in modulus until about -30°F. At this
point a rise in modulus is noted in the data obtained. The -30OF point appears to coincide with
the Mr' point derived from the electrical versus temperature measurements.
Heating following cooling considerably below the M s point, as shown graphically in figure
16B, does not significantly change the modulus until the temperature is increased above 100°F,
the A s point. At this point a slight increase in temperature greatly increases the elastic mod-ulus. Examination of the data shows that modulus increases are largely confined between the
A s temperature and about 180°F, the M d temperature.
The preceding discussion describing the behavior of 100- and 20-mil diameter Composi-tion B materials applies, in general, to the other compositions and forms tested.
Degree of shape recovery. - Testing was conducted to determine the ability of Composi-tion A, B, and C wire and foil to recover to its annealed shape after being plastically deformed
near its M s point. No external load was applied during recovery.
Tensile recovery: A typical tensile recovery test sequence consisted of uniaxially elon-gating a specimen to a predetermined strain level. The straining force was then removed andthe amount of elastic recovery (spring-back) was measured and recorded. Composition Amaterials were deformed at -65°F while Composition B and C materials were deformed at 75°F.
This data is summarized in figures 17, 18, and 19, which show both elastic recovery(measured at the deformation temperature) and the total degree of recovery (consisting of elasticand thermal recovery) as a function of strain. The recovery limit was found to vary between6 and 8 percent strain for the differe,t compositions of wire 100, 20, 15, and 10 mils in dia-meter. Variations, although not large, are thought to result from the different processing his-
tories, such as the amount of cold-drawing prior to final annealing, etc. The effects of pro-cessing are more clearly seen by comparing the performance of Composition B and C foil ma-terial with that of the wire. The foils that had Httle cold reduction prior to final anneal exhibited
full recovery up to about two percent strain level. The behavior of Composition A foil was
superior to that of Composition B and C foils in that it recovered completely up to the six per-cent strain level. This behavior was not unexpected, since the cooling and heating cycles of the
electrical resistance versus temperature curves (fig. 8) encompassed more area than the B
and C foil material (figs. 9 and 10). Performance differences, however, may not result totallyfrom processing history variations and may be partly attributed to the composition of the alloy.
Compressive recovery: Recovery tests similar to the tensile recovery tests were alsoconducted in the compressive mode. The test data obtained on 100-mil diameter CompositionB and C rods are presented in figure 20, which shows the degree of shape recovery as a functionof initial compression strain. Comparing the data obtained on the two compositions, it is seen
that full recovery from compression was obtained up to about four percent strain in the case ofComposition B, while Composition C was somewhat less, since considerable degradation in re-covery occurred after about two percent strain.
27
z_ HEAT _"l x_
IF-"
-_ "- 6m
.13
Initial strain (_)= __6 x 100L
_$
Spring-back recovery (Rs) = -_- x 100
Total recovery (Rt) = 6s + 6m x 1006
Z
--,I
F--
LJ-
O
ZI.LJ
U
Ill
I.DI
U
"l-
u.l
ZI--
100
9O
80
7O
60
50
4O
30
2O
10
1" I I I I I I
100-MIL DIA ROD
O 6-MIL THICK FOIL
THERMAL AND ELASTIC RECOVERY
ELASTIC RECOVERY
I I I
2 4 6
,,....,,O. ,_
NOTE: STRAIN TEMPERATURE =-65°F. -
I I J I
8 10 12 14 16
INITIAL STRAIN, E, PERCENT
Figure 17. - Tensile shape recovery versus initial strain curves for
Composition A materials.
28
_L ......
E_ ,.F--
L"_-_m
Initial strain (e)=' 6TxlO0
6sSpring-back recovery (Rs)= _- x 100
Total recovery (Rt)= 6s + _m x 100
Z
_" I00I--I,/I
< 90i,,,-
-- 80
0_ 70Z
u,_ 60
_ 50
_ 4O
" 30
20
l0ZiiiI--
0
I I I I I I I
_l • MLDAR D_l A 100-MIL DIA ROD (B-6) _h _ _._\-11. o 20-_LD,AW,RE " _ \ "_. -
_ O 15-MiLDiAWiRE _ \_
- _t o IO-MILDIAWmE _ _ \_ _0 0 6-MIL THICK FOIL _ _
_ - THERMAL AND ELASTIC RECOVERY I
- _. _ "---'--ELASTIC RECOVERY _ I_
- -,, .-___-_.o- I- '---_" NOTE: STRAIN TEMPERATURE = 75°F. 1
I I I [ [ I I
2 4 6 8 10 12 14 16
INITIAL STRAIN, _, PERCENT
Figure 18. - Tensile shape recovery versus initial strain curves forComposition B materials.
2g
.D
-13
_m
.J
Irfitial straTn (_): _- x 100
6sSpring-back recovery (Rs) = _- x 100
_s t- _m x 100Total recovery (Rt)- 6
100-MIL DIA ROD (C-4)
& 100-MIL DIA ROD (C-6)o 20-MIL DIA WIRE
[] 15-MIL DIA WIREO 10-MIL DIA WIREO 6-MIL THICK FOIL
THERMAL AND ELASTIC RECOVERYm _ i ELASTIC RECOVERY
Z
lo0
__ 90I--
z-- 80u_
0_- 70Z
U 60u_l
ill
40Uu.f
" 30
.<-i- 20
_ 10zI--
0
I I
RUN 1
\
I I I I I
RUN 2
RUN 2
RUN I
_ NOTE: STRAIN TEMPERATURE = 75°F.J I I [ I I I
2 4 6 8 10 12 14 16
INITIAL STRAIN, _, PERCENT
Figure 19. - Tensile shgpe recovery versus initial strain curves forComposition C materials.
3O
100
9O
_Z 8O
Ou 70l_ ..,I
uJ<
,,, _ 50>
_,,, Z 40Q-U
_ 30OU
20
10
I00
9O
80
u 70
w<
•" _, 60<i12
-, _ 50>m _
m,,,,,Z, 40,, U
_, 30OU
2O
10
I I I T I
A A
COMPOSITION B
I I I I I
I I I 1 I
COMPOSITION C
I I 1 _L__._L
2 4 6 8 10 12
COMPRESSIVE STRAIN, PERCENT OF]NIT1AL LENGTH
z_ 100-MIL DIA ROD
THERMAL ANDELASTIC RECOVERY
ELASTIC RECOVERY
Initial strain (e)= _ x 100L
Sprlng-back recovery (Rs)
as= --xl00a
Total recovery (Rt)
a s + am-- x 100
Figure 20. - Compression shape recovery versus initial strain curves for
Composition B and C rods.
31
o
w"-I-UZ
u_lI,--w
Zw
Z
O
Y
g:D
_D
O
!
v-I
g O
IN3D_i:::Id '_3_1:1 _lAO NI NIV_IS
O
32
Bend recovery: Of the possible deformation modes, bending perhaps holds the most in-terest because it is through this means that a large surface area structure, such as a parabolicantenna or passive communications satellite, could be tightly packaged for subsequent self-
erection in outer space.
Bend recovery data were obtained on 100-mil diameter Composition A wire and 100-,20-, 15-, and 10-mil diameter Composition B and C wire. Data were also obtained on 6- and3-mil Composition A, B, and C rolled sheet (foils).
Tests were performed by bending the specimens 180 degrees about mandrels of varyingdiameters. Actual mandrel diameters were selected so that the effect of varying degrees of
outside fiber strain could be evaluated. Figure 21 shows how outside fiber strain varies as afunction of inside bend diameter. Following deformation, the specimen was released and the
elastic recovery (spring-back) was measured. Composition A specimens were bent at -65°Fand Composition B and C specimens were bent at 75UF prior to heating. Martensitic recoverywas then brought about by increasing the specimen temperature until full recovery was obtainedor until it was noted that an increase in temperature did not produce any additional recovery.
Where possible, simultaneous measurements of sample temperature and degree of recoverywere taken so that the degree of recovery could be closely related to specimen temperature.Due to the fast heating response obtained when 6- and 3-mil thick Composition A specimenswere exposed to room air temperatures, simultaneous measurements on the group were notobtained.
Figures 22, 23, and 24 show the bend recovery of Composition A, B, and C specimens
OO
X
LU
uuJ
p-
ZLU
UevLufL
100-
90-
80-
70-
60-
50-
40-
30-
20--
0-80
I I I I I I I
/
T 7 oo:: X'2!I F�IId 0 r = 0.450" (10.00)P //" O r = 0,365" (12.00)
/_ V r = 0.305" (14.00)
-" _RS SHOWN IN PARENTHESESDESIGNATE PERCENT STRAIN INOUTER FIBERS.
-- "STRAIGHT
ANNEALED"
-CR.T, BEND
SPRING-BACK
-- HEAT
R.T.
I I I I i I I
-40 0 40 80 120 160 200 240
TEMPERATURE, °F
Figure 22. - Bend recovery versus temperature curves for 100-mil diameterComposition A rods.
33
"STRAIGHTANNEALED"
R.T. BEND
SPRING-BACK
60 100 140 180
TEMPERATURE, OF
20-MIL DIA WIRE
HEAT
NOTE: NUMBERS SHOWNIN PARENTHESESDESIGNATE PERCENTSTRAIN IN OUTER
FIBERS.
300
Figure 23. - Bend recovery versus temperature curves for Composition B materials.
34
I I _ I I
1 wJ_
.,.g-
_,1oo 70- A r : 0. 240"
60 - O r : 0.158"[] r = 0.115"-> O r : 0.090"
_ 5 _ O r = 0.073"_: 40 _ r = 0,061"
I"- i ® r = 0.053"
Z
30 g r = 0.046"• r = 0.035"
" 20
10l ._---SPRING-BACK
0 _ l ) I
(2.9)(4.4)(5.9)(7.5)(9.0)(10.6)(12.0)(13.6)(17.4)
15-MIL DIA WIRE
I t
8 8o
× 70
605o
4o
20a.
10 --_ .,_-----SPRtNG-BACK
6O 100 140 180
TEMPERA1URE,
I 1 I
A r = 0.245" (2.0)O r = 0.120" (4.0)[] r = 0.054" (8.4)O r = 0.036" (11.9)
O r = 0.025" (16.0)
10-MIL DIA WIRE
220 260
OF
_00
"STRAI.GHTANNEALED"
CR.T. BEND
SPRING-BACK
HEAT
NOTE: NUMBERS SHOWNIN PARENTHESESDESIGNATE PERCENTSTRAIN IN OUTERFIBERS.
Figure 23. - Continued.
35
OO
X
o>-l.u
8
l--
z
U
a_
gX
8uiii
l--
ZILl
u
100
9O
8O
7O
60
5O
40
30
2O
I0
O
I00
9O
8O
7O
6O
5O
40
30
2O
I0
0 ) I6O I00
I I I I
/r = 0.157" (1.9) -
O r = 0.072" (4.0)[3 r = 0.047" (6.0) -O r = 0.035" (8.0)
O r = 0.027" (10.0) -V r = 0.022" (12.0)® r = 0.018" (14.0) -
SPRING-BACK
I
6-MIL THICK FOIL -
I I
I I I
/A r = 0.043" (3.0)O r = 0.023" (6.0)D r = 0.015" (9.0)
O r-- 0.011" (12.0)
--SPRING-BACK
I
140
3-MIL THICK FOIL --
I I I
180 220 260 300
TEMPERATURE, OF
L D
"STRAIGHT
ANNEALED"
CR.T. BEND
HEAT
NOTE: NUMBERS SHOWNIN PARENTHESESDESIGNATE PERCENTSTRAIN IN OUTER
FIBERS.
Figure 23. - Concluded.
36
100
90
8O
X
6O
u.l
soU
,_ 40
Z
U
I.IJ
4 ....
10RING-BACK
z_r= 1.25o"(3.9)O r = 0.800" (5.9)[] r =0.575" (8.0)
O r = 0.575 '' (8.0)o r= 0.450" (10.0) -
r= 0.365" (12.0)
® r = 0.305" (14.0) -@ r = 0.260" (16.0)
u,.l
L9
t.LI¢}.-
r = 0.061"® r = 0.053"
r = 0.046"
100-MIL DIA ROD -
I l
8 8o "X
t! 7060 _, r = 0.240"O r= 0.157"
u..l
50 cs r = 0.115"
0 r : 0.090""" 40 0 r = 0.073"
60 I00 140
TEMPERATURE,
(4.0)(6.o)(8.0)(i0.0)(12.0)(14.0)(16.0)(18.0)
20-MIL DIA WIRE
L____.__L I180 220 260
OF
"sTRA GHTANNEALED"
CR.T. BEND
SPRING-BACK
HEAT
- NOTE: NUMBERS SHOWNIN PARENTHESES
- DESIGNATE PERCENT
STRAIN IN OUTER
- FIBERS.
_00
Figure 24. - Bend recovery versus temperature curves for Composition C materials.
37
8I---
x
Ill
kJ
Z
U
1DO
9O
8O
7O
60
50
4O
3O
20
10
0
g
x
ILl
r_u.J
I'--
ZuJ
U
u.J0,...
100
9O
8O
7O
60
5O
40
3O
20
10
O6O
,PRING-BACK
I I
I00
I i
A r = 0.240" (2.9)
O r = 0.158" (4.4)O r = 0.115" (5.9)
O r = 0.09O" (7.5)0 r = 0.073" (9.0)"9 r = 0.061" (10.0)O r = 0.053" (12.0)
r = 0.046" (13.6)
& r = 0.035" (17.4)
15-MIL DIA WIRE
I i
I I
Z_ r = 0.245" (2,0)
O r = 0.120" (4,0) _rn r = 0.078" (6.0)O r = 0.046" (9.6)
O r = 0,036" (11,9)'9 r = 0.029" (14.5)® r = 0.025" (16.0)
10-MIL DIAWIRE -
:,RING-BACK J T I J
140 180 220 260 3uO
TEMPERATURE, OF
"STRA GHTANNEALED"
CR.T, BEND
HEAT
R.T.
NOTE: NUMBERS SHOWNIN PARENTHESESDESIGNATE PERCENTSTRAIN IN OUTER
FIBERS.
Figure 24. - Continued.
38
100
9O
c_ 8O
x
7o
ILl
ULLI
l.-
zu.l
U
60--
50-
40--
30
2O
10
0
100
9O
8O
x 7Oo
60
50U"' 40p-
Z_ 30U
a_ 20
10
I I I I I
r =0.157" (1.9)
O r =0.072" (4.0)[] r -- 0.047" (0 r =0.035" (8.0)
0 r = 0,027"r =0.022" (12.0)
6°MIL THICK FOIL
I I
I I I I
z_ r = 0.046" (3.1)0 r = 0.023" (6.0)[] r = 0.015" (9.0)
0 r = 0.011" (12.0)
06O
3-MIL THICK FOIL -
SPRING-BACKi I I I
loo 140 180 220 260 300
TEMPERATURE, OF
"STRAIGHTANNEALED"
CR.T. BEND
SPRING-BACK
HEAT
NOTE: NUMBERS SHOWNI N PARENTHESESDESIGNATE PERCENTSTRAIN IN OUTER
FIBERS.
Figure 24. - Concluded.
3g
as a function of sample temperature. Examination reveals that increasing the temperatureresults in a slight increase in recovery until the A s point is reached, whereupon small tem-perature increases produce large recovery rates. In general, the A s point agrees favorablywith that predicted on the basis of electrical resistance versus temperature tests; however,some increases are apparent in the highly deformed specimens.
Another interesting type of behavior that can be noted in examining these curves is thatspecimens tend to return slightly to their bend shape as the temperature is lowered from thelevel where the highest recovery was obtained.
Figures 25, 26, and 27 show the effects of strain (in the outside fibers) on martensiticand elastic recovery (spring-back) of Compositions A, B, and C.
It can be seen from these curves that complete bend recovery varies from 6 to 10 per-cent outside fiber strain for the types of wire tested. Furthermore, the fall-off is gradualbeyond the full recovery point, so that many of the specimens exhibited as much as 90 percentrecovery at strain levels as high as 14 to 16 percent. Recovery performance of the foils wasconsiderably below that of the wires. It was found that specimens could be strained only 1 to1.5 percent before full recovery was lost. Outside fiber strain in excess of this amount pro-duced gradual but constant fall-off in recovery.
Recovery stress. - To gain a better understanding of the forces associated with the shaperecovery process, tensile recovery stress characteristics of each of the compositions and pro-cessed forms were determined. To accomplish this, each specimen was uniaxially strained(elongated) a predetermined amount. After the load was removed, each wire was heated andthe force (stress) required to prevent recovery was measured as a function of temperature.
z 10o
9o
-- 7Ob__ 6o!l.u
l.uQ_
4O
O 3O.,u_ 20
.,z,_ 10
"STRAIGHT R,T. BEND SPRING-BACK HEAT R.T,ANNEALED"
I .... I I I I I I
THERMAL AND ELASTIC
RECOVERY
_, IO0-MIL DIA ROD
O 6-MIL THICK FOILV 3-MIL THICK FOIL
I I I I I I i I
0 2 4 6 8 10 12 14 16
STRAIN IN OUTSIDE FIBERS, PERCENT
18
Figure 25. - Bend shape recovery versus initial strain curves forComposition A materials.
40
z 100
< 80
- 70
_. 6oILl
sow
4o
3o
o 20z
10
"STR_.tGH [ R.T. BEND SPRING-BACK HEAT R.T.
ANNEALED"
I I I 1 I I I I
o 20-MIL DIA WIRE
[3 15-MIL DIA WIREO 10-MIL DIA WIREO 6-MIL THICK FOIL
x7 3-MIL THICK FOIL
THERMALANDE'AST,CRECOVERY\\ _. ELASTICRECOVERY
-'_._-__. _._ ._.____________----_-v,,. k_-,_- "--- --. ...._._
I I I I_ I I I I2 4 6 8 I0 12 14 16 18
STRAIN IN OUTSIIgE FIBERS, PERCENT
Figure 26. - Bend shape recovery versus initial strain curves forComposition B materials.
z t00
_h 90
__ 80
6o
_ soo-
t.u
_ 2oN lo
"STRAIGHT R.T. BEND SPRING-BACK HEAT R.T.ANNEALED"
! I I I I ! I I
:& ]OO-MIL DIA ROD _C_o 20-MIL DIA WiRE[3 15-M1L DIA WIRE<> 10-MIL DIA WIRE0 6-MIL THICK FOIL
- V 3-MIL THICK FOIL -
k - _ THERMAL AND ELASTIC RECOVERY
_ _ ELASTIC RECOVERY
r- _ ,,_._ ......,,._ -
_'----__ --_'__.:_C_-"----i-_-_
I 1 I I I I t I
2 4 6 8. 10 12 14 16 18
STRAIN IN OUTSIDE FIBERS, PERCENT
Figure 27. - Bend shape recovery versus initial strain curves forComposition C materials.
41
UZ
Uu__o
Z
8O
7O
6O
5O
4O
30
2O
10
I I I I
NOTE: HEAT RATE = 40°F PER MINUTE
I I
6%
4%
INITIAL
STRAIN
0 I I I60 100 140 180 220 260 300 340 380
TEMPERATURE, OF
Figure 28. - Tensile recovery stress versus temperature curves for100-mil diameter Composition B rods atvarious strain levels.
A representative set of curves obtained in this manner is shown in figure 28. These curvessh6w the tensile recovery stress performance of 100-mil diameter Composition B rods atvarious strain levels. It can be seen that higher initial strain produces a higher recoverystress. Furthermore, as initial strain is increased, higher temperatures are required to at-tain maximum recovery stress. A maximum recovery stress of If0 000 psi was recorded for20-mil diameter Composition B wire strained initially to 8 percent. This stress was produced
by a force of 34.5 pounds. This stress was the highest value recorded of all the materials
tested. A summary of the recovery stress performance of each specimen is presented in fig-ures 29, 30, and 31. Maximum recovery stress obtained upon heating is plotted as a functionof initial strain (elongation). From the data it is clearly seen that a maximum strain level (ap-proximately 8 percent) beyond which recovery stress diminishes is inherent in each specimen.It is believed that these variations are the result of the initial processing histories.
Mechanical work. - The ability of each specimen to perform mechanical work was meas-
ured by applying a constant force to the end of a vertical specimen. The force (weight) uniaxi-ally strained the wire. Following straining, the specimen was heated while the force wasmaintained, and the recovered strain (deflection) was measured as a function of temperature.
The product of the deflection times the constant force per unit of material volume (specificwork) was then plotted as a function of temperature for both the heating and cooling cycles. Afamily of curves was obtained for each material by varying the constant load (initial strain).
An interesting observation was made in connection with the cooling cycle. As the speci-
men was cooled, with the constant load imposed, it would elongate plastically beyond the initialstarting length if a suitable stop were not provided. This "super" plastic behavior is initiatedwhen the temperature is lowered to the M s ' point and the specimen becomes very soft and
42
7-UZ
1.1.1
C_
wo_
z
OQ,.
ii
O
z
8"l-
u,,.I
I,-,.
u,,.I
8U1.1.1
Z)
120
100
8O
60
4O
2O
I I I I I I
_ O 6-MIL THICK FOIL
_ 3-MIL THICK FOIL
NOTE: STRAIN TEMPERATURE =-65°F
I I I I I I
0 2 4 6 8 10 12 14
INITIAL STRAIN, PERCENT
Figure 29. - Maximum recovery stress versus initial strain curves forComposition A materials.
43
IUZ
u_l
Or
LI.I
C5z:2)O
U,_
O
C5Z
O.1_I--
LLI
I-
LL#
UIJ_l
2C)
120
100 -
80 -
60 -
4O
2O
I I I I I I
A 100-MIL DIA RODO 20-MIL DIA WIRE
[] 15-MIL DIA WIREO 10-MIL DIA WIREO 6-MIL THICK FOILV 3-MIL THICK FOIL
NOTE: STRAIN TEMPERATURE = 75°F
I L I 1 I I0 2 4 6 8 10 12 14
INITIAL STRAIN, PERCENT
Figure 30. - Maximum recovery stress versus initial strain curves forComposition B materials.
44
-I-Uz
u_lr.r
C_
l.u
c_Z
OQ_
U_
O
z,<
O"I"l--
l.u
>-ill
Ul.urv
x
120
100
8O
6O
4O
I2O
I
I I I I I I
z_ 100-MIL DIA RODO 20-MIL DIA WIREn 15-MIL DIA WIRE
O 10-MIL DIA WIREO 6-MIL THICK FOIL_7 3-MIL THICK FOIL
NOTE: STRAIN TEMPERATURE = 75°F
I I I I I I
0 2 4 6 8 10 12 14
INITIAL STRAIN, PERCENT
Figure 31. - Maximum recovery stress versus initial strain curves forComposition C materials.
45
t
easily subjected to plastic deformation. The M s point is somewhat higher than M s and repre-sents the martensitic transformation temperature for the material under stress.
The data obtained on 100-mil diameter Composition A, B, and C specimens are pre-sented in figures 32, 33, and 34. A close examination of these curves shows that the maximummechanical work output increases with increasing initial strain until about 7 percent. Beyondthis strain level a sharp decline occurs in the maximum mechanical work. A summary of the
maximum mechanical work as a function of initial strain level for the specimens tested is pre-sented in figures 35, 36, and 37. Since there is a considerable variation in the maximum
mechanical work output among the various specimens tested, the relative performance of eachspecimen can be established.
In general, the mechanical work performance of the 20-rail wire was superior to all othersizes. The 20-mil Composition C material recorded the highest value - 2900 inch-pounds percubic inch. The performance of the 15-mil wire closely followed that of the 20-mil wire,whereas the 10-rail wire performance fell considerably below that of the 15-mil wire, but
slightly above that of the 100-mil rod. The 100-mil Composition A rod followed this group;however, it exceeded the performance of the 6- and 3-mil foils, with the 3-mil foils showingthe poorest performance.
The data obtained shows improvement in performance as the result of drawing and proper
annealing and the overall importance that processing had on final material behavior. It is seen
1600
i 14o0U
z
_u 1200co
U
o_ I000
z
0 8OOi
IU
_z 6ooff
N 40O
U
z 200I
0
I I I I I I I I I I
STRAIN STRESS
,_/ (%) (LBIIN. 2)
/_/ _ 3.75 10 8000 4.10 11 800[] 6.00 35 600
I O 7.50 73 500
-200 i I I I 1 I I I-80 -40 0 40 80 120 160 200 240 280 320 360
TEMPERATURE, °F
Figure 32. - Mechanical work versus temperature curves for 100-rail diameterComposition A rods.
46
1600
I- 1400-U
Z
U_-, 12_ -DU
a. lO00-
p,Z -
O 800 -
_ _U
z- 600 -
N 4OO
U
< 200I
,v,0
STRAIN STRESS(b_) (LB/I N. 2)
6 2.0 16 700
0 4.0 19700[] 6.0 39 4000 8,0 75 600
-200 I I I . I I I I I60 100 140 180 220 260 300 340 380 420 460 500
TEMPERATURE, OF
Figure 33. - Mechanical work versus temperature curves for 100-mil diameterComposition B rods.
1600
i 1400uz
_u 1200
U
o_ 1000
z
O 800a.
±U
z_ 6OO
_2
N 40O
U17" 2O0
I
2 0
-20060
I I I I I I I I I I
1 I .........
_./of- /! STRA,N STRESS
I I I I I I I I I I
100 140 180 220 260 300 340 380 420 460 500
TEMPERATURE, °F
Figure 34. - Mechanical work versus temperature curves for 100-mil diameterComposition C rods.
47
3000
zu 2500Z
U
22)U
U_l
a 2000Z
0
!"I-Uz
1500
0
.-.,I
uz< 1000"I-ULid
_ 500
I 1 I I I
NOTE: STRAIN TEMPERATURE ---65°F
A 100-MIL DIA RODO 6-MIL THICK FOIL
V 3-MIL THICK FOIL
0 10 12
INITIAL STRAIN, PERCENT
Figure 35. - Maximum mechanical work versus initial strain curvesfor Composition A materials.
48
3000
'3-u 2500Z1
Ui:n
,Z)U
I.l.IQ_
r-, 2000Z
0!
IUZ1
::2 15oo
0
-..J
<U
Z< 1000-r"UM.J
_ 500
I I I 1 I
NOTE: STRAIN TEMPERATURE = 75°F
A I_-MIL DIA RODO _-MIL DIA WIRE
[] 15-MIL DIA WIREO 10-MIL DIA WIREO 6-MIL THICK FOIL
V 3-MIL THICK FOILm
I
0 2 4 6 8 10 12
INITIAL STRAIN, PERCENT
Figure 36. - Maximum mechanical work versus initial strain curvesfor Composition B materials.
4g
3000 I I I
NOTE:
STRAIN TEMPERATURE = 75°F
zu 2500Zm
U
U
2000Z
0!
IUZ
ff 1500
0
uz
i I000U
500
A 100-MIL DIA RODO 20-MIL DIA'WIRE
n 15-MIL DIA WIREO 10-MIL DIA WIREO 6-MIL THICK FOIL
V 3-MIL THICK FOIL
0 2 4 6 8 10 12
INITIAL STRAIN, PERCENT
Figure 37. - Maximum mechanical work versus initial strain curves
for Composition C materials.
5O
that cold work in theform of drawingandperhapsrolling improvesperformance, the maximumbeingreachedat the 20-mil diameterlevel. Further reductionbeyondthis point appearstolower performancesomewhat. Thereasonfor this is notclearly understood. However, it ispossiblethat a reductionin performanceof the 10-mil diametermaterial is broughtaboutbyoxidationduringannealing. This wouldhavea more significanteffect onfine diameterwireandwouldalso helpexplainthevery low performanceseenin the 3-mil foils.
Shape recovery fatigue, - To gain a better understanding of the repeatability of the ten-sile shape recovery process, testing was performed to measure the effects of a repeated de-formation and heat recovery cycle. Three indicators were used to monitor changes:
(1) Degree of no load recovery after heating above the M d point
(2) Change in the load versus elongation curve
(3) Change in the electrical resistance versus temperature curve
These items will be discussed in more detail later in this subsection. Testing was per-formed on 20-mil wire for the two strain levels of 6 and 8 percent. A maximum strain level of
8 percent was selected on the basis of previous tensile recovery tests (figs. 17, 18, and 19),which indicated that full recovery could be anticipated up to about this point.
Degree of recovery: The degree of recovery as a function of the number of cycles for
the 6 and 8 percent strain levels is shown in figure 38 for Composition B wire. These datashow that, following 6 percent strain, the specimen recovered 96 percent after the first cycleand maintained nearly the same degree of recovery following 10 additional cycles. Following100 cycles, the material still recovered 93.4 percent. Increasing the strain level to 8 percentproduced a significant change in recovery behavior. As can be noted in figure 38, the speci-men recovered only 91 percent for the first cycle and 86 percent after 10 cycles. After ap-
proximately 20 cycles the rate of change declined considerably, so that following 100 cyclesthe sample recovered only 81.percent.
Change in load elongation curves: The overall changes in load versus elongation curvesfor the 6 and 8 percent cyclic strain tests are illustrated in figure 39. Examination of thesecurves indicates that for both strain levels, repeated cycling brings about a small but
ZLLI
L)LUO-
LU
ULUrv
100
95
9O
85
80
/ I I I I i,b......__J. J 6 PERCENT_CONSTANT STRAIN
0 10 20 3o 4o so 6o 7o. 8o
NUMBER OF CYCLES
1I!
90 100
Figure 38. - Tensile recovery versus number of deformation cyclesfor 20-mil diameter Composition B wire.
6 and 8 percent strain.
51
I_ 1 1 ] 1 I 1 1 t I 1I t I I I I 1
,Z
IO
0
Z
0
I
1J
Ittt111
SaN flOd 'aVOl
u
Z
U
< o
zZ 0 ,r-4
U I--
b z bG o _...a _
_._
,,_ "_ _O
<_ _0
Z_
o_ _o_
<o o;_
U _e_I
g _,-:N _g __ g
g
52
significant reduction in the yield point characteristics. This change appears to progress atabout the same rate as the recovery degradation. In addition, it can be noted that a smallanomalous change in the linear portion of the load versus elongation curve disappeared afterabout 10 to 20 cycles.
Electrical resistance versus temperature curves: The electrical resistance versus tem-perature curves shown in figure 40 were obtahled on the 6 and 8 percent strained wire follow-ing 20 and 15 deformation-recovery cycles, respectively. Curves were not obtained on wireswith a greater number of cycles. In general, the cycling produced little change in the charac-
teristics. The M s point for the material remains unchanged; however, a small increase ofabout 15°F is noted in the A s point when the data is compared with that of the uncycled wire.
In summary, it appears that repeated 6 percent cycling produces a very small, but meas-
urable, degradation in tensile recovery after the first cycle. At 100 cycles the wire recovered93.4 percent of its initialgage length. Increasing strain to 8 percent brings about a significantincrease in recovery degradation in that, following the first cycle, the specimen recovered o_.y91 percent and continued to decay to 81 percent recovery at 100 cycles.
Thermal cycling effects. - A series of tests was performed to determine the effect ofcyclic exposure to temperature extremes on shape recovery behavior. All testing was con-ducted on Composition B and C 20-mil wire.
Testing consisted of subjecting specimens to temperature extremes while beil_g held in apredetermined strained condition. A test temperature range of -100 ° to 300°F was initially
selected, but was later modified to -320 ° and 160°F when it was proven, through preliminarytests (one cycle), that 300°F exposure destroyed a major portion of shape recovery. Loweringthe temperature to 212°F brought about less degradation, but still proved to be unsatisfactory.After further testing, 160°F was selected as the upper limit, since it produced insignificantdegradation following a single cycle exposure. Recovery after exposure to -320°F was com-pared with that obtained after -100°F and found to be identical.
Two different levels of deformation in bending (6 and 8 percent outer fiber stress) and
three different degrees of restraint were examined as a function of temperature cycles. Theywere as follows:
(1) Specimens were fully restrained so that the initial 6 or 8 percent outer fiber
stress was maintained throughout the cyclic temperature exposure.
(2) Specimens were permitted to recover elastically (spring-back) after bending, thenrestrained at the recovery point throughout the cyclic temperature exposure.
(3) Specimens were plastically deformed in bending, allowed to spring back, then
heated slightly to provide partial martensitic recovery before being restrained.
The degree of strain maintained under each of these restraint conditions is given intable VI.
Specimens were subjected to 0, 1, 10, 50, and 100 thermal cycles by dipping them alter-nately in liquid nitrogen (-320°F) and hot water (160°F). After the desired number of cycles,^specimens were removed from their restraining mandrels and submerged in boiling water (212UF)to bring about martensite shape recovery. They were then removed from the water, allowedto assume room temperature, and measured to determine degree of recovery. It is importantto note that the degree of recovery at 75°F (room temperature) is somewhat lower than the level
53
:
I I I IMATERIAl TiNi
COMPOSITION BHEAT NO. D-4028
PROCESSING Cold Drawn I 20_ AreaReduction Between 700°C Anneals
FINAL ANNEAL Strm_d Annealed I 500oc
SAMPLE NO. B-2CROSS SECTION SIZE 0.020 in. Dia.TEST LENGTH 2.0 in.
STARTING RESISTANCE 0.24b OHMS
STARTING TEMP, __8_._0= °FREMARKS Strained 6/_ and Heated. 20
C_'cles Total.. (Shape Recoverer Fatibme --
Test)
I I
IIII IIII llll IIII It[l I I I I I I I I50 100 150 20O 25O 3OO 350
TEMPERATU RE (°F)
(A) 6% STRAIN, AFTER 20 CYCLES
54
ZO --
\- /
o,% _
-I1_/_ I 11 I
-loo -50
MATERIAL TiNi
- COMPOSI TION BHEAT NO. D-4028
PROCESSIN'_-'_ br_n, 20_,_ AreaReduction Between 70Ooc Anneals
__ocSAMPtENO. B-2TEST LENGTH 2.0 in,-----_
STARTINGRES_ "_'_SSTARTING TEMP. 93 :3FREMARKS Strained 8E_,and heated. 15
cycles total (.Shape Recover_t Fatigue --Test)
II11 1111 1tll llll I 1 II I I I 1 I I I 1
0 50 100 150 200 250 300 350
TEMPERATURE (oF)
(B) 8% STRAIN, AFTER 15 CYCLES
Figure 40. - Electrical resistance versus temperature curves followingrepeated deformation cycles for 20-rail diameterComposition B wire,
TABLE VI. - DEGREE OF STRAIN MAINTAINED IN THERMAL
CYCLING SPECIMENS
Initial strain,percent
6
8
Fullyrestrained
6.0
8.0
Outer fiber strain, percent
After elastic
recovery
After partial
memory recovery
reached when the sample was placed in boiling water. The behavior was previously noted and
can be clearly seen in the data presented in figures 23 and 24.
The results of the tests performed on 20-mil Composition B and C wire are summarizedin figure 41. The curves show the degree of bend recovery as a function of cooling-heatingcycles.
Examination of the data obtained reveals that the 6 percent strain condition produces lessinitial degradation in recovery for both compositions than the 8 percent strain condition.However, after 100 thermal cycles in both strain conditions, recovery is about equal. The high-est recovery recorded after 100 cycles was 95.8 pel"cent for Composition B and 95.3 percentfor Composition C.
In general, the completely restrained condition produces a greater initial degradation butbecomes better than the other two constrained test conditions beyond about 10 to 20 cycles.
Following 100 cycles, Composition B specimens, restrained following partial recovery, indi-cated a slight improvement over the elastic recovered specimens. This may not represent atrend since the same restrained condition produced identical behavior in the Composition Cwire.
It has been concluded that while the magnitude of change brought about by thermal cyclingbelow 160°F is small - about 2 percent loss for 100 cycles - a trend exists that should be con-sidered in any future design. Most important is the serious degradation to recovery (memoryloss) that occurs if strained materials of the type examined are subjected to temperaturesabove about 160°F.
Although heating much above 160°F will produce measurable degradation, it should not beconstrued as a sharp limiting point. While the mechanism surrounding this behavior is notclearly understood, heating above the M d point (approximately 160°F) appears to introduce aslightly different "remembered" shape and should be avoided whenever a specimen is con-
strained in a configuration other than that originally introduced.
55
100
99-- --
97 _ --
96 - _
95
94
---i
i '. (A) 6% STRAIN, COMPOSITION B
99 ....................
98 --
,,u 97' _ - -_ 96 _I.__ _ _
94 --
lI
(B)8%
IIPlIII I i I I I
STRAIN, COMPOSITION B_ __I L____
llrllll(C) 6% STRAIN, COMPOSITION C
I I I I
I
] _ ] _l ...... ! _ j _ I ] I_ 1--1 l--
0 COMPLETELY RESTRAINED
98 [3 RESTRAINED AFTER ELASTIC RECOVERY
,', _RESTRAINED AFTER PARTIAL RECOVERY
10
(
96
95
940
(D) 8% STRAIN, COMPOSITION C
20 30 40 50 60 70 80 90 100
NUMBER OF THERMAL CYCLES
Figure 41. - Thermal cycling effects on bend recovery for20-mil diameter Composition B and C wire.
56
CONCLUSIONS
Engineering data has been obtained on typical 55-Nitinol alloy compositions in sufficientquantities and detail to permit NiTi utilization in practical design concepts. Furthermore, ithas been proven that when careful processing procedures are utilized, reliable and consistentshape recovery performance can be obtained over a range of material compositions and wiresizes.
Careful processing of the various wire and thin sheet forms has led to a greater under-standing of the variables that must be controlled to provide good recovery behavior and hasfirmly established the relationship between post-drawing annealing temperature and shape re-
covery performance.
Through the widespread use of electrical resistance versus temperature profiles as a
transformation temperature (Ms and As) monitor tool, it was possible to broaden the under-standing of how processing affects transformation, starting temperatures, and overall recoveryperformance to such a degree as to make possible qualitative and quantitative predictions of the
wire performance on the basis of preliminary small ingot sections.
Therefore, the following conclusions can be drawn from the results of this experimentalinvestigation:
(i) Nitinol alloys having M s temperatures that are near room temperature can beprocessed readily into fine wire and thin sheet by using hot-swaging and cold-drawing or rolling techniques that are customarily used for advanced materials.
(2) Nitinol alloys having M s temperatures considerably below room temperature canalso be drawn into wire; however, special die cooling procedures must be followed.
(3) The shape recovery ability of a 55-Nitinol alloy is closely linked to processing
history, e.g., prior cold work introduced by cold-drawing or rolling and finalannealing temperature. Therefore, processing must be carefully controlled.
(4)
(5)
Electrical resistance versus temperature curves provide a reliable means fordetermining Ms, As, and Md temperature points of the martensitic transforma-tion that occurs in 55-Nitinol alloys. In addition, they provide a means of pre-dicting the relative shape recovery performance of a particular Nitinol alloy.
A decrease in the elastic modulus is associated with a decrease in temperature
and precedes instability and transformation at the M s point. In the case of 100-mil diameter drawn rod, elastic modulus decreased from a maximum of 12 millionpsi to a minimum of about 3 million psi at the M s point. Heating from below the
M s point causes a reverse behavior whereby modulus increases, starting at atemperature (A s point) somewhat in excess of the Ms point and continuing until amaximum is reached at about the Md point.
(6) The 0.2 percent offset yield stress decreases as temperature is lowered below theMd point and reaches a minimum at M s. In the case of 100-mil diameter drawnrod, the 0.2 percent offset yield stress decreases from a maximum of 90 000 psito a minimum of about 5 000 psi. Upon heating, a reverse behavior occurs.
However, the increase is initiated at As, a temperature point somewhat in excessof M s .
57
(7) Plastic deformation when introduced at temperatures below or somewhat above theM s point causes transformation in amounts determined by the degree of strain and
the strain temperature. The highest degree of transformation for a given degreeof deformation occurs at the M s point. As temperature is increased above theMs, the amount of transformation for a given degree of deformation continues to
diminish to zero at M d. Deformation below the M s point appears to provide ahigh degree of transformation.
(8) Wire can be strained 6 to 8 percent and still exhibit 100 percent recovery. How-
ever, the material can be strained considerably beyond 8 percent and still exhibitover 90 percent recovery, which should be satisfactory for many possible appli-cations.
(9) Recovery tensile stress over 100 000 psi can be generated by 15- and 20-mildiameter wires strained initially at 8 percent.
(10) Mechanical work capability of over 2500 inch-pounds per cubic inch was consis-tently demonstrated with 15- and 20-mil diameter wires that had undergone approxi-mately 7 percent initial strain, with a maximum value of 2900 inch-pounds percubic inch being reached.
There are indications that these limits could be raised by using improved processingtechniques.
From these results it is recognized that 55-Nitinol has many potential applications inadvanced space structures, especially where requii'ements for expandable and erectable struc-tures or self-actuating devices in space areneeded.
58
REFERENCES
o
o
o
o
1
1
.
.
Buehler, W.J. ; and Cross, W.B. : 55-Nitinol Unique Alloy Wire for Self-Erectable Space
Structures. Presented at the Wire Association, Inc., Non-ferrous Div. Meeting, Hunts-ville, Alabama, 3 - 4 April 1968. (To be published in the Wire Journal.)
Buehler, W.J. ; and Wang, F.E. : A Summary of Research on the Nitinol Alloys and Their
Potential Application in Ocean Engineering. Ocean Engineering, vol. 1, No. 1, July 1968.
Wang, F.E. ; Buehler, W.J. ; and Pickart, S.J. : Crystal Structure and a Unique '2Ylarten-sitic" Transition of TiNi. J. Appl. Phys., vol. 36, No. 10, October 1965.
Wang, F.E., ed. : Proceedings of Symposium on TiNi and Associated Compounds, held atU.S. Naval Ordnance Laboratory, Silver Spring, Maryland, 3 - 4 April 1967. NOL TR68-16. (These proceedings have been published in the J. Appl. Phys., vol. 39, 1968.)
Dautovich, D.P. ; and Purdy, G.R. : Phase Transformations in TiNi. Canadian Metal-
lurgical Quarterly, vol. 4, No. 2, April -June 1965, pp. 129-143.
Purdy, G.R. ; and Parr, J.G. : A Study of the Titanium-Nickel System Between Ti2Ni andNiTi. Trans. AIME, vol. 221, June 1961, pp. 636-639.
Wasilewski, R.J. ; Butler, S.R. ; and Hanlon, J.E. : On the Martensitic TransformationTiNi. Metal Science Journal, vol. 1, 1967, pp. 104-110.
Wang, F.E. ; DeSavage, B.F. ; Buehler, W.J. ; and Hosler, W.R. : The Irreversible
Critical Range in the TiNi Transition, J. Appl. Phys., vol. 39, No. 5, April 1968.
BIBLIOGRAPHY
Bradley, D. : Sound Propagation in Near-Stoichiometric TiNi Alloys. J. Acoust. Soc. Am.,vol. 37, No. 4, April 1965, pp. 700-702.
Buehler, W.J. : Intermetallic Compound Based Materials for Structural Applications. Pre-sented at the Seventh Navy Science Symposium, 14-16 May 1963, ONR-16, vol. 1, 1963.
Buehler, W.J. ; and Wang, F.E. : Martensitic Transformation in the TiNi Compound. Reac-tivity of Solids. Fifth International Symposium, Munich, 1964.
Buehler, W.J. ; and Wang, F.E. : Study of Transition Element Intermetallic Compounds. Pre-sented at the Ninth Navy Science Symposium, 5-6 May 1966.
Buehler, W.J. ; and Wiley, R.C. : The Properties of TiNi and Associated Phases. Trans.Am. Soc. Metals, vol. 55, No. 2, June 1962.
Buehler, W.J. ; and Wiley, R.C. : TiNi - Ductile Intermetallic Compound. Trans. Am. Soc.Metals, vol. 55, 1962, pp. 269-276.
59
Buehler, W.J. ; Wiley, R.C. ; andWang,F.E. : Nickel-BaseAlloys. PatentNo. 3,174,851,March 23, 1965.
Buehler, W.J., et al: Effect of Low-TemperaturePhaseChangesonthe MechanicalProper-ties of Alloys Near CompositionTiNi. J. Appl. Phys., vol. 34, No. 5, May 1963,pp. 1475-1477.
Cooper,J.E. ; Bowker, D.E. ; andCross, W.B. : Investigationof the UniqueMemory Proper-ties of 55-Nitinol Alloy. (Tobepresentedat the 15thNationalSAMPESymposiumon29and30April and1May 1969at the Statler Hilton Hotel, Los Angeles,California.)
Golf, J.F. : Thermal Conductivity,ThermoelectricPower, andElectrical Resistivity ofStoichiometricTiNi in the 3° to 300°KTemperatureRange. J. Appl. Phys., vol. 35, No. 10,October1964.
Goldstein, D.M., et al: Effectsof Alloying UponCertain Properties of 55.1 Nitinol. NOLTR64-235, U.S. NavalOrdnanceLaboratory, Silver Spring, Maryland, May 1965.
Hanlon, J.E. ; Butler, S.R. ; andWasilewski, R.J. : Effect of Martensitic Transformationonthe Electrical andMagneticProperties of NiTi. Transactionsof theMetallurgical SocietyofAIME, vol. 239, September1967,pp. 1323-1326.
Rozner, A.G. ; andBuehler, W.J. : Effect of ColdWork onRoom-TemperatureTensile Prop-erties of TiNi Intermetallic Compound.Trans. Am. Soc. Metals, vol. 59, 1966.
Rozner, A.G. ; andBuehler, W.J. : Low TemperatureDeformationof the TiNi IntermetallicCompound.NOLTechnicalReport66-38, March1966.
Rozner, A.G. ; andWasilewski, R.J. : Tensile Properties of NiAl andNiTi. J. Inst. Metals,vol. 94, 1966,pp. 169-175.
Rozner, A.G., et al: Effectof Additionof Oxygen,Nitrogen, andHydrogenonMicrostructureandHardnessof CastTiNi Intermetallic Compound.Trans. Am. Soc. Metals, vol. 58, No. 3,September1965.
Schuerch,H.J. : Certain PhysicalProperties andApplicationsof Nitinol. NASACR-1232,November1968.
Spinner,S. ; andRozner, A.G. : Elastic Properties of NiTi As A Functionof Temperature.J. Acoust. Soc. Am., vol. 40, No. 5, November1966,pp. 1009-1015.
Wang,F.E. : EquiatomicBinary Compoundsof Zr with Transition ElementsRu, Rh, andPd.J. Appl. Phys., vol. 38, No. 2, February 1967.
Wang,F.E. : TheMechanicalProperties as a Functionof TemperatureandFree ElectronConcentrationin StoichiometricTiNi, TiCo, andTiFe Alloys. Proc. First InternationalConf.on Fracture, vol. 2, 1965,pp. 899-908.
Wang,F.E., et al: Growthof TiNi SingleCrystals by a Modified "Strain-Anneal" Technique.J. Appl. Phys., vol. 35, No. 12, December1964,p. 3620.
60 NASA-Langley,,969-- ,_ CR-1433