(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Activation energies for high temperature creep of magnesium-osdmium polycrystalline alloys."

ACTIVATION ENERGIES FOR HIGH TEMPERATURE 



ClJ-iJ V ( MAGMDriUM-CADMIUM 



F LYCRY STALLENE ALLOYS 






ARNOLD H. MHWKJK.V 



'V 1 m ' Eh ■<■•''' 
'■■ SB ■'':'■•■' : '"' 

Wm ''■■"■:"'■■ : v ;' 

W$$KfiiS»aB^ ;■. 

■■ ,'J. BMMh .-,',.,■' , 















School 



ACTIVATION ENERGIES FOR HIGH TEMPERATURE 
CREEP OF MAGNESIUM-CADMIUM POLTCRYSTALLINE ALLOTS 



******** 



Arnold H. Madbury 



ACTIVATION ENERGIES FOR HIGH TEMPERATURE 
CREEP OP MAGNESIUM-CADMIUM POLYCRXSTALLINE ALLOYS 



by 
Arnold Holmes Medbury 
Commander, United States Navy 



Submitted in partial fulfillment 

of the requirements 

for the degree of 

MASTER OF SCIENCE 

IN 

MECHANICAL ENGINEERING 



United States Naval Postgraduate Sohool 
Monterey, California 



19 55 



•n,-, 



osis 



/??13 






This work is accepted as fulfilling 
the thesis requirements for the degree of 
MASTER OF SCIENCE 
IN 

MECHANICAL ENGINEERING 



from the 
United States Naval Postgraduate School 







HIEFACE 

Current practice in the engineering world is to demand more and more 
from the materials of construction in order to gain new increments of effi- 
ciency,- speed, or fuel rate. Specifically, the engineer wants a lighter 
and stronger material which will withstand the higher temperatures that 
are required. 

The onus is placed upon the research engineer and metallurgist to 
provide the advances in metals which will allow the developments proposed 
for the end use. 

In order to provide the answers for these converging requirements a 
comparatively new field of high temperature metallurgy has emerged. It 
has been imperative to develop limited basic knowledge, theories, and ex- 
perimental results into usable data in an expedient fashion. A part of 
this larger study is the investigation of the mechanism of creep phenom- 
ena. 

The purpose of this thesis is to contribute to the data and funda- 
mental understanding of the mechanism of creep of a polycrystalline solid 
solution alloy. The data here developed supplements the already extensive 
study made by the Institute of Engineering Research at the University of 
California in Berkeley sponsored by the Cffice of Naval Research. 

The writer wishes to express his appreciation to Professor Alfred 
Goldberg for his interest and capable guidance which helped considerably 
in the accomplishment of this work. 



ii 



Item 



TAB IE OF CONTENTS 



Title 



Page 



Chapter I 
Chapter II 
Chapter III 
Chapter IV 

Chapter V 
Chapter VI 
Bibliography 
Appendix I 
Appendix II 



Introduction ••• , 1 

Materials and Preliminary Work 4 

Equipment . 8 

Experimental Technique 14 

A. Preliminary 14. 

B. Test Procedure 16 

C. Sample Calculation 17 

Results 19 

Discussion of Results 20 

25 

Material Analysis 26 

Typical Strain vs Time Curves for each^ H Deter- 
mination (Figs. 6-16) 27 



iii 



LIST CF imJSTRATICNS 

Figure Page 

1 The Creep Test Specimen 7 

2 The Experimental Set Up 9 

3 The Extensometer 11 

4 The Thermocouple-Specimen Arrangement 13 

5 Activation Energy vs Composition for the Cd-Mg System 22 
5 A Strain Rate vs Composition for the Cd-Mg System ... 24 

6 Strain vs Time for 99.85 Percent Pure Magnesium (as 
extruded) 28 

7 Strain vs Time for 99.85 Percent Pure Magnesium 
(Annealed 300°C) 28 

8 Strain vs Time for 99.85 Percent Pure Magnesium 
(Annealed 500°C) 29 

9 Strain vs Time for 88.4 Atomic Percent Magnesium and 

11 • 6 Atomic Percent Cadmium (Annealed 500°C) • • • . 29 

10 Strain vs Time for 73.2 Atomic Percent Magnesium and 

11.6 Atomic Percent Cadmium (Annealed 475°C) . . . . 30 

11 Strain vs Time for 73.2 Atomic Percent Magnesium and 

26.8 Atomic Percent Cadmium (Annealed 450°C) . • • • 30 

12 Strain vs Time for 52.2 Atomic Percent Magnesium and 

47.8 Atomic Percent Cadmium (Annealed 350°C) . . . . 31 

13 Strain vs Time for 52.2 Atomic Percent Magnesium and 

47,8 Atomic Percent Cadmium (Annealed 380°C) . . . . 31 

L4 Strain vs Time for 74.8 Atomic Percent Cadmium and 

25.2 Atomic Percent Magnesium (Annealed 475°C) . • • 32 

15 Strain vs Time for 91.4 Atomic Percent Cadmium and 

8.6 Atomic Percent Magnesium (Annealed 2l6°C) • • • 32 



iv 



(List of Illustrations continued) 



16 




Table 


I 


Table 


II 


Table 


III 


Table 


IV 



Strain vs Time for 99.3 Percent Pure Cadmium 
(Annealed 200°C) 

Nominal Atomic Percent of Alloys . 

Summary of Annealing Temperatures and Grain Counts . , 

Experimentally Determined Activation Energies for the 
Cd-Mg Binary System 

Chemical Analysis of Metals in Percent by Weight . . . 



Page 

33 
K 
5 

19 
26 



TABIE OF SIMBOIS 
(Listed in the order of their use in the text) 

£. - Creep rate (sec"* 1 ) 

Q - Structure parameter (sec" 1 ) 
^H- Activation energyf calories per molej 
R - Gas constant (cal. mol." °K ) 
~J~ - Absolute temperature ( K) 

a - Constant (Square inches per pound) 

(J"~ - Stress (psi) 

A - Constant (sec ) 



vi 



CHAPTER I 
INTRODUCTION 

Currently there is considerable interest in the behavior of metals 
lander stress at high temperatures. The aeronautical field is interested 
in magnesium and its alloys. The nuclear energy field is interested 
in cadmium and its alloys. ^'Essentially all research and development 
engineers have out of necessity become involved in the field of high 
temperature creep. 

The experimental data developed to date are not sufficient to formu- 
late rational lavs governing high temperature behavior of stressed metals. 

(3) 

'It is believed that if more were known of the mechanism of creep, which 

may be defined as the time dependent deformation of solids under stress, 
such knowledge and understanding could ultimately remove the limitations 
of empiricism and allow formulation of useful equations from fundamental 
understanding. Such equations based on short time test results could be 

used to predict long time behavior of stressed metals at high temperatures. 

(3 l) 
To this end numerous mathematical expressions have been presented. ' ' 

These are of either an empirical nature or have as a basis Erying's Re- 
action Rate Equation. 

Dora and co-workers in investigating the plastic properties of high 
purity aluminum and its alloys met with considerable success in formulat- 
ing several creep laws which are related to Erying's Reaction Rate Equa- 
tion which necessarily implies that creep is an aotivation process. They 

(7) 
develop the basic equation. 

€ = Se RT e 9 a) 



where 

£ r creep rate (see" ) 

O = a parameter which is a function of the creep stress 
or strain (structure parameter) (sec." ) 
AH = activation energy (cal.Mol. ) 
R s the gas constant (cal»Mol.~ K ) 
7" * the absolute temperature (°K) 
§ ■ constant (square inches per pound) 

tf~ ■ stress (psi) 

(3) 

and showed that for constant load tests the equation reduces to 

-AH 
C = A&*T (2) 

where 

A = constant (sec. ) 
£*H a constant independent of temperature over wide ranges of 

temperature (above that where rapid recovery occurs which 
is about 0.45 the melting temperature). It is also independent 
of creep stress, creep strain, grain size, sub-structures devel« 
oped during creep, small alloying additions, and cold work. 
Thus, the activation energy A H for creep may be determined by eval- 
uating two creep tests at different temperatures T-, and T~ using the 
relation 

AH = -; — (3) 

r « " T, 

Activation energies have been determined for several pure metals in- 



eluding cadmium ' and magnesium and reported to be respectively 
22,000 and 31,000 calories per mole. Essentially no work has been re- 
ported showing the effect of solid solution alloying on this constant. *' 
This thesis involves the experimental re-determination of A H for pure 
cadmium and magnesium and for five alloys across the oadmium magnesium 
binary system. 

The results show that there is no simple systematic variation of 
AH as a function of composition. There are significant positive and 
negative deviations from linearity. 



CHAPTER II 
INTERIMS AND PRELIMINARY WORK 

Specimens vera machined from extruded bars 0.75 inches x 0.10 inches 
obtained from Dow Chemical Company. The nominal atomic peroents for the 
alloys are given in Table I. The chemical analyses are listed in Table 
IV of Appendix I. 

Table I 
ft9ffllll)»l AVflBl? P«reants For Matels 
AUov. 

70 
71 
72 
73 
74 
75 

Preliminary to the creep experimental work it was necessary to study 
the effeot of the proposed high temperatures on the structure and surface 
of the metals. By means of the Smally X-rays technique it was established 
that the materials were as extruded (not annealed) and that a preferred 
orientation existed. A microscopic study of the specimens revealed that 
the grain count on the Surface normal to the direction of extrusion in- 
creased across the binary system from 80 grains per inch for cadmium to 
700 grains per inch for magnesium. Although it was later proved that the 

4 



Percent Cd. 


Percent >fe 





100 


100 





91.4 


8.6 


74.8 


25.2 


47.8 


52.2 


26.8 


73.2 


11.6 


88.4 



activation energy was independent of structure (for alloy 69) it vaa con- 
sidered necessary to minimize the effect of grain size by re crystallizing 
the metals to obtain the sains nominal grain count. Inasmuch as cadmium 
re crystallize s at room temperature and also had the largest grain size, 
this metal was selected as the standard. However, it seemed advisable to 
anneal cadmium to a temperature above that of the creep test so that the 
structure would not be altered prior to loading. With these two thoughts 
in mind it was experimentally determined that the alloys should be annealed 
at temperatures ranging from 200 to 500°G to give a nominal grain size of 
100 grains psr inch. Temperatures used and the grain size obtained are 
listed in Table II. 

XfibJ&JI 
Si™^.™ of An^Aniw Temperatures and Grain Counts 
Alloy. Annealing Temperature AT \Pfffl\!rff ^."ft Number 

(°C) (Hours) (Grains/inch) 

69(a) As extruded 700 

69(b) 300 1 450 

69(c) 500 1 130 

70 200» 1 80 

71 216 1 103 

72 220 1 100 

73 362 1 75 
U 450 1 60 
75 500 1 110 

•All specimens were furnace cooled ex- 
cept alloy #70 which was air cooled. 



Difficulty was encountered with alloys 69(c) and 75 due to oxidation of 
the surface at the selected annealing temperature of 500°C. A coating of 
silicone oil was found to prevent oxidation of alloy #75. However, it was 
necessary to anneal pure BAgnesium (alloy 69(c) in an inert atmosphere. 
To obtain this, helium was passed continuously through copper chips heated 
at 600°C, then through anhydrous magnesium perchlorate chips, and into a 
tube furnace where the specimens were annealed. The heated copper chips 
removed traces of oxygen and the perchlorate chips removed any moisture 
present. 

The dimensions of the specimens are shown in Figure 1. "Tolerencea 
of "t 0.001" along the reduced section were obtained in machining j how- 
ever, after annealing a total of 0.015" was tolerated in the extreme case. 



•Similar specimens were employed in the creep 
University of California.^ 1 ) 



studies performed at the 



— it" 



.250 



A" 

16 



DIA 




Pig. 1 

7 



CHAPTER III 
EQUIPMENT 

The test machine used for the creep experiments was a conventional 
single lever, constant load type manufactured by Baldwin-Lima-Hamilton 
Corporation. This machine was originally designed for testing heavier 
metals (6 ton capacity) and thus had to be modified to Increase sensitiv- 
ity for the lighter metals and to accomodate the extensometer. Figure 2 
shows the experimental set-up. The alterations on the testing unit con- 
sist of the following i 

(1) Addition of an adjustable counterweight to the lever arm in 
order to "zero" (balance the lever with the specimen installed and 
no applied stress) the machine with no load on the pan. 

(2) Shortening the lever arm ratio to increase sensitivity. 

(3) Addition of a gage to determine the angle of the lever during 
the test. 

U) Modification of the lower pulling tab connection to accomodate 
the reset tumbuokle. 
The significant features of the machine are as follows: 

(1) Loads are transmitted by means of 3 oase hardened knife edges 
in the lever arm system. 

(2) The lever arm assembly has a dowel pin at right angles to the 
direction of the knife edges thus giving the system three degrees of 
freedom. 

(3) A balanced lever of this type If symmetrical compensates Itself 



8 



COUNTERWEIGHT — 



EXTENSOMETER 
DIAL 



LOAD ROD 



FURNACE- 



THERMOCOUPLE 
SELECTOR SWITCH 

POTENTIOMETER- 



TURNBUCKLE 



COLD JUNCTION- 



CONTROLLER 




FIG. 2 THE EXPERIMENTAL SET UP 



so that the lever ratio is the same regardless of the angle. Due to 
unsymmetrical cross section of the lever arm, the counterbalance which 
was in line with the knife edges could only be truly balanced in one 
position; however, the error is insignificant for small angles. 
U) The reset turnbuokle allows an initial adjustment of the lever 
before loading and reset after a specimen was elongated appreciably. 
(5) The specimen is always uniaxially loaded which is an important 
prerequisite for accurate creep measurement. 

The extensometer shown in Figures 3 and U was specifically designed 
for use with this creep machine and is suitable for the temperatures em- 
ployed. The elements of the gage are shown in the figures while its 
features consist of the following i 

(1) 2" gage length. This is accurate to +0. 002% -0.000" by means 
of a detachable spacer. 

(2) The gage blocks contain hardened steel points which penetrate 
the specimen 0.010" on one side and have a spring loaded clip on the 
other. This is necessary to accomodate for the reduced thickness of 
specimen as creep progresses. 

(3) The upper gage block is rigidly connected to a tube which passes 
through a guide and finally attaches to the gage dial which is free 
to move vertically. The dial foundation has a guide slot for move- 
ment on the upper pull tab. 

GO The lower gage block is rigidly attached to a rod which moves 
through the above-mentioned tube and provides a resting platform for 
the dial gage stem which is spring loaded. 

10 



DIAL GAGE 



GUIDE 




TUBE 



GAGE BLOCKS 
SPECIMEN 



FIG.3 THE EXTENSOMETER 



u 



(5) All high temperature components of the gage were made from 

stainless steel* 
An L. H. Marshall 115V niohrome -wound cylindrical type furnace 16" long, 
7" O.D. and 2" I.D. was used. Eight taps for shunting were provided for 
temperature control. However, it was not necessary to use these shunts 
because of the satisfactory temperature gradient obtained across the 
gage length of the specimen by careful insulation of the top and bottom. 
The desired temperature was obtained and kept constant by a variac with 
the aid of an ammeter* 



12 



UPPER PULL TAB 



THERMOCOUPLE 



LOWER PULL TAB 




FIG. 4 THERMOCOUPLE-SPECIMEN 

ARRANGEMENT 



13 



CHAPTER IV 

EXPERIMENTAL TECHNIQUE 

A. Preliminary: 

In high temperature creep vork of this nature there are four main 
factors to consider. They are stress , strain, temperature, and material. 

Inasmuch as these experiments are of the constant load type only the 
initial stress is of concern. In uniaxial loading the measured area is 
multiplied by the desired initial stress to determine the loading. This 
load when divided by the mechanical advantage of the lever establishes 
what weight shall be added to the loading platform. 

Strain rate varies exponentially with applied stress (Equation l) and 
must be carefully controlled. The test machine lever arm ratio was call* 
brated with a Baldwin SR-4 Type U (0-2000 lbs) standard load cell which 
has an inaccuracy of— l/U% of full range at a point. The lever arm ratio 
was found to be 14*75*1. With this instrument, it was demonstrated that 
the lever gave no detectable loading error within the range of - 7.5 de- 
grees from the horizontal. Because all the experimental work for this 
thesis was done on the same test machine and was of the constant load 
nature in which tests of identical loads were compared, errors that might 
have arisen by the machine would be of a systematic nature and automatically 
cancelling. Thus, the errors introduced in controlling stress are reduced 
to the errors of determining a representative cross sectional area of the 
specimen and measuring the applied weights. The representative sectional 
area was determined by multiplying an average of three widths by an average 



U 



of three thicknesses. Weights were measured on a balance accurate to 
2 x 10 lbs. Combining these extreme errors a stress variation of less 
than 10 psi would result. 

The extensometer was calibrated by mounting in a tension-testing ma- 
chine with a calibration specimen with SR-4 gages between the gage points 
on both sides of the specimen. The extensometer was found to be accurate 
to its sensitivity of 2 x 10 • 

A substantial part of the preliminary work on this thesis was devoted 
toward getting satisfactory temperature control in the furnace and a sat- 
isfactory gradient across the specimen. For constant stress, the creep 
rate varies exponentially to A. It is mandatory to minimize the tem- 
perature gradient across the gage length in order to get accurate data. 
Because of the nature of the -^ H equation (Equation 3) it is essential 
to determine the temperature accurately. This was done by means of three 
duplex fiberglass insulated chrome 1-alurael thermocouples attached with 
glass thread to the specimen at the center and just outside of the gage 
blocks as shown in Figure A. The thermocouple loads were led through a 
cold junction ( ice water ) to a selector switch connected to a potentio- 
meter. This thermocouple arrangement was calibrated at 0°C and gave read- 
ings of •+ 0.75°C for each thermocouple. 

The furnace temperature could be controlled automatically by putting 
the center thermocouple to the controller input; however, cyclic variations 
of 5°C were obtained. This could be improved by the proper amount of re- 
sistance placed in parallel with the control relay. However, the voltage 
variation over a 2A hour period was of the order of 10 percent which would 



15 



again broaden the temperature variation. Thus, better control was exercised 
by establishing the required constant current to maintain the desired tem- 
perature and then noting the ammeter and constantly changing the variac to 
maintain a constant ammeter reading. Utilizing this technique the gradient 
across the specimen was reduced to — 2°C and the average temperature varia- 
tion during the critical part of the run was less than 0.5 C. It is es- 
timated that the variation and uncertainty in temperature introduced an 
error of less than 3 percent in the determination of^H • 

B. Test Procedure: 

The determination of an activation energy using Equation 3 requires 
that two creep tests be run at two different temperatures under the same 
initial load. The problem then was to determine the load and temperatures 
such that sufficient experimental points could be obtained at the high 
and low temperatures within about five minutes and twenty hours respectively, 
This often required frequent preliminary runs as each new alloy was tested. 

The recrystallized specimen was mounted in the extensometer and thermo- 
couples attached (see Figure 4). The furnace was brought into position 
and insulated at the top and bottom with aluminum foil and felt. The 
furnace was then turned on and allowed to come up to temperature. The 
test was started after the desired equilibrium conditions were established. 
This took from three to five hours after the power was turned on. An auto- 
matic timer was used for the rapid test. 

The low temperature test (50 - 100°C less) was generally discontinued 
after reaching a strain of about 6 percent, the important part of the test 
being in the region of 3 to 6 percent extension where frequent readings 

16 



were taken and fine temperature control maintained. 

The data from the two runs were reduced and plotted on strain-time 
coordinates. The creep rates ( € ) were obtained by drawing tangents to 
the curve at a given strain. The exact temperatures at the time corres- 
ponding to this strain were used. Thus the four unknowns of Equation (3) 
are available for evaluation. 

In the experimental work on this thesis satisfactory results were 
obtained when two high temperature runs gave essentially the same result 
with one low temperature run. 

/\H was determined for the seven metals shown in Table III. Three 
separate AH values were determined for alloy #69 to demonstrate that 
j&H is essentially independent of structure. 

C. Sample Calculation - 

(1) To calculate ^H it is necessary to substitute a set of ex- 
perimental values into Equation (3), 

\ 



- 



1 - J T T 

~T~ ~t7" * " 2 



Consider the isothermal strain-time curves of Figure 6 where, 
T x = 574°K 

T 2 = 473°K 
By evaluating the slope at some particular strain say, 

€ = m x icr 3 



17 



the creep rates are found to be - 
£,» 177 x KfVsec 
€ x = 0.486 x lO^/Sec 

( 177 \ 
AH= (574)(473)R ln\0~^8c"y 
574 - 473 

= 30,500 cal/mole 
(2) To calculate £ standardized to 200°C at 2000 psi consider 
Equation 3 rearranged, 



in/ 6, \ = *H ttl - tg) 

Nov consider the data of Figure 8 where 



T x = 559°K 

€/ = 97.3 x lO^ec 

let the standardized temperature = 473°K and ^H s 30,000 cal/mole 
then; 



In 97.3 x IP" 6 Z (30,000) (86) = 4.39 

4 * (2) (473) (559) 



£ 2 = 97.3 x 10" 6 = 0.732 x lO^Sec" 1 
133 



18 



CHaPTER V 

EXPERIMENTAL RESULTS 

The activation energies for the metals tested are reported in Table III 

and are shown graphically in Figure 5* Typical strain vs time curves for 

each alloy used for determining A H are given in Appendix II, Figures 6-16. 

Table III 
Experimentally Determined Activation Energies 
For the qd-qg Binary System 



AH97 
No. 


Magnesium 
(Atomic %) 


Cadmium 
(Atomic %) 


Temperature Annealed 

(80) """ 


Activation 
Energy (cal/mole) 


69 


100 





As Extruded 


32,400 


69 


100 





As Extruded 


30,500 (Fig 6) 


69 


100 





300 


30,300 (Fig 7) 


69 


100 





300 


30,300 


69 


100 





500 


30,300 (Fig 8) 


69 


100 





500 


29,800 


75 


88.4 


11.6 


500 


38,500 (Fig 9) 


75 


88.4 


11.6 


500 


39,200 


74 


73.2 


26.8 


475 


26,100 (Fig 10) 


74 


73.2 


26.8 


475 


25,900 


74 


73.2 


26.8 


450 


2*6,400 (Fig 11) 


73 


52.2 


47.8 


350 


33,600 


73 


52.2 


47.8 


350 


33,900 


73 


52.2 


47.8 


380 


71,700 (Fig 12) 


73 


52.2 


47.8 


380 


57,600 (Fig 13) 


72 


25.2 


74.8 


220 


18,500 


72 


25.2 


74.8 


220 


18,400 (Fig 14) 


71 


8.6 


91.4 


a6 


23,300 (Fig 15) 


71 


8.6 


91.4 


216 


23,300 


70 





100 


200 


22,100 (Fig 16) 



19 



CHAPTER VI 
, DISCUSSION OF RESUUS 

The results show an absence of any systematic variation of AH 
across the Mg-Cd system. However, the good agreement of the results on 
both pure Cd and Mg with those reported in the literature gives 
confidence to, the other determinations in this thesis. It is significant 
that the activation energies for Cd and Mg are insensitive to minor dif- 
ferences in metal preparation, purity, and structure. The duplication of 
^H for Mg in the "as extruded", annealed to 300°C and 500°C (grain 
sizes in Table II) implies that for this pure metal AH is essentially 
independent of grain size and small amounts of strain. 

An attempt was made to obtain identical grain sizes (re crystallized) 
for all alloys. The small variations in the final grain sizes used should 
not affect the AH values. It is thus felt that the results shown in 
Figure 5 are real. In what follows is an attempt to delve into some of 

the possibilities that might account for these values. 

(12) 
Jette v 'in analyzing Vegard's law of additivity states that negative 

and positive deviations from linearity exist due to interactions which are 
both chemical and physical in nature. These deviations might well be ex- 
pected after considering the extraordinary differences in the component 
metals of this solid solution. The major differences of this system 
follow;, 

(l) significant difference (8^) in atomic radii thus resulting in 

dlstortional strain energies. 



20 



(2) high (2 volts) difference in electrochemical negativity giving 
strong interaction associated with tendency toward compound formation. 

(3) high exothermic heat of formation (4.6 K cal./g. atom) compared 
to other solid solutions (the order of ~ 1.0 K cal./g. atom) which is 
due to strong interaction and can be related to negative deviations 
of the atomic radii based on the law of additivity. 

The interaction should give a larger A H ajid thus a positive deviation 
from the straight line joining the values for the two pure metals. The 
greatest effect of interaction per atomic percent added would be felt on 
the first addition of the solute to the host metal j namely at 10/6 on 
either side of the Cd-hjg system (the lowest alloy compositions studied). 
This was true for the 10$ Cd alloy only. 

The results of the 25$ Cd, and 75$ Od are unexpected and difficult 
to explain; however, the drop is not large. On the other hand variable 
results were obtained with the 50$ Cd alloy. 

An examination of the phase diagram shows the existence of an ordered 
structure, X , below 250°C for the 50$ composition. (This also corres- 
ponds to the maximum temperature). ' 

The lowest AH value reported for this composition was obtained 
from tests run in this temperature range. Additional tests were run at 
temperatures above 250°C (where the random structure, °< , exists.) Two 
sets of tests gave surprisingly high and different AH values. The re- 
liability of the data on the one hand and the great spread of values for 
the 50$ Cd alloy on the other hand suggests that the creep mechanism is 
in some manner affected by the disorder-order reaction taking place (slowly) 



21 



30,000 

70,000 

60,000 

<\ 50,000 



40,000 



w 

25 
O 



g 30,000 __ 



o 20,000 t 



10,000 







'. 






o 














1 


i 

o 
























a 














o 








i_ o 


>- — a 


_. 


— — - 




' ~~ 




1 








o 



































Cd 10 20 30 40 50 60 70 80 90 
COMPOSITION (ATOMIC PERCENT) 

FIG. 5 ACTIVATION ENERGY vs COMPOSITION FOR THE Cd - Mg 
SYSTEM 



Mg 



22 



in this composition. It is felt that additional study of this composition 
would be of benefit in furthering our knowledge not only of creep but also 
on the order-disorder reaction and its influence on creep. 

It might be of interest to note that the 50-50 alloy showed the great- 
est creep resistance. The creep rate for all compositions stressed at 
2000 psi were standardized to 200 C. A plot of In € vs alloying is shown 
in Figure 5A. As a reference, a straight line Joins the creep rate of the 
pure metals (this supposes that the increase (or decrease) in rate is pro- 
portional to a composition change). A comparison between Figures 5 and 5A 
shows that negative deviations from £ (increased strength) and positive 
deviations from Z2i H (increased bonding interaction) are consistent, thus 
giving further evidence to the validity of the data obtained. Again, the 
abnormal behavior of the 50^6 Cd alloy is exhibited in that a significant 
deviation (increased creep resistance) as might have been expected was 
not obtained. 



23 



10 3 



10' 

x-x 

V 

s 

I 

2 10 

o 



OS 



CO 



10 



-1 



10 



-2 



■ I 




















\ 

\ 




















\ 






















\ 




















•\ 




















\ 


\ 




















\ 

N 






















\ 




















\ 




















\ 




















> 


\ 
\ 




















\ 


\ 








































\ 




• 
















\ 


V 




















\ 

\ 




















\ 


\ 




















\ 




















\ 






















\ 




















\ 






















\ 




















\ 




















u 




































































































• 























Cd 10 20 30 U0 50 60 70 
COMPOSITION (ATOMIC PERCENT) 



80 



90 Mg 



FIG 5A STRAIN RATE vs COMPOSITION FOR THE Cd - Mg SYSTEM 
Standardized to 200°C at 2000 pel 



24 



BIBLIOGRAPHY 



1. Anonymous 



2. Murray, R. L. 



3. Schvope, A.D. and 
Jacks on, L.R. 

4. Sully, A.H. 



5. Eyring, H. 



6. Kauzman, V. 



7. Sherby, O.D. and 

Dorn, J.E. 



8. Sherby, O.D., Orr,R.L., 

and Dorn, J.E. 



9. Frenkel, R.E., Sherby, 
O.D, and Dorn, J.E. 



10. Sherby, O.D. and Frenkel, 
R.E. 



11. Dorn, J.E., Goldberg, A., 
Tietz, T.E. 



12. Jette, E.R. 



MA.TERIAIS APPLICATIONS OF TOMORROW - 
M4TERIAIS AND METHODS, p 130-131 
October 1952 

INTRODUCTION TO NUCIEAR ENGINEERING 
Prentice Hall 

A SURVEY GF CREEP IN METAIS. NACA Technical 
Note 2516 

METALLIC CREEP AND CREEP RESISTANT ALLO© 
Inter science Publishers, Inc. New York, 
1949, p 103 

VISCOSITY, PIASTICITY AND DIFFUSION AS 
EXAMPIES OF ABSOLUTE REACTION RATES 
Journal of Chemical Physios 1936 
p 283-294 

FLOW OF SOLID METAIS FROM THE STANDPOINT 
OF THE CHEMICAL RATE THEORY. Trans. AIMS 
(1941) 143, p 57-83 

AN ANALYSIS OF THE PHENOMENON OF HIGH 
TEMPERATURE CREEP, 34th Technical Report, 
dtd Jan 15, 1954 

CREEP CORRELATIONS OF METAL AT EI£VATED 
TEMPS, Journal of Metals, Vol. 4, dtd 
dtd Sept 1952, p 959 

ACTIVATION ENERGIES FOR CREEP OF CADMIUM, 
INDIUM AND TIN, Minerals Research Labora- 
tory Inst of Engineering Research, Univ. 
of Calif., 36th Technical Report dtd 15 
April 1954, dtd Oct 15, 1953 

DISCUSSION OF PAPER, "CREEP BEHAVIOR OF 
EXTRUDED EI£CTROLYTIC MAGNESIUM", by 
C.S. Roberts, Journal of Metals AIMS, 
Sept 1953, p 1121-1126, Univ. of Calif. 
Inst, of Engineering Research, Berkeley, 
31st Technical Report 

THE EFFECT OF THERMAL-MeCHANICAL HISTORY ON 
THE STRAIN HARDENING OF METALS, Univ. of 
Calif., Berkeley, Engineering Research Pro- 
ject First Technical Report dtd March 1948, 
Figure 4 

INTERMETALLIC SOLID SOLUTIONS, Trans. AIMS 
(1934), 111, p 75-93 



25 



APPENDIX I 
Table IV 

Chemical Analysis of totals In Percent by Weight 



H3.97NO, 


Percent Me 
(analyzed) 


Percent Cd 

(by subtraction) 


70 





100* 


71 


1.98 


98.02 


72 


6.79 


93.21 


73 


19.15 


80.85 


1U 


37.08 


62.92 


75 


62.35 


37.65 


69 


100* 





•The cadmium magnesium alloys were prepared from 99.9 


pure cadmium. The 


impurity analysis of the magnesium follows : 


5 


Impurity Percent 





Al 


0.012 


Ca 


< 0.01 


Cu 


< 0.001 


Fe 


< 0.001 


m 


0.001 


Ni 


< 0.001 


Pb 


0.005 


Si 


0.022 


Sn 


< 0.01 


Zn 


0.052 



26 



APPENDIX H 

Typical Strain vs Tin© Curves 
for each ^ H Determination (Figs. 6 - 16) 



27 



TIME t2 (hours) at Temp T2 




FIG C 



ISO 200 220 240 260 
TIME tx (sec) at Temp T^ 

6 STRAIN vs TIME FOR 99 .85 PERCENT PURE MAGNESIUM (AS 
EXTRUDED) 



TIME T2 (hours) at Temp T2 




220 2^0 260 280 300 320 340 
TIME tl (sec) at Temp T^ 
FIG. 7 STRAIN vs TIME FOR 99.85 PERCENT PURE MAGNESIUM 
(ANNEALED 300°C) 



28 



TIME t 2 (hours) at Temp T2 
15 16 17 18 19 




160 180 200 220 240 260 280 300 320 
TIME ti (sec) at Temp T^ 

FIG. 8 STRAIN vs TIME FOR 99.85 PERCENT PURE MAGNESIUM 
(ANNEALED 500° C) 



TIME t 2 (hours) at Temp T 2 
10 11 12 




240 260 280 300 320 
TIME ti (sec) at Temp T^ 

FIG. 9 STRAIN vs TIME FOR 88.4 ATOMIC % Mg and 11 6 ATOMIC 
% Cd ANNEALED 5CO»C 



29 



60 



55 



Q 



i 50 



45 



CO 



40 



35 



!*■* 15 



TIME t2 (hours) at Temp T2 



H^L 



16 16.? 17 



MAGNESIUM 73 »2 ATOMIC % 
CADMIUM 26 .. 8 ATOMIC £ 
ANNEALED 475 
2000 PSI 




360 330 400 420 440 
TIME ti (sec) at Temp T^ 

FIG. 10 STRAIN vs TIME FOR 73.2 ATOMIC % Mg and 26,8 ATOMIC 
% Cd ANNEALED 47 5° C 



TIME t2 (hours) at Temp T2 
15.5 16.0 16.5 17.0 17.5 18 18^ 19 




4000 



5500 



4^00 5000 
TIME ti (sec) at Temp T x 

FIG. 11 STRAIN vs TIME FOR 73 2 ATOMIC % Mg and 26.8 ATOMIC 
% Cd ANNEALED 450° C 



30 



TIME t2 (hours) at Temp T 2 




1300 1500 



FIG. 12 



1700 1900 2100 2300 
TIME ti (sec) at Temp Ti 

STRAIN vs TIME FOR 52.2 ATOMIC % Mg and 47.8 ATOMIC 
% Cd, ANNEALED 350°C 



60 



55 



\ 



50 



45 



CO 



40 



35 



80 



TIME t 2 (min) at Temp T 2 
90 100 110 120 130 



MAGNESIUM 52 . 2 ATOMIC % 
CADMIUM 47 p 8 ATOMIC % 
ANNEALED 380 
750 PSI 




140 



582°K 



= 546°K 

s 625X10~ b /sec 

S 1136X10" 6 /sec 



50 60 



FIG. 13 



70 80 90 100 

TIME ti (sec) at Temp T^ 
STRAIN vs TIME FOR 52.2 ATOMIC % Mg and 47 08 ATOMIC 
% Cd ANNEALED 330 C 



31 



60 
55 



3 50 
I 



3 45 



15 



TIME t2 (hours) at Temp T2 
16 17 18 19 20 21 



22 



23 



CO 



40 
35 



1 1 1 1 - 

CADMIUM 74 o 8 ATOMIC % 
MAGNESIUM 25.2 ATOMIC % 












ANNEALED 475°C 
2000 PSI 


b*s 




• 






































/ 1\- 477°K 
To = 363°K , 


fy""! 












6 1 
€2 


= 300X10 "°/aec 
s Oc706X10 =b /sec 



140 160 180 200 220 240 260 
TIME ti (sec) at Temp T^ 
FIG. 14 STRAIN vs TIME FOR 74.8 ATOMIC % Cd and 25.2 ATOMIC 
% Mg 



60 

55 
I 

O 

? 50 

as 
3A5 

I 

CO 

40 
35 



14 15 



TIME t2 (hours) at Temp T2 
16 17 18 19 20 



21 



1 1 1 1 

CADMIUM 91o4 ATOMIC % 
MAGNESIUM 8.6 ATOMIC % 


/ 






Y 




ANNEALED 2l6°C y 
2000 PSI £ / 


r 




A 










Tl 






/? Z 
























T 1= 434°K 
To=378 K 
















€l=54 


„6X10" 
068X10 


/sec 
" 6 /se< 



900 1000 1100 1200 1300 1400 

TIME tx (sec) at Temp T-^ 
FIG. 15 STRAIN vs TIME FOR 91.4 ATOMIC % Cd and 8 6 ATOMIC 
% Mg 



32 



TIME t 2 (sec-l(P) at Temp T2 



? 

Q 



55 

CO 



70 




2 


! 3 


4 


5 


6 




7 


fi 


9 




65 


PURE CADMIUM (99o8£) 
ANNEALED 200°C 
2000 PSI 




/ 








60 










55 






















50 










1 / 


s 


/ 


r 






45 








m / 


/• 












40 








T y 


y*s 




T l 
*2 

^2 


= 453°k' 
- 371°K 


35 
















861X1C 
4.54X] 


r 6 /sec 
.0" 6 /sec 

i i 



40 50 60 70 80 
TIME ti (sec) at Temp Ti 
FIG. 16 STRAIN vs TIME FOR 99o8* PURE CADMIUM (ANNEALED 200°C) 



33 



M423 



Thesis 
MW3 

c.1 



8464 



2 

Medbury 

Activation energies $b 
for high temperature 
creep of magnesium- 
cadmium poly-crystalline 
alloys . 




28464 

Medbury 

Activation energies for l" 4 gh 
temperature creep of magnesium- 
cadmium polycrystalline alloys. 



SSn energies for h,g hi== 




3 2768 001 88563 5 

DUDLEY KNOX LIBRARY 



HI 
Willi 

mm 







','■'■■'' 












'':.»•. 



.',:. um 






"i •!-■ I 

, , w ... ■'■.',! 



fffl&KMK 



lEj ■ f :,-'•' i.' t w . ! , .i;i 1