Skip to main content

Full text of "DTIC AD0840063: ROOM TEMPERATURE CREEP IN TI-6AL-4V"

See other formats


UNCLASSIFIED 



AD NUMBER 


AD840063 


NEW LIMITATION CHANGE 
TO 

Approved for public release, distribution 
unlimited 


FROM 

Distribution: Further dissemination only 
as directed by Air Force Materials Lab., 
Attn: MAM, Wright-Patterson AFB, OH 45433, 
Jun 1968, or higher DoD authority. 


AUTHORITY 


AFML ltr, 7 Dec 1972 


THIS PAGE IS UNCLASSIFIED 













AD840063 


AFML-TR-68-171 


ROOM TEMPERATURE CREEP 
IN TI-6A1-4V 


WALTER H RE1MANN 
Air Force Materials Laboratory 


TECHNICAL REPORT AFML-TR-68-171 


JUNE 1968 


n n C 




ill- 


SlP } V 12o3 


; \£VJ 


This document is subject to special export controls and each transmittal 
to foreign governments or foreign nationals may be made only with prior 
approval of the Metals and Ceramics Division (MAM), Air Force Ma¬ 
terials Laboratory, Wright-Patterson Air Force Base, Ohio 45433. 


AIR FORCE MATERIALS LABORATORY 
AIR FORCE SYSTEMS COMMAND 
WRIGHT-PATTERSON AIR FORCE BASE, OHIO 




room temperature creep 
IN Ti-6A1-4V 


WALTER H. REIMANN 
Air Force Materials Laboratory 


This document is subject to special export controls and each transmittal 
to foreign governments or foreign nationals may be made only with prior 
approval of the Metals and Ceramics Division (MAM), Air Force Ma¬ 
terials Laboratory, Wright-Patterson Air Force Base, Ohio 45433. 





FOREWORD 


This report was prepared by the Strength and Dynamics Branch, 
Metals ana ceramics Division, under Project Number 7351, "Metallic 
Materials", Task Number 735106, "Behavior of Metals", Subtask 
735106-043, "Fatigue and Reliability of Aerospace Materials". The 
research vork was conducted in the AF Materials Laboratory, Wright- 
Patterson Air Force Base, Ohio, by Dr. W. H Reimann of AFML. 

This report covers work performed from May 1967 to February 

1968. 


This manuscript was released by the author June 1968 for 
publication as an Technical Report. 

This technical report has been reviewed and is approved. 


W. J 



Chief, Strength and Dynamics Branch 
Metals and Ceramics Division 
Air Force Materials Laboratory 






ABSTRACT 


Recent investigations have emphasized that Ti-6A1-4V can 
show appreciable amounts of creep at room temperature. The 
present study was conducted in an attempt to establish the 
design limitations imposed by this behavior. It was concluded 
that as long as the applied stresses remain below the limit of 
proportionality the alloy exhibits no instability. However 
after plastic deformation, and under reversed loading, the alloy 
shows drastic dimensional instability at very low stresses. It 
is essential therefore that after forming, a component be given 
adequate stress relief prior to being put into service. 


This abstract is subject to special export controls and each 
transmittal to foreign governments or foreign nationals may 
be made only with prior approval of the Metals and Ceramics 
Division (MAM), Air Force Materials Laboratory, Wright- 
Patterson Air Force Base, Ohio 45433. 


in 




LIST OF ILLUSTRATIONS 

Figure Page 

1 Tensile specimen configuration. 9 

2 Torsion specimen configuration. 10 

3 Load-time profile used in slow cycle tensile 

tests. 11 

4 Overall view of torsional straining teBt rig. 12 

5 Close up view of specimen with mirror collars 

in place. 13 

6 Incremental loading with 30 minute hold time 

(skill annealed condition). 14 

7 Incremental loading with 30 minute hold time 

(STA condition). IS 

8 Effect of 100 0-c v load cycles at a frequency 

of 0.5 Us. (mill annealed condition) 16 

9 Effect of 100 0-o m load cycles at a frequency 

of 0.5 Us. (STA condition) 17 

10 Effect of slow cycling with hold time on the 

two heat treated conditions. 18 

11 Torsional stress-strain diagram for solid mill 

annealed specimen. 19 

12 Torsional stress-strain diagram for solid STA 

specimen. 20 

13 Torsional stress-strain diagram for tubular 

mill annealed specimen. 21 

14 Typical torsional creep curves for the two 

heat treated conditions. 22 

15 Effect of reversed torsional loading on tubular 

mill annealed specimen. 23 

16 Effect of intermediate anneal plus reversed 

torsional loading on tubular mill annealed 
specimen (same specimen as in Figure 15 . 24 

i v 











i 


INTRODUCTION 


Recent investigations by Wood (Refs 1, 2) have shown that 
Ti-6A1-4V can exhibit appreciable amounts of creep at room 
temperature. Furthermore this creep can take place at stresses 
below the nominal yield point of the material. 

The phenomenon of room temperature creep in titanium and 
its alloys is not new. Adenstedt (Ref 3} in 1949 was the first 
to show this for pure titanium and since then a number of other 
workers have obtained similar results for titanium alloys (Refs 
4, 5, 6, 7). Of the common titanium alloys Ti-6Al-4V appears 
to show the least amount of room temperature creep. 

The significant feature of Woods study has been to point 
out that the creep observed is very dependent on the loading 
method. The earlier investigations were restricted to uniaxial 
loading; however Wood has clearly shown that under torsional 
loading, the room temperature creep observed is far greater 
and takes place at stresses much lower than those required to 
initiate creep under simple tensile loading. 

In view of the existing and planned wide-spread use of 
titanium alloys for aerospace applications where stability 
over long time periods is of vital concern, it is necessary 
that this creep phenomenon be fully understood and design 
limitations established. It was with this in mind that the 
AF Materials Lab undertook the following investigation. 


II. PROCEDURES 


All specimens tested were of T1-6A1-4V. The specimens 
were tested in two standard heat-treated conditions - mill 
annealed and STA. 

Two series of tests were performed, one series using 
uniaxial loading and the other using torsional loading. 

Figs. 1 and 2 show the specimen configurations used in each 
case. For the torsional tests some additional tubular speci¬ 
mens were made by drilling a 3/16" dia. hole along the axis 
of the specimen. After machining the specimens were heat 
treated prior to testing. 


1 




For the uniaxial testing a standard MTS testing machine 
was used. initially an matron ollp-oit caL cnaumclei was used 
to measure strains; however this was not successful because 
the long testing times involved in seme of the tests permitted 
too much "drift" in the zero point of the strain measuring 
circuit. For these tests extensions were measured by a 
Tuckerman optical strain gage. This allowed extensions to be 
determined with an accuracy of better than 5 microinches, but 
had the disadvantage that the results could not be automat¬ 
ically recorded. 

It was felt that the probability of observing creep would 
be greater for very low strain rates, therefore slow loading 
rates were used in most of the tests. The actual strain rates 
during the loading part of each test are shown on the re¬ 
spective figures. 

In addition to simple load/hcld tests, same specimens were 
subjected to cyclic load/unload ai i load/hold/unload testing. 
For these last tests the load-time profile shewn in Fig. 3 was 
enscribed on the data track of the MTS testing machine. It was 
considered that this profile would approximate many expected 
aerospace service conditions. 

For the torsion tests a special testing rig was designed 
and built which incorporated an optical torsional strain 
measuring systems. An overall view of the test set-up is 
shown in Fig. 4, The specimen forms a link between two shafts, 
one of which is rigidly restrained while the other is free to 
rotate. The specimen grips contain wedge blocks which bear 
against machined flats on the specimen and thus ensure that 
there is no slippage between specimen and grip. Rigidly 
attached to the end of the rotating shaft is a 8" dia. pulley 
with cable and pan so that the specimen may be dead loaded 
and the applied torque will remain constant with deflection. 

To measure the resulting torsional strains, two mirros 
are fixed to two collars which clamp directly onto the test 
section by means of pointed positioning screws. By using a 
spacer attachment these collars can be clamped on to give an 
accurate 1" gage length. A close up view of the specimen 
with the collars in place is shown in Fig. 5. Deflections 
are observed by means of a telescope with cross hair eyepiece 
which focuses on the. images of two vertical scales in the two 
mirrors. Deflections as small as 0°02' can be readily observed 
by this method. For the strain calculations deflections were 
determined as the deflection of the movable end minus the 
deflection of the "fixed” end of the gage length. In this way 
elastic deflections in the test rig itself can be eliminated. 
These elastic deflections turned out to be quite appreciable. 
For small angles of twist this proved to be a very accurate 


2 





method. For larger angles of twist (>10°) the accuracy de¬ 
creased somewhat because the mirrors had to be periodically 
rezeroed. The error introduced by the fact that the two 
mirrors do not lie along the axis of the specimen may be 
ignored as long as the mirror-scale distance is kept large. 

The method for determining the torsional stress-strain 
data was as follows. After the specimen was set in the 
testing rig the zero point for each of the mirrors was read. 

A dead load was then applied and the deflections of each of 
the mirrors again noted. If after 1-2 hours no further 
deflection was observed the specimen was unloaded and the 
zero rechecked. The load was then increased by one step and 
the process repeated. If the specimen started to creep under 
the applied torque, the load was maintained constant until 
the specimen had again come to rest. This proved to be very 
time consuming since 25-50 hours were frequently required. 
At the highest stress levels used the specimen did not come 
to rest but continued to creep although at a decreasing rate. 
In these cases the load was maintained for a period of 150 - 
300 hours and then removed. 

Torsional stresses and strains were calculated from the 
following standard equations! 


Shear strain y * —- 

L 

Where r * specimen test section radius 
6 * angle of twist (radians) 

L ■ gage length 


Maximum shear stress 


T max 


T max 

Where M T 



16 M T (solid specimen) 

TIT 

16 M t (tubular specimen) 



applied torque (in-lb) 
outside diameter 
inside diameter of tube 


3 





Ill 


RESULTS 


A. Uniaxial Loading 

Prior to any other testing, one specimen from each heat 
treated condition was pulled in Bimple tension to determino 
the 0.2% yield strength and the limit of proportionality. 

The values obtained are shown belowt 


0.2% Y.S. L.P. 


Mill annealed 133.4 ksi lid ksi 

STA 152.0 ksi 127.5 ksi 


Figs. 6 and 7 show the effect of simple incremental loading 
with a 30 minute hold time at each level. It can be seen that 
under these conditions no creep is observed below the limit of 
proportionality of the alloy. 

The result of cyclically loading the specimen can be seen 
in Figs. 8 and 9. Each specimen was subjected to 100 cycles 
zero to maximum stress at a frequency of 0.5 Hz. Again it is 
clear that no strain accumulation takes place below the limit 
of proportionality. 

Fig. 10 summarizes the results obtained from using the 
load-time profile shown in Fig. 3. Again it can be seen that 
even at 90% of the yield stress the strain accumulation with 
cycling is very small. 

Although not shown in Fig. 10, one more test for each heat 
treated condition was run, in which the load was cycled from 
■*■80% to -40% of the yield stress. Again, no creep was observed. 

All of the above results confirm the early published data 
which indicate that room temperature creep in titanium alloys 
does not take place below about 85% of the yield stress (e.g. 

Ref 7). Therefore it was concluded that under uniaxial loading, 
room temperature instability of titanium alloys does not present 
a design problem! nor is there any reason to doubt the validity 
of the published data. 

B. Torsional Loading 

The stress-strain relationships obtained from simple dead¬ 
weight torsional loading are shown in Figs. 11 and 12. 

Zt was felt that the beBt reference point to study insta¬ 
bility effects would again be the limit of proportionality, 


4 





wnicn was taken as the first observable departure from the 
elastic line. The values of the torsional limit of pro¬ 
portionality for the two heat treated conditions were 
determined to be 62.5 ksi for the mill annealed conditions 
and 72.0 ksi for the STA condition. 

Fig. 13 is the equivalent stress strain relation for a 
mill annealed thin walled tubular specimen and it can be 
seen that the behavior is essentially identical. The value 
of 61.0 ksi for the limit of proportionality is in good 
agreement with that obtained from the solid specimen. 

Fig. 14 illustrated some typical torsional creep curves. 

As was the case in the uniaxial tests, it can be seen that 
under simple torsional loading, no instability is observed 
below the limit of proportionality. 

Fig. 15 shows the effect on a mill annealed tubular speci¬ 
men of interrupted and reversed torsional loading. The speci¬ 
men was loaded to the point where continuous torsional creep 
occurred and allowed to creep extensively (to a plastic shear 
strain of approximately 0.03). After this creep the specimen 
was unloaded and allowed to rest for 50 hours under zero load. 
During this rest period a very small amount (.0009 shear strain) 
of reverse plastic creep was observed. The specimen was then 
reloaded in the same direction. No creep was observed until 
the previously applied maximum load was reached, when the speci¬ 
men again exhibited instability. At this stag.* the specimen 
was unloaded again and loaded in the reverse direction. As can 
be seen from Fig. 15 this had a very drastic effect on the creep. 
The specimen started creeping as soon as the load was reversed 
and there was no clearly defined elastic region. This exagger¬ 
ated Bauschinger effect has already been emphasized by Wood 
(Ref 2) and the present results merely confirm it. Upon re¬ 
versing the load again back to the original direction the same 
effect took place with the specimen exhibiting instability as 
soon as the load went through zero. After this symmetrical 
hysteresis loop was established, the specimen was removed from 
the testing machine and reheat treated to the mill anneal con¬ 
dition before reloading. The effect of this heat treatment on 
the loading can be seen in Fig. 16. It is clear that the effects 
of the prior plastic deformation have been completely eliminated, 
and the specimen follows the initial loading curve. 

The specimen was taken just into the creep region (a shear 
strain of approximately 0.003 as against 0.03 initially) and 
then the load was reversed again. It is clear from Fig. 16 that 
again the specimen exhibits instability at stresses far below 
the original limit of proportionality. However a comparison of 
Figs. 15 and 16 show that the amount of creep at any intermediate 
load Is much smaller in the second case. This would appear to 
indicate that the instability observed upon load reversal is a 
function of the prior plastic strain. 

5 






IV. 


DIJCUSSION 


The above results re-emphasize the well known fact that 
titanium alloys can exhibit appreciable amounts of creep at 
room temperature. However the point of concern at present is 
the influence of this phenomenon on the applications for which 
these alloys are being considered. 

For the present applications at least there does not 
appear to be any real instability problem. It is clear that 
as long as the stresses to which the component will be sub¬ 
jected remain within the elastic range, whether those stresses 
are steady or cyclic, the alloy will not exhibit creep. The 
design loads used in the various applications all lie well 
within this region. 

The marked Bauschinger effect shown by this alloy is a 
very significant feature that must be taken into consideration. 
This Bauschinger effect occurs where the load is reversed after 
prior plastic deformation. As already stated the design loads 
are well within the elastic range and therefore would not pro¬ 
duce this effect. However two cases come to mind where it 
could be significant. 

The first of these is the common practice of creep-forming 
large titanium sheet components. It is imperative that ofter 
creep forming, components that will be suS^ected to reversed 
loadings be given adequate stress-relief treatments to ensure 
stability (see Ref 2). 

The second case to be considered is the possibility of 
plastic overloads resulting from gust loading, severe maneuvers 
or clear air turbulence. This case is not as simple as the 
first, nor is the solution as readily available. More investi¬ 
gation is required before quantitative answers can be supplied 
but it is probable that this also is not as severe a problem as 
it might appear at first glance. It must be remembered that 
the results shown in this investigation and the observations of 
Wood (Ref 2) were obtained by dead weight loading where the 
specimen was given every opportunity to deform by maintaining 
the load for a period of many hours. The overloads resulting 
from gust etc, take place at very high loading rates. The 
plastic deformation experienced would be extremely small (if 
any) and therefore the induced instability upon any load re¬ 
versal would also be correspondingly small. 

It would be of interest to determine the extent of the 
Bauschinger effect under reversed uniaxial loading. Unfortun¬ 
ately th_ present test facility did not permit this to be done 
as any attempt to apply plastic compressive loads resulted in 
severe buckling problems so that the measurements were rather 
meaningless. However there is no reason to suppose that the 
same effects would not take place under simple uniaxial loading. 


6 



V 


REFERENCES 


1. W. A. Wood, Institute for the Study of Fatigue and 
Reliability, Columbia University, Tech. Report No. 

45, April 1967. 

2. W. A. Wood, Institute for the Study of Fatigue and 
Reliability, Columbia University, Tech. Report No. 

54, January 1968. 

3. H. K. Adenstedt; Metal Progress, Vol. 65, p. 658, 1949. 

4. W. B. Aufderhaar; Mallory-Shanon Titanium Corp., 

Research Report, March 21, 1958. 

5. J. A. VanEcho; Battelle Memorial Institute, DMIC Tech. 
Note, June 8, 1964. 

6. A. J. Hatch, J. M. Partridge and R. E. Broadwell; J. of 
Materials Vol. 2, p. Ill, March 1967. 

T. F. Kiefer and F. R. S^hwartzberg; Martin-Marietta 
Corp., Denver Division, Final Report Contract NAS9-5842, 
June 1967. 


7 




ACKNOWLEDGEMENTS 


Th» author wishes to thank Mr. F. J. Peck Jr. of th 
Federal Aviation Administration, for his interest and co 
operation, and Mr. D. R. Donaldson, of the SST Division, 
Boeing Aircraft Company, Seattle, for the supply of 
material and specimens for this investigation. 


8 















Figure 2. Torsion specimen configuration 










Figure 4. Overall view of torsional straining teat rig. 










Figure 5. Close up view of specimen with mirror collars 
in place. 


13 









0 ~ KSI 



Figure 6. Incremental leading with 30 minute hold time 
(mill annealed condition). 


14 






w 


0.6 


0.8 


1.0 


0 


Figure 7 


Q2 


L.. 

% 6 


Incremental loading with 30 minute hold time 
(STA condition). 


15 









Figure 8. Effect of 100 0“« max load cycles at a frequency 
of 0.5 Hz. (mill annealed condition) 






8 § §> 2 


ISM _i? 


Figure 9. Effect of 100 0-o max load cycles at a frequency 
of 0.5 llz. (STA condition) 


17 


3 % 




STRAIN % 



Figure 10. £ffect of slow cycling with hold time on the 
two heat treated conditions. 


18 




ShtAR STRAIN 





001 










o 


2 


O 

flO 


ISM 


O 

« 




NW1S UV3HS »*a 



Figure 13. Torsional stress-strain diagram for tubular 
mill annealed specimen. 


21 










\J£\J 



Pigure 14. Typical torsional creep curves for the two 
heat treated conditions. 


22 































UNCLASSIFIED 


DOCUMENT CONTROL DATA - R & D 

(Security ela$ title allot) of till*, body of abatract and indating annotation nun bo on lorod whom tho owmrmtl rwport la ctmaaitiod) 


* ARIAiNAriNA 1CTIVITV /f' n mnrmlm 

Strength and Dynamics Branch 
Metals and Ceramics Division 
Air Force Materials Laboratory, WPAFB, Ohio 


». REPORT TITLE 


RrRORT «*ril«ITV Cl aiiifitation 

UNCLASSIFIED 


a*, croup 


ROOM TEMPERATURE CREEP IN Ti-6A1-4V 


*■ Ollcnmm NOTH (T,p» ol rmpot! m,d incluoiom dolma) 


• ■ au thopmS) (Fir* 1 nomo, middlo initial, Imat namo) 


Dr. Walter H. Reimann 


• REPORT DATE 

June 1968 


M. CONTRACT OR CHANT NO. 


7a. TOTAL NO. OP PACE! 7b. MO- OF RIFI 

27 7 


to. ORIGINATOR** REPORT NUMBKfMH 


:t mo. 7351 


AFML-TR-68- 171 


«. TASK NO. 735106 


OTHER REPORT NO(B> (Any thaw nwm h oto that a 
tfifi taparl) 


io oktriiution This i s subject to special export controls and 

each transmittal to foreign governments or foreign nationals may be made 
only with prior approval of the Metals and Ceramics Division (MAM), Air 

n . >• i 1 _ a _ * _ t j ■ .1 — u x n_ a x a i n n . _ 4 r J O n 


Force Materials Laboratory. Wriqht-Patterson Air ForceJlase. Ohio 45433 


I 


■ SUPPLEMENTARY NOTES 

IS. BRONSORINB MILITARY ACTIVITY 


AFML (MAMD) 

Air Fbrce Systems Command 


Wright-Patterson AFB, Ohio 45433 


ABSTRACT 


Recent investigations have emphasised that Ti-6A1-4V can shorn appreciable 
amounts of creep at room temperature. The present study was conducted in an 
attempt to establish the design limitations imposed by this behavior. It was 
concluded that as long as the applied stresses remain below the limit of pro¬ 
portionality the alloy exhibits no instability. However after plastic 
deformation, and under reversed loading, the alloy shows drastic dimensional 
instability at very low stresses. It is essential therefore that after 
forming, a component be given adequate stress relief prior to being put into 
service. 

This abstract is subject to special export controls and each transmittal 
to foreign governments or foreign nationals may be made only with prior 
approval of the Air Force Materials Laboratory (MAMD), Wright-Patterson 
AFB, Ohio 


FORM 

I NOV «l 


1473 


UNCLASSIFIED 

Security Classification