LIBRARY
Michigan State

 

|

 

| University

 
PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRC/DateQue.p65-p.15

 
CONSTRUCTION OF THE EQUILIBRIUM DIAGRAM
FOR AN ALIOY.

A Report Submitted to
The Faeulty of the

Michigan Agricultural Oollege

by

Keith A. Weston

Candidate for the Degree

Bachelor of Science

June, 1921
’
CORTEN TS.

PAR? I
PAR? II
PAR? ITI

INTRODUCTION
METHODS
RESULTS

104075
PAR? I.

The stady of the action of metals under differ-
ent temperatures has become of immense importance during
the last few years and so some mans was sought by which
the action of a particular metal could be represented

graphically as it passed thru the different ranges of

temperature. The outcome of this has bem the equili-

brium or constitutional diagram. At the present time

the iron-carbon and several other common alloy diagrams
have been thoroughly worked out but there are still in-
numerable possibilities in alloys which are not so well
known and for which it is necessary to have the eqili-
brium diagram in order that they shall be thoroughly under-
This Report was the outcome of the desire on the

stood.
part of the author to have a more thorough knowledge of

the method of experimentally constmoting one of these
diagrams.

Some difficulty was experienced in finding a
suitable metal for the experimental work, but it was
at last decided to use an alloy of magnesium and copper
made by the Dow Chemical Company of Midland, Michigan.
The well known Dow metal is an alloy of aluminum and
magnesium end it should be kept in mind that the alloy
used in connection with this Report is not the seme.

At this point it may be interesting to note

some of the interesting facts concerning Dow metal as

leo
it may be considered a very close relative to the metal
worked with. The Dow Chemical Compeny found that there
was large amounts of magnesium in the brine pumped from

wells st Midland and which was lost when the brine was

finally thrown away. They accordingly have been extract-

ing the magnesium from the brine by electrolysis, but

as magnesifm is not a metal that is used mech, there was

no merket for it. The Company has been trying to make

a market for this metal by using it for pistons in

automobiles since it is extremely light. The coefficient
of expansion of Dow metal is higher than that of the cast
iron cylinders which caused a great deal of difficulty
in this field although at the present time there are
some engines in actuai use with Dow metal pistons and

they seem to give very satisfactory service. |
It seems strange that the Dow Chemieal Company

in all their experiments with this metal had no equilie
brium diagram for it. In the winter of 1920-1921,
Professor H. Le Publow of the Michigan Agricultural College
eonstructed the equilibrium diagram shown on the next

page for magnesium-aluminum alloys up to 35% aluminum,

Some of the photomicrographs of these alloys follow the
constitutional diagram and it is interesting to compare

them with those accompanying this Report of the mgnesium

copper alloys.
The construction of an absolutely accurate

equilibrium diagram would have been impossible in the time
allowed but the principles of the work will be @-seri bed

in the following pages.
aLe

 
 

 

 
 

A
568). ag

 

18 as received...

pay

All above photocra

A

-_—T——
PaRT If.

There were six semples of the magnesium eopper
alloy that were procurable end they haa the following

ecomposi tion:
253 0.55 Co Yr
254 1,03 ; os
255 1.86 "
256 53.43 © "
267 7.12 "
258 11.65 "

The per eent of copper only is given, the rest of the
metal being menesium. Note should be taken of the
numbers corresponding to the vazious compositions as the
alloys will be referreé to from now on by their numbers.
It was desirable to first find the transform-
ation points of the different samples. This was done on
a Brown Transformation Point Recorder from which the
following set of curves was produced. Page 6 shows a
diagram of the wiring of this Instrument, the operation
of which is briefly as follows: The neutral body is a
piece of nickel 1 inch long and 3/4 inch wide. The test
piece should be about the same size and in the middle of
each piece there should be a #10 hole. Two thermoeouples
ere used as shown in the wiring diagram, one being placed
in the test piece and the other in the neutral body. The
nickel body would, if not plaeed next to the sample, give
@ smooth heating and eooling curve as it has ne transforn-
ation points but when placed next to the test piece in
the furnac3, it will sve up heat to the sample at one of

the alloy transformation points on heating and Phsorb
heat from the alloy at these points on sooling. |

Ro
In the recording Instrument, the movable element
is wound with two coils in fixed mechanical contast bat
well insulated from wach dther, forming a differential
system. When the instrument is used for a transformation
point determination, the left hand cirauit records the
true temperature throughout the test. The right hand
cirouit contains an interrupter run by clock works by
means of which the circuit is made and broken every fifteen
seconds. When this circuit is closed the current from
the thermocouple in the test piece opposes that of the
test thermo-circouit. As one winding of the movable element
has a sensitivity of about ten times that of the test
thermocouple circuit a magnified jog or buck will osour
at the transformation points in the difference ourve,

A difference curve was made for each of the
samples and as some of the points at the low temperatures
were very indefinite, curves were made for three of the
Samples running the temperature up very slowly with the
results shown on Page R, These curves show very little
that cannot be seen on those run up to higher temper-
atures so they were not made for the remaining samples.

@sre had to be used not to run the temperature
of the zlloy up too high as this metal being composed of
chiefly magnesium burns very easily. By excluding the
air as well as posaible it was possible to run the
temperature up as hich as 975°P, without burning the
metal, but on cooling, the outside of the test sample
 Dacame covered with a fluffy black oxide.

In order to plot the eqilibrium diagrat it is
necessary to know the per cent copper ut which the alloy

=6o
 

: Pd

papesaay — ¢4S

We Ek
LAO#H oie eh Sued {

eur2y IS

Pree ment Tl Lier VEEN

           

      

oe RIMS:
S Ne ACA)
4c
SOL SISA
Yd

 

ae J

 

 

 

NOM eT lan 4 hdd
~~

f19D Y?0T LW il athe

   

—/

 

 

 

 

 

 

 

 

vp

 

 

 

 

 

 

 

99 4/ SOlSegsy

 

oS eee ANA

 

 

 

 

 

 

 

 

 

 

 

~Pa- nae ee A fA ee) Wg Ie Te ee Loe ll

 

 

 
{
|
;
an
' ol
hn Oe, <2
; 400 ™
aly)
} »
™
.
_
| tao
! hs
>a
ss
/
_ | }
i» :
yi
~~ ’
ati
«
200

 
 

/
|

  

 

   

NT fap asette +
cee 2 ee
700 nn EY Pee nee cL

- “TT,
ea

 
 

 

 

 

 

* Lm ns
: ot care a ae

oie

ed a ee ents

 

 

 

 

 

 
 

 

 

 

 

 

 

 

 
 

 

 

 

 

 

 
 

 

oe

 

 

a attri ia ee.

fe =i

 

 

 
}

|
/
|
|

1000

;

Pt

ioe

   

    
 

ih amie UTE
wile

;

 
 

 

 
 

 

 

 

ES

 

 

 

 

 

 

 
is all eutectic, At this point the metal will go from

the solid into the liquid stage immediately, therefore,

@ melting point determination was tried on each sample
to see if one of the samples was eutectic, The rapidity
with which the alloy oxidized and the tendency to burn

at temperatures above 900°F, made this determination ex-

tremely difficult. <A hole was drilled in the sample large

enough to asocommodate the end of a mall thermocouple.

The sample was then put in a erucible with the thermocouple

in place, covered with a heavy layer of fine charcoal,

and set in an electric furnace. <A Brown Heat Meter was

used to record the temperature of the piece as it was

heated.
During the melting period there would have been

a@ slowing up of the inorease in temperature until the
metal was all melted after which there would be a rapid
rise. If the piece had been composed of eutectic there
would have been a point of constant temperature while the
metal was melting and then a sudden and rapid increase
in the reading of the pyrometer recorder, It so happened
that there was no piece of pure eutectic among the samples
and they started to burn before the melting temperature
wes reached, even with the heavy coat of charcoal surround-
ing the metal.

As it is absolutely essential that the per cent
composition of the alloy at the eutectic be found in order

to construct the equilibrium diagram, it was necessary to

find it by photographic measurement. Eaeh of the six

aJo

 

 

 

 
 

#252

.

Received.

As

 

Same as above but

improperly focused.

B=
 

Received,

As

G
—-J—
 

#255

fs Received.

ite
 

As Received,

oe] Jae
LAE teat
PL eT Ty

ks ae S A

 

#257

As Received,

 

Same as above but

etched slightly deeper.
SAS :

7

 

#258
As Received,

 

put

Same as above
etched slightly deeper.

ae
samples was polished, etched, and photographed. Phese
photomicrographs are shown on peges (8-13) inclusively
and are all 100 diameters larger than the actual semples.

These alloys of m@gnesium and copper are very
soft and it was necessary to use extreme care in polishing
them. The best way to polish these samples was to: lst,
smooth up the surface to be polished with a fine carborune
dum wheal using plenty of water; 2nd. take off these rough
sorstehes by using Alundum F on a eanvas covered wheel
and polishing crosswise to the first soratchesy 3rd.
polish en a broadcloth wheel using Alundum XP as an abraes-
ive and polishing crosswise to the soratohes made by the
eanvas wheels 4th. polish on @ brosdoloth wheel using
levigated alumina in suspension in water and spraying it
on the wheel by means of a wash bottle. It should be noted
here that plenty of water should be used in order to
euccessfully polish the picsce, |

The photomierographs were taken on a Bausch &
Lomb Metallurgical Microscope, An electric are was used
for the licht and a green filter was placed over the eye-
piece of the microscope in order to soften the rays be-
fore they ceme into contact with the photographie plate.
fhe adjustment af the light is a very important essential
for zood phcetographs. The beam of light from the are
must first be exactly focused on the aperture entering the
microscope. To do this the diaphram on the are was olosed
leaving only a very smell cpening méd@ this small beam is
then adjusted on the sperture. The diaphram on the micro-

olde
scope was then alossd and the quarts reflector adjusted
until the beam as reflected on the ground glass screen
comes in the center then vhen the diavhrams are opened
the light on the soreen is at its mximam intensity.
Proper foeusing is also very essential for a good picture,
This may be best accomplished by examining the image on
@ clear portion of the glass with a 7.5 or a 10.0 eye-
piece and adjusting until the image as seen through the
eyepiece is as sharply outlined as possible. The photo-
graphic plates used were Cramers 5-1/4 x 4 Ilentern Slide
Plates.

The plates were developed in the following

solution:
Hydrochronon S grams.
Sodium Sulphite 7 "
Potassium Bromide 16 Drops.
Sodium Carbonate 7 grams.
Water 250 6eGe

and the following solution was used as a fixing beth:

A. Water (1 gallon)

128 ounces.

Sodium thiosulphate g2.2C=+«*7
B, Water sei”
Dry sodium sulphite a
Sulphuric acid C.P. 1/2"
Powdered chrom aiun eg *

Pour B into A while stirring well.

After the photomicrograph of each piece md bem
made a tracing was mde of eaoah photograph on cross section
peper. The number of squares of eutectic in each sample
per unit area was then caloulated and hence the per cent
of eutectia in that pertionlar sample could be computed,
In order to be sure that the average per cent should be

el] Be
,
ij
;

;

’

HCI TF

[.

|

|
a
|

/

|

|

 
 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
found, three tracings were mde of the same photograph
and these averaged. Blue prints of these tracings sre
shown on pages ({ 16-17-18 ). The number of squares of

eutectic in each of the photographs and the average are
es follows;

#253 #254 #255
Sheet 1 62 13 £19
Sheet 2 41 58 189
Sheet 3 45 74 192
Average 49 68 200

#256 #257 #258
Sheet 1 227 137 ene
Sheet 2 178 183 271
Sheet 3 237 171 282
Average 212 164 276

There are 900 of the emall sqares in each one
of the tracings so if each of the above averages are divided
by 900 the result will be the per cent of eutectic in the

respective samples, which is as follows:

#253 #254 #258 #286 4 8=—6 #2. 7 #258
5.45% 7.46% 21.9% 23.6% 18.2% 350.5%

=~19-
FaRT Jil.

It now becomes necessary to find the composition
of the eutectic frem the data available. In order to finé
the relation of the various per cents of eutectie of the
geamples in relation to the per cents of copper, the graph
on page 24was drawn. It can be seen that these points vary
widely although #257 and #258 lie on the same line and the
rest of the points lie very close to & mean line. As the
amount of eutectic in samples 253 and 254 is not large the
error in tracing would probably be very large so they hea
better be negleoted in ooming to a conclusion in regard
to the constitutional diagran. |

If these points are now plotted ag in the curve
on page 2£5it will be seen that the eutectic would occur
at 37% copper if the two points ‘25y and 258 were taken as
correct or it would be at 14.75% copper if the points
254 and 256 were to be assumed as correct. The points
given by samples 253 and 266 may be regarded as incorrect
as the line drawn through them would show a eutectic com-
position of 8.5% eopper md 91.5% magnesium which we know
to be incorrect as a sample of 11.65% copper and 88.35%
magnesium was not a eutectic. If line B is considered as
correct, showing 14.75% copper in the euteotie, then
sample 258 whioh contains 11.65% copper would from this
curve contain 78.8% of eateetic.. But from actual measure-
ment and from inspection of the photomicrograph of this
specimen it will be seen that this cannot be true; there-

fore curve B earmot be used in determining the eutectic,

of 0=
The elimination of curves A and B leaves only surve 0
which, since we are dealing with higher percentages of
eopper in the samples forming this curve and therefore
the chances of error are liable to be less, we will assume
to be correct.

fo sheck this assumption a difference curve was
ran on the Brown Transformation Point Recorder, running
the temperature of the cample as high as possible in order
to determine the melting temperature, As has been stated
before in this report, it is very hard to melt this alloy
without burning it up before resching the melting point.
In order to lessen the chances of burning, the sample to-
gether with the neutral body was tightly packed in asbestos
and this arrangement gave the curve shown on page Q. The
same transformation points ean be noted on this curve that
have been noted above but at temperature of 1040°F, there
is a decided although emall jog in the difference curve and
from the rapid rise in temperature that follows this is
undoubtedly the melting point. The rapid rise in tempera-
ture shown by the succeeding points would give rise to the
belief that shortly after the melting temperature was
reached the metal started to burn. Inspection of the metal
after it had cooled in the furnace showed it to be burned.
If the point of 1040°F, is assumed to be the melting point
as shown by the difference curve, then the corrected melt-
ing point would be 1120°F, Wow referring to the eqili-
brium diagram as it has been so far conetrueted with 37%
eopper as the eutectic and 1205°F, as the milting point of

a2le
magnesium and a point representing 11.65% copper and
88.35% magnesium and melting at 3120°F, would lie almost
on a straight line between the eutestoid point and the
melting point of magnesium. This shows that the melting
eurve for this portion of the diagram is a straight line.

From inspection of the difference curves it can
be plainly seen that there were three traneformation points
that occur in each sample. These points oocur at the
Zemperatures 190, 610 and 840 as shown on the curves,
There was an error in the recorder of 100 degrees too low
at a temperature of 1300 which mde a proportional temper-
ature correction necessary for the transformation points,
These pointe as corrected would come at 205°F., 655°F,, and
905°F, The trensformation point indicated by the line of
905°F. eannot intersect the 100% Mg. composition ordinate
@s pure magnesium has no transformation point at that
temperature, The difference eurve for semple 263 which oon-
tains only 0.55% eopper (page L) shows a transformation
point at this temperature very plainly which would indieate
that the amount of g011d solution in the alloy is very
slight. This solid solution area has been indieated on
the equilibrium diagram although its authenticacy cannot
be vouched for as no samples of less than 0.55% eopper were
obtainable.

Examination of the photomicrographs shows two kinds
of structure very plainly,namely; large crystals of pure
magnesium and a eutectic structure in between these orystals.

dust what the nature and structure of the eutectic
is, as shown in the photomicrographs, could not be determined

 

~28-
fn the time allowed mt from a study of the difference
curves and the tentative equilibrium diagrem offered it
ean be seen the stmeture mast be very complex.

Comparison of the equilibrium diagram offered
with this Report with that offered by Professor Publoew
for magnesium-aluminum alloys shows them to be very similar,
the eutectic points eoming within 6% of eaon other. The
solid solution area on the Mg-Al diagram is, however, very
mach harger than on that offered for the Mg-Ou series.

The transformation points also ocaur et different points
but generally speaking the two diagrams are very similar,

One of the things peoliar to this metal, that
is worthy of note, is that there is practically no lag of
the trensformation points in the cooling curve. It maturally
follows then that this series of alloys is ineapable of
being hardened by heat treatment although som very interest-
ing changes in the grain size can be obtained,

The resalts of this report are summed up very

completely in the equilibrium diagram offered. (Page 26)
Although some of the remilts obtained in many eases heve
been indefinite and depended upon deduetive reasoning from
the work performed to reach a definite conelusion, this dia-
gram should serve as a tentative guide for the action of
Mg-Cu alloys up to 37% copper until a more thorough investi-
gation of this field will prove or disprove its validity.
It ean be plainly seen now that as stated in the introduct-
icn the field in the metallography of non-ferrous alloys is
very great as practieally nothing is mown of their struct-
ure and their action under varying conditions.

~23~
 
 

 

 
 
 

 

 
570 904

30577

4
 

MICHIGAN STATE LIBRARIE

nN Uh HIM

1293 02236 0790

       

 
 

be Me
7 — ie _ aw

 

        

ee

ec seee  ee

 

ear =. ~~ —
L ? é

2. _.. _ 372

  

>
te

lineage

ar i
 

MICHIG. a STATE LIBRARIE

MUN

1293 02236 0790

       

Mob
L

 
 

     

= . & | = .

a
i
‘

   

 

Se
ee

= ~S —
; ad : 7 ?
a . a wits 5% '
oo 1 _— >

4 SO ty — 2
Ng

 

=<

 

>
a
 

MICHIGAN STATE LIBRARIES

vl TT

1293 02236 0790

       

hy

 
 

 

“Ss
a —- — - ae
‘. — i CT
ent) I pena oa
+
‘ . - n } ‘
oS e 4 >

 

 
  
   
   

i

4
nse. an
ee _ itt -
er
1 a
- Am a
' a. Z

~~...
 

MICHIG. AN STATE LIBRARIE

MMM us mV

1293 02236 0790

       

Mel
bh

 
 
 

     
   

ATE LIBRARIE

il ii i

1293 02236 0790

 

    

Mb
As

 
      

 

~ ~~
oe oo. er re NY —
Be r & -y,

 

 
 

SO —_ gee . _ —
Hi %

ge UN

fate

a
le
a: r
+
7 ed hn

      

_—-
 

MICH aAN STATE LIBRARIE

nn il TU

1293 02236 0790

       

7"

 
 
 

by

 

 

 

     

 

   

 

 

MICHIG. AN STATE LIBRARIE

UU

 

NN

1293 02236 0790