H 532;";__5;::2_:_:_:____:_:_:_:_E3::

198
18H

11T

 

.5,

'.
'1‘.
R‘O
. 5

;
v
a

agree a";

ts
a ad

;‘

i.‘
1

THESIS

0-169

WWWWWWWWWIWWW WWWWW

293 01093 1156 ...-

 

This is to certify that the

thesis entitled

"An Investigation of the Notch
Sensitivity of Several Materials
in Torsion-Impact"

presented by

T. Bruce Henderson

has been accepted towards fulfillment
of the requirements for

JL_S_’ degree in ME 0

WWW

Major professor

 

 

 

 

, rhfi—y . -77 ‘A2 _ 2.. .7 -_-

AN IN 'ESTIGATIOI‘E Oi" "[17:17 I‘WOT‘CH S'WSITI'V'ITY

OF 53 BHAL MATERIALS 1H lQmSION—IMFACT

Submitted to the School of Graduate 5 J ies of Michigan
State College of Agriculture and App ied Science
in partial fulfillment of the sequizements

for the decree of

H

rtp r‘ 1t‘1w- ’\i'1 ."‘ nrgx \
1.111011le bf LJUihuiv...‘

Department of Mechanical Engineering

TH ESIS

W/ 2/5'3

Acknowledgement

Q.)
(1"
,4.
O
:3
d-
0

The author wisfies to express his apprecit

Professor Paul J. DeKoning of the Applied Mechanics

)

L

epartment, School of Enrineeriny, flichiran State
College. Professor Defioning generously extenficd his

time and knowledge on the details of the investiyation
and the instrumentation of the Torsion-Impact Macnine.
The author further acknowledges the helpful interest

L.)

of Ur. Holland T. Minkle of the Mechanical Engineerinc

'5

Department, School 01 Engineering, Nichiyan State College.

‘lflti.i $26;

TAbLE OF COETEKTS

Page

I. Introduction

II. Descriptive Material

A.
B.

C.

Historical Sketch

O O O O O O O O O O O O O O O

The DeKoninc Torsion—Impact Machine

The Torsion-Impact Specimen

III. Experimental Material

A.

IV. Discussion and Conclusions

V. ist of References . . . .

Variation of Energy of Rupture with Diameter of

Test Section

0 O O O O O O O O O O O O O O O O 0

Variation of Energy of Rupture With Notch Anyle
for Two Ductile Materials . , , , , , . . . . .

Variation of Energy of Rupture with Notch Angle

for Two Irons .

but-J

28
3O

35

TABLE uF EIGUALS

Picture, torsion machine setup and relatefi
equipment . . . . . . . . . . . . . . . . . .
Plot, energy of rupture vs. flywheel speed . .
Two typical oscillOUraph tapes . . . . . . . .
a. Leaded brass, square notch
b. ZhS-Th aluminum, full radius notch

Description of standard test notch

a., Picture, whole and broken specimens . .

Q;

h. Drawing, standar test Specimen . . . .
Picture, machine-tool setup for specinen
production . . . . . . . . . . . . . . . . .
Energy and torque of rupture vs. diameter
of test section . . . . . . . . . . . . . . .
Energy of Rupture vs. diameter of test
section, log-log . . . . . . . . . . . . . .
Comparison of notch sensitivity of two ductile

materials 0 O O O O C O O O O O O O O I O O 0

Comparison of notch sensitivity of two irons .

H
l
‘33

ll

1h

17

G3

~ *,"th,:"\*\“’romIo}Y
1 l , ‘ .
I. . _L-‘ A l\/I"‘"‘ ’\' L V‘ L1

 

Many mechanical devices have members suhiected to

torsional stresses; often the designer is forced to impose
discontinuities in the stressed section such as shoulders,
grooves, keyw js, splines and others. The factor of impact
is also usually present.

The purpose of this investipation is to present the

features of experiments with notched specimens for several

materials in torsiondimnact; the DeKoninp Torsion-Impact

A

A

Machine was employed to study certain asrocts of notch
sensitivity in these materials. Some problems encountered
and discussed are the mass production of accurate test
specimens, the variation of enerpy of rupture in torsion—
impact with diameter of test.section, and the variation

of rupture energy with notch anrle in some typical ductile
and brittle materials.

_The DeKoning Machine was recently offered as a new
instrument in the field of materials testing. The scope
of this investigation was designed to facilitate specimen
production and testing technique, standardize the test
notch, and indicate the general variation of data that
may be expected from this machine in its present state
of development. Also, data is presented concerning the

behavior of several materials in notched torsion-impact.

II. DESCRIPTIVB MATEMIAL

A. Historical Sketch

 

The general field of impact testing has been recognized
as extremely helpful in the selection of the proper material
for a given shock loaded member. Shock or impact leading
causes a severe increase in stress in a member calculated
to resist a nominal static load, often times causing
failure. Impact overloads may result from such simple
situations as back-lash in a pear, starting and stopping
machinery, or even the sudden.arplication of a load to a
member without breaking contact.

It follows that there is a decided need for information
on a material's ability to resist such loading, enabling
the designer to cope with the problems of higher operating
speeds, reduced masses, and lower factors of safety. It
was with this thought in mind that the American Society
for Testing Materials met and published a symposium on
impact testing.22 Sam Tour opened the discussion by
‘ labeling impact testing as an "art" and "science" and
making a plea for utility and the use of non-standard
impact tests.

A few years prior to this meeting, two unique and
non-standard impact testing machines were developed to

O

rupture a specimen in torsion, rather than the ordinary

f1)

tension or bending type impact. Luerssen and Greene
designed and built a torsion-impact machine and tested
specimens of plain carbon, and high alloy steels.lo The
final standard specimen was one-fourth inch in diameter
with an effective length of one inch, allowing a compromise
between energy absorption in rupture and fairly Food heat-
treatability. The data of their tests showed some sub-
stantial peaks in rupture energy indicating that optimum
tempering conditions do exist for that particular material.
Further, these peaks were not demonstrated in the usual
standard Charpy and Izod type impact tests, thus assuring
the position of the torsion-impact test.

In 1933, Kititosi Ithihara published the first of a
series of articles on the results of torsion—impact tests
conceived at the Tohoku Imperial University, Senday,
Japan.18 The test section was 8 mm. in diameter and 10 mm.
long; torque measurements were read from a spring deflec-
tion, and twist angles recorded by optical means. The
investigation included tests run with a variable shear
rate and in general indicated that maximum torque increased
with increasing velocity of impact. Impact values offered
by Ithihara indicated substantial increases over static
values; a 25 Percent increase in torque and a 10 percent

to 20 percent increase in angle of twist was reported.

B. The DeKoninp_Torsion-Impact Kachine

In 1950, Paul J. DeKoning designed and built a
torsion-impact machine embodying several new themes.3
The standard test specimen had a severe notch and an
extremely small gage length, where failure occurs. This
allowed another variable, notch sensitivity, to be intro-
duced into the test with its subsequent mechanical and
metallurgical features. Most shaftinp, and other members
subjected to torsion, have threads, notches, shoulders, and
various discontinuities that present the problem of selecting
a material with proper strength levels, along with a low
notch sensitivity or resistance to stress risers caused by
abrupt changes in section.

The DeKoning machine, as pictured in Figure 1, in—
volves a rotating flywheel to produce and measure the
energy of rupture, and electrical strain pages, with their
related equipment, to measure the resisting torque of rup-
ture exerted in the specimen support. A detailed descrip-
tion of the machine in its present stare of development is
as follows: the stationary specimen holder, mounted
rigidly in the tail-stock of the machine, resists the
exerted tornue of rupture through two small pillars, equi-
distant from the central or twisting axis pf the test piece.
Mounted on these pillars are four 55-h resistance tvpe
strain pares, arranred with two in series in parallel With
two in series, and a "dummy" or compensatinp rage on the

other lea of a wheatstone circuit to balance out temperature

pcefimwzdm bepwamm baa ocwfiomz

powQEanowmaoe mcHCome one

 

Fig. 1

\fl

differentials. The net result of the four active gages

is to give average readings of axial strain on the pillars,
without error introduced by bending. The signal from the
strain pages is fed through a Brush Strain Analyser, and
thence to a Brush Magnetic Pen Oscillograph, recording a
"pip" on the oscillograph tape that may be calibrated
directly into inch-pounds of tornue.

The flywheel contributes the necessary energy to pro-
duce rupture. It is brought up to the nominal testing
speed, h20 rpm, by a small, 110 volt, A.C. driving motor,
connected through a Variac transformer for speed control.
The initial energy level is determined approximately with
a stroboscopic synchronizing device, consisting of a wheel
driven by a rubber belt from the shaft of the flywheel and
a neon glow lamp, shining on the wheel rim on the operator's
side.. Stripes painted on the wheel rim are calibrated to
"stand still" when the proper initial test speed is reached.
‘The very accurate Speed determination, needed for energy
results, is made by means of a unique photocell tachometer.
Painted on the flywheel rims are white lines on a black
background, suitably spaced, that reflect pulses of energy
from an exciter lamp to the photocell. The pulses are
then fed through a Brush D.C. Amplifier and thence to the
oscillograph tape, causing the deflection of a second os-
cillograph pen. These deflections may be calibrated in
terms of the energy state of the flywheel and will be dis—

cussed in detail later. The advantage of Such an arrange-

ment is that there is no mechanical connection to the
flywheel for an energy measurement and thus no energy loss
through friction. The flywheel shaft is mounted in bear-
ings such that the resulting friction energy loss is
negligible over a short time interval.

The entire operating apparatus is wired through a
central control panel to facilitate testing. The circuit
includes a timing switch for proper flywheel engagement,

a switch to energize the photocell tachometer, and a
Aswitch to start and stop the oscillograph tape for the
duration of a test run. Thus, the testing operation is
greatly simplified and requires only the clamping of a
specimen in a dog and in the tailstock support, moving the
tailstock into position, checking the flywheel speed with
the stroboscope, and energizing the system with a master
starting switch. The completed operation takes less than
a minute allowing many tests to be run in a relatively
short time.

The energy state of the flywheel, rim, and shaft at
any speed, is a function of the mass moment of inertia

of the system and the square of the angular velocity.

That is, E = i I w2
where E = kinetic energy,
I = mass moment of inertia,
and wr= flywheel speed.

If the spacing of white impulse strips on the flywheel

rim and the diameter of the flywheel are known, the angular

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

               

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

71 A 4 7 l 4 1 71
. 7 . a 7 . . fi 7 .7 7 .7... . 7 . .. . .7.. 7... 7.
. . 7 . .. ... . . 7. . . . a .7.“ . 7 . 7.... 7...7.... 7 . ....
. k . 7 7 .77. .77.7m.7 7.. 7. . .7 7 ......770 7 e, . < 7 ..7.. .77 7777.. l
. 0 e 7.. t I I “ V 7 - dt I b O l O I717~. v p OI I 775 70 A V 70
. l. w v . .
ll777o77l7i¢77777777737 77.7.0 77.7777f.777..97.77. w 7H7 47 a 77 lOIfllllVOATtltlolllkiltil'l 0 ll”
. . _ e .9. .. 7 7 o. .. v7. . .77 .7 ~ 7 . .77. 7 . 77..
. . . m . .H 7. .s .7 7.. 7.. g 77.. . 7 . .. 7....fi ..77l.L.A
. . 7. 7 .777 7 .. 7‘. . ..a 77 o. ‘7 7 . ..7777 7.77A7v!V7
7 . . * tab . .7. 7 7.... . . 7 . 777r.7 h 7.7 ...
1|. . l
. A . o 7. HF . 7 7M .7 .7 7 7 . . 7. . 7 ..l v 77. 77..
7 . 7 x, 7 7 o 7 7 .7 .h777 H 7 7 7 77 7.4 .777 c. .7 o .7 3w . h.- 7 77.: 77v.
. 7 . . H - . .. .7.... 7. _ . . f... 7 , . . _7.. . ....r......
l . 7. . . . fi .7 . 7 7 7... 77 7M . *7- . * W. 77.4. 7.7.7 7.7
:7;-;a 7:77:777737 777: 77+7 7 . 7- 7 7. . 7 . 7
_ . v . . . .. .. .. 7 7+. .4. *7. ... .....7. ... ...... . 4 .7.7+..
. . . .. 7 77. 7 » 7. 7 7 . . .7 .7 .77. . . . . 7... . ..17+ 7777
w 7—7 . . .. .H ”7 7 *7 7 . .H.. . . 7 7 77” 7 7.7V 77. . .7 a477l7;7.ho7.
. 7 . . 7. 7. s... 7 7 7. 7. 7 7. 7 7 7 7. .77. .7 77 7 . .
.. . r a n ml.7 w w 7. ..m ...H.... .7m . .. 7 ...M... 7.. .7 7. . .77 7 7
u . P . 7 ~. H .7 .7 7... 7* . .7 ...oo. 7 . . .7 . . H. 7 777 .0'7
. .. . . .. 7. 7 7 7 . . . 7 7.. .. . . . 7 7. . 77 . 7 . .77 7.
7 _, all r H H a .
71.47.77ll77777l7 l 7777:7l7 | 71117 T 4 4 .Y 7 l 7 _l l
# 7 . . o 7 a. .* m 7 . H: 7 7 4.‘ 7 7 7 .fi A 7.7".7 . 7 . . .7j7.
. 7 . . . . 4 7.“ . . 7. .. . 7 ..’.7 7. 7. . 7. 7.. 7. 7. . 777.
. . _ . . i» . . . 7 77 .. 77 . ..w.. . . 7H 7 .. .a. 7 77 7 7
_v 7 . . a . 7W .7. 7 7 .v . . .7. . .. 7.” .77..7. 7 . .l .
.P 7 4 4 l l L7 » l7
. . w . . 4 F . H . . y. . 77 7 .7 . 7 .477 .7 .7 4.. . . .f 7 . .H... .....7 77
. . . fl . .. 7 .p .77. ..., ...7. .. a.. . 7M7. 7.7. 7...; ...... .
. _ 7 . . . 7.. ......7. .. . . W . 7 .w 7 7 7. 7 ..... ..7 9777*. 7.. 7.. 77.7 .
. .. . W 7 7 . 7 . 4 7 .7 . .. .77‘7 7 7.7.7 . .7477 . 7.7W779777
. . .ll‘ 77.97 l 7 l. 77 '7 ...l7 7 7777777774 .lolllo. 7777.-- 777| 777....77ll7 . I ll? ll I£Il7 .. N $ l
. a .H m .7 . .. o 7 7. o 7 . . . . u .. < 7 . f 77 7777...77 7 ...V.*.. 7A
. .. . 7 _ 7 .. . ~ . A . . . .. .. . e. 7 7... _ 7 . 7.. 7. .w... 7.... 77 7
7. . . a . .. a . . . 7 . 7 m. 7 . 7.4 o 7 .... . 7 ... ... 77.... . 7
7 7 .. . . 7 . . 7 . 5 7 o 7 . . . . 7. . .7 7t. 7.? 70 4 I
.7 L 7 _ 7 . 7 t t H . e a 7H7.
J7 71 4 7 .. 1
7 7. u 7 p .. H o .0 7 7* ... V b. 7‘7 .77. 7 7H .7 .70‘. . 79 l.w~..0 7
7 7.* 7 .7 77 7 77. . 7 v 76. q . ... . o7 . 7 7.7 7.4. 7 .Y..7. . v.7 L 7.777.. 7.. 7. .... 7
o .. . .. . 7.H7 . 7 ¢ .7 7.74 . 7 . 7. 7 ..7. 7. 7 747 777477 777 7 7 4 o 7*.77 7 ......7
. 7. . 7 fi .. ..... . . 7 .. . . 7 . . ...7 .. .. .r. . 77...».7 7.4.. . 7 . 7A 77P... . ..H...
I7IOL7I|DIIQI*7.lOQl‘77A . |9ll0i.TthIlJIliilJT-tln 7 d '7'... 1 . r797 l a l
. 7 .. k . 4 . .* . ... . .7. . 7+. ...7 ... .. 4. 7. .7 ¢. _. 7
.. . . . o a .7. .p .7 7 . . .. 7. a. . ... 7 .7 ... 7. 7 ... .7 777 7.... .
. .fl .7 . . .. i “ . . 7 A. .. .a . 7 74 . .~ . 7 .7 7. ....I.. 77 ...
7 . . . .. . m7 . . . .‘7 ... 77 .. .r... . o . ... 7 .. 7
s . F l .
l 4 l4 4 . l
7 fl _ .. .. ... . h 7 . . r ¢ ....7 7 7 7‘ 7 o 7 o .7 ..c...
v 7 . ».7 .7. .. ..— 7 7 . 7 c e. . 77 . . W .7 7 7 .. I 7 + .Vl ...k.7.. A.
. 7 . ... .i. 7 . . . 7 . .. A . ... . .. 7 .. 7 . . ...777. ...a... ..7‘77. 7......
7 7 . 7 . o .. h 7 . .n 7 47 . 7 7.. . . 7 o . . . 7 . 7 .7 77 >7... 7 7 .7 7. ..77. . ..>77..
f 7 7707777 . 777 fl .-ol 7771.777 7 ll7liT77- 77777777! 477 7 77 771777 4|-.7I7 l liw7ll7L lolfllfilllll L l. 7777477 . llTl... 70.777177777le l .7
7 7 .7 . . . . 74 . .... . .7... 7 7 . . . . . . .7 . . 7 7 .7 .... ..Ml... 7 7... .
. _ p . u . . ... .7 .. . . . . .. . . .. 7.. ..77.7 7 7. ...7 .7. . 777......777. 7777*.
. v 7 .. H . M M . 7 M 7. . 7 .i . 7 . 7 . 7 .7 H ...7 .7..“7..77 .77Mzo7l7 77..V775LV7 ufi'OI
7 F . .u 7 . 7 H 7. w l . H H. .7. . .7H.7 . 7.7777777. 57 1977. Well it
4 i 4 l l 7 1 l .
. . .. . . _. u 7 .. . 7 .. .7 . . 7. 7. .7. u. .7“ 7 «.... .7 7 . c 7. 7707 777
7 a . 7 7 . § H . o 7 . 7 7 . .77. . .. . . . .7 .. .. p . . . 7 ..7. . 77L7a7. .77 7&4 .7 .I.
. H ‘ . . . _ 77 . . _ ... . . 7 . . 7 ...b...7. 7 .. .7.. 7 7 T .7 .7777. .fi7‘ 7
. _ 7 . o 7 7 7 . . .. .. . . 7 7 . . 77.7.77 7 . ..7 . 7 77 7 .. 7 ..7 Qic 7.7 777.7;
fi70ll0 7 .47 77777 7 III 77.0.7777.7H.7|7lv.’.| l7im ul .7 .77 77 0.7; l70.OI777clc.Hllll7l7‘l77lll7lL107$l5l. 1l7.777l§.l77v O77|7T0lrviolt+il7inun . 77'] Out 11l 7|; *l .4 l7 Hlfl
. . . . h . , fl .. _4 h. ... i 7 . . ,. .. .... ... 7.... ...H .. . 77.;
_ k ._ . . . . 4 . . v. .. . 7... .. .. 7 . .7..~ -77 77 .7...
. . . . . . . u 7 .+ . . n . . .7 . . 7.7 . 7 ...».77 7 ..Y.
. . 7 . . 7 . . .. .... . u 7... .7 . . ...7 . a ...... .7. .77.
vi 77lo|7l7|7l 7 illlzc 7|- 77‘7'1 71!..77 llllalo. l 7 .17? 7|.'7 ’7. $7.7‘l l +7 ll? lo 47 w
. .H .. . 7‘ 7 . 7 . 7 7 7 7 *i. 7 7 77 L .7 77k .7 .. 777. o 7777
. . 7 . . 7 7 i . . 7 7 .7 .. 7 . 7.. 77. 7.. 7 .. . .. 7 7. 7 u a
. 7 7 . . . h . . . * a . . 7 . 7 . 7H 7. . -7“ .7 .7 77 7. 7.77777. .7.H.A 9777
. 7 . 7 . h . . .k . . . 7 . 7 7 7.7.7.7. 7 7. 7 7 770. 7 77 ..7. 77x77770
77 . . o 7.. It; . 7 7 0707 7 I; .l 7 . 7 7L7. 77177.7 7VIL 7770777] 0 . 7‘ 7.. 77701 7 777 0|. 7 7777 nolo *I 77717.7f77l . Vi '77 it» 7I'IIIAv7l 7iiQIOI. 7Jx77.777l& It'AT‘iI‘ux 77077O|ofillo7lllx 77.10,}: AI! .7 vI'IA
. e. . M . . . . . 7 . . . . . .w . . 7 . .7... .u.... ..7.7....I77 £w0.rl7
‘ . 4 7 . 7 . . 7.. .7 7 7 7 _ 7.. .. ....7 a 7. .7777 fi7777 7 fintfl
— . i
a . . . _ . . 7A . v . 7 .7 ..H 7L...Oo .777 .791 7777...
7 . 7 . .7 . . 7.7. 7. 777 .. 77 . .7 7. .77.. 7 7o . 7 7
7. . .7 .l l . a 7 .7 a a _ i H . $7 7. I7.
74 77 l l 7 7 _ 7 4l 7 7 4 ‘7 7
. 7 . . . a . d .7 7 7 7 7 .... . 0 . ..7~.. Q77777 77 77 ¢Y77H777IA
. . .7 . . . 7. .. . . .. . 77. e 7 . . .. ...... .77 . ...?b777 F7777. 77777
a . . ¢ . .7 . 7 .H o: - ‘7. 4 v 7 7. 7 7.7.1} 7". O, to . LQvio.
. . . 7 _ o 7 . .. . 77 7.. .le .7. 7 . 777. 77. .7.77| 17 .7. m117'7977
7-777.777 7. .7. 7-7 7. . L7777l+ ..7 l.. + 7. F.7 I77 7 i A7777. +l7747777r777l7717677777h77 .777 77
7 _ . _ . . H7 .7. . 7 w. 7 .. 4 7 . 7w. . ..7 ......9 77.1
. ... 7 v 7 .... c . .7. .7 7. 7 ... l 7! ...7
7 _ . 7 . . 7 . .7 ...77 . 7 7 7 7 747. . T 0.
7 M . . W 7 .. . 7W . H. 7 .77 7 ..77 77p.
Yill‘llllil lol 717‘ 7777 I. l 7? l7 Ldr .
_ 7T 4. . .7 H 7 .7 . .7. 7 .. 7. 7
_ 7 . 7 7 .. 7 .. . 7 . _ .. ...7. 7 ...7.
. W .7 . .97 .. 7 7m. _ 7 7 .7 ... . 7
0 . fi . . fl 7 .. 7. .. . L.. m 7 7. 7 o L. a . 7 7.? 7 7 7797..
l.-. . 777 7 7 7 . 7 7 . I 77 O . 7.47 777's... 777i 1 ill... 97. I 77 7‘ 7|! iIlIArI7-I71 7- 11 7+ ‘77 7 70 All 77"? 770700 l
. v n 7 . . .. . . . . . 7 7 4!!! 7 .*7 .. 77a 7 77 . 7771.177
7 . . . . 7... . .. .7.” .. .4 . ...+ . .... ... ., . 7 7 777 . 7 707..
H _ . i 7 .. . w 7 7 . c. . .77... . .4 .7 7 . o 7.7 .75. O7?
. 7 . i 7 7 7 .7 c 7 c ... ... ...». .9 7 .. _ 7 . 7 777
. 1'7! I” I. l “III l! l 71 P. a l 1 1 4 1 l“ 7% ‘ 7i 4 P
. . — 7 7 . . 7 v . 7 . .e .. .7. 7.. 7 . +.77 .. o 7 7 7. .. . 9.7.
. _ 7 o . 7 ~. .. . v. a .7.“ n .w.. ......77 . 7 v 7 7 | 47. 97 77 7 «... .777 .0
. 7 . ... w . 7 7 ... .7 a. 7 . ... 7...... . .o .. . ... 7 ...7 ...7 7--
. . . _ 7 7 U .7 H..7 . .7. 7 7.. . 7.7.77 .7”: . 7....77.
. 7......73 as. a. ewe..\<§7w$ezww7m§i<§ 7 77 7777777 .
. . . .. ..... . . 7.... .. . .. ... . ... 7.. ..
_ 7 a . A . . .. . 7. -. 7.7.7: .. . -.. ....
. . . a . o 7 fl . .7 7. .....7 . .7... 7 . 77 .7 . 7.77 . . 77 o 7. .77 7
g 7 .. . ... ...v 7 a.. . .7. 7 ... . .7 .7 . 7': ..J . .V. 77
MLllllillel7 7.7777? . 7 l 4|l77....l77. 7 .7777 l77.7 fililiolci r . .77 H . 7 w 7 777
7 . 7 7 7 7.. 7 ... . . w. 7 7... . . 77
. . fl « a. 77 7 7 7.‘ 7 4 ‘ 7»... . 7 ...”.J
. ~ . .7. 77‘ . 7 7» . . . . 7. .. ..A 777wYt7
. . . v 7.».. . . . 7 . . _f.. 7~.L . 7. .7.
7. .77 . 7 77.0 L .4 7% Lr Ah: 4 7 7“.
7 . . 7» 7 v 7 o 7 ... 777‘ 7 7 7.... I .7 7..
. W . 77.7.. I 77.7».7 7 7 7a. .7. 7.7.7... 7* .7 W 7.7HA7.7
. # . .t. 7 7 . 777. .7 7. 7 7 .7u7. 7 .7.... VI..o77-v7.
. a * . . .Lw. . ”.7. H . . .. .77 .. 7..7777 . .4 7
{it‘ll‘7 l .‘ ‘ " 1 1 1 4 l“ . 7
. . 7 N a . 7.77.. 7 o 7‘47 .Iw 7 . . . . 7.7777 . 9 77 77 1H . 7 do
. _ . 7 . . 7 k. 7 .+... .7.77.. 7 . . .7 7 . . 7 ...... 7. +7 ..
. fl . 7 V . 7 . . m ..a. .. . .o. . . . . . .. fi— ..7. . . ..L ..fil. 7
7 . 7 . . . 7 .7 ..7 . . . 7 7 '7. 7.9 .. . ... 77
.7 7777 o I, ll 77I’ I I ’777 7 [6707.770] a 7 7 xf7‘i7 bltl Yln‘lbl kl HIA‘II (ll! 0 741 l l R a 'ngli'ii 177:7.L 11.7
. ~ . . . . c . .. . .. 7 _. . ... u v a 7 . — 7‘ .7 .. .o ‘.
~ fl 7 . w . . 7 7 7 . . 4 . ....7 .17 .... 7.. fi . . 7 ‘ a.
. . . 7 . . . . . . 7... 7.; 7*... ..7 . . 7 .. . “7 . 7. _ ..
. . . . P . H7 7 . 7 . . 7.. . 7 .. . 7. . .. .7 ..
r) Il- l‘lu filli'O all ll‘ ll7nui§llfilviT 171.777. i l i i W
W . . . 7 . V . 7. fl 4 . . ... . .7 .. . 7 . . . . . .. . .77...
_ . . . . . .. 7.. 7 7 7 .r.. 7 . 7‘ 7. 7.77.. . 7. . . ... 7... . 7
. _ . . . 7 n 4.7. . 7 ... .7 . 7. .. .7 7 . . .7. 7 7 .
_ . . . fi .. . . . . . . . 7 . 7 . .7 . . . ..7 7 fi.7 . . H .
.7 .7 77A 77.7 ..l.‘ -.7%00l7777777 ..77777 717.77‘7177 V7'3077l77 7-07l0'll7i4.liloll70l 7‘O7I+7779ILI710707li7ivvl771|0l0 lou'177bl’4lll7'? 1Lll791'7vtol.llbuta'QIi-Vl‘vlf77wv7‘ui7fi77‘.‘J-7.0||7A
. a . . 7 7 _ . «7 7 .. . . ... . . ... . . 7 7 7 77 h a. . .9. .7;
7 . . .7 . _ H .. ... . o. . . .. . ... .. 7 . 7. 7‘ . 7 7.7 . .. .. ..~ 7 M.. . 7 . 7. ..77. .477..
_ . . .. . . a . . ... _ . . ~ .7. 7 7 .7. . .. 7 . 7. ...7 7......” .......
. 7 .. 7 .7 7.. . .. . 7 . . . 7.777 . .77.. .7. ..7. .. 7 .. . 77.0
7 7.7 . . . . M L. . a .7 H a 7 H n H H H

 

 

 

 

 

l

.( .7 .3 Z. u.7.(2 IUZ- Eml 0N XON
AJU ZuUNFaTu MILAAUDL XMl(l —LQ(K.J ZHUNFW7Q ON. On.” 02

Fig. 2

8

velocity may be expressed in terms of a linear, peripheral,
velocity. Further, since these impulses are transmitted
through the photocell to the oscillograph pen and tape, if
the tape speed is known, a direct expression results relating
the kinetic energy in terms of the wave length of tape pen
oscillation; For a tape speed of 125 mm. per second, the
energy equation is

E = l§.5 in foot-pounds
2
m

where 13.5 = machine and circuit constants,
and m I wave length of ten oscillations
in inches.3

The plot of this expression, as shown in Figure 2 ,
provides a convenient and accurate method of determining
the flywheel energy of a given tape measurement of ten
wave-lengths. The plot used for energy data for the fol—
.lowing tests was made on 2h x 36 inch graph paper, insuring
good accuracy to a hundredth of a foot-pound. The initial
energy state usually varied only within a small ranne, but
since the Curve is rather critical there, it was deemed
more accurate to calculate the correspondinr energy.

Figure 3 shows two sample oscillograph tapes taken as
typical for the tests run. Figure 3a is a test run of a
leaded, free-machining brass with a square notch and .005
inch corner radii. This type of notch imposes a definite
.ObZS-inch test length and-represents the practical upper
limit of energy absorption for the machine. Ten wave-lengths

before rupture measured 0.87 inches, with an energy of 17.80

Typical Oscillograph Tapes

 

a
a.
n
S.
H
O
3
A
3
_I.

-

 

   

       

       

                                       

.41; Txfi _ 4
J _ u . . — L .
-. ,3; P\.\+l .
_m _ _ H _
+1, - ‘4; ~“4. - n
.4 . _ _ _ ,
u — _
4. - t \«i... ...“-
_ 4 _ _
_w‘,‘ ..- m'||v+9tl to
_. m .4
_
..rl

        
         

   

'__ _
I

              

   

  

                        

  

!

figh—

. K

R!

    

O
.4 ---—
o
I
'

Q
..-- .4.- ,-
.
' I
O
V
l

t

‘
*——-.———§-—— o——.—
‘ I
- \
- 4
A
0
q--_.__—‘ _ p-.—..- _—
I

u
I

- --—._ .-
l
O

..
I

      

Q

\

O

1

‘|

L—
V

1'..—

.
.0- -——’- ...—_—
X

 

.

‘._
+—-—
. _

s

‘

C
A
O

--
o
'0‘

. _ . 2 .' ~
0 l
. '4‘ . I
I - g -
W
0 . V ‘

I
’ l
I
J____.'._._.,' __ -- _L
‘
l

v 0
--.—_.. --.._— *.—._—’<

-_:-P-_;. _

            

     
 

_
w _ "
'Fu‘ 0.. . ' I*\\IL.!V I- _I It‘ll II
_ w _ _
.Il _ . mu! {T to col... _

---“; q .71; .t 31.4 _ _ . _. .7

4 P c m . _ . _ _ n
\r:.1 ,-+ .1 \raz-: :£:T:+;

. _ . _ . . . . _ _ n_
_ _ _ _ _ n _ _ _ .
Q3 3 .. 4 o I .i? T . «. .f.l..4. 4
_ n," r! I‘f I L 'D .D- I* 0

Fig. 3b

foot-pounds; ten wave-lengths after rupture measured 2.06
'inches, with an energy of 3.10 foot-pounds. The resultant
energy of rupture then, is lh.él foot-pounds. The torque
pip is approximately 5.7 divisions high or representing
285 inch-pounds of torque. The tape length during rupture
is 0.08 inches long corresponding to an elapsed time of
0.39 seconds.

Figure 3b is a typical test run of a full radius
notch specimen of 2hS-Th aluminum. The enerny of rupture
was 6.h2 foot-pounds with a torque of 230 inch-pounds.

The elapsed time durinn rupture was 0.20 seconds. The
full radius notch was later adopted as a control for the
series of experiments since it seemed best suited from
the standpoint of machineability and tool wear. Over the
range of notches investisated, the full radius notch also

gave the most consistent data for the materials tested.

C. The Torsion-Impact Specimen

~Finure ha is a vied of whole and fractured specimens,
adopted as standard for this torsion-impact test. b‘iggure
Ab shows a detailed drawing of the specimen. The press
dimensions are .500 inch diameter and .8125 inches long
with a milled clamping flat .059 inches deep. The notch
is turned in the center of the length to a nominal root
diameter of .3125 inches, and a constart width of .0625

inches.

Torsion-Impact Specimens

Fig. ha

11

 

 

 

STANDARD 7755 T SP5 O IME IV
DE'KO/V/IVO TORSION - IMPA O T MA CHINE

.031 no. .500 0,, r

 

 

 

i!
4]
/
/fl

 

 

 

 

 

.44/ FLAT

 

 

 

 

 

 

3/25 ROOT
DIA.

 

. 8/3

 

 

 

.031 x 45' any.
90m (”as

DOUBLE SIZE VIEW OF FULL RADIUS
NOTOH, ADOPTED AS CONTROL

 

 

 

“ll-ll
[HI-ll

ACTUAL SIZE

 

 

Fig. uh

12

13

Te intelligently run a series of tests of this nature,
a few hundred Specimens are needed. To maintain hiph stan-
dards of dimensional accuracy, a machine-tool setup is
needed with an adequate method of checking these dimensions.
Figure 5 shows the small Hardinpe production lathe selected
for the job. The Specimens were turned from 1/2 inch 0.D.
bar steCk, with a collet type chuck for good cencentricity
between 0.D. and root diameter. Teol blocks were designed
and built specifically for this problem and were keyed and
bolted rigidly to the movable cross—slide of the lathe.

The machine cycle is as follows:

1. Set depth of grooving tool, mounted on rear teel
block, With aid of long-travel dial indicator and adjustable
step

2. Open collet and brinp stock out to adjustable
length-step and lock collet

3. Back in grooving tool to depth set on step

h. Run in cut-eff teel, chamfer with a file, and cut
off.

After the stops were set properly, the entire cycle
took only a few seconds. A vertical millirg maching was
employed to machine the clamping flat and the specimen was
ready to test. .

Accurate dimensional control of the cutting tools,
notches, and finished reet diameters is a necessity if a
quantitative test is to be run. An optical comparctor or

"shadow-graph" Jas used for checking the profile of the

Production Machine-Tool Setup

Fig. 5

 

cutting tools and notches. With the aid of an over-sized
micrometer screw, a table travel was mace for checking root
diameters easily to a ten-thousandth of an inch. The com-
parator was calibrated accurately to give a magnification
of 50 to l of objects placed in front of the lenses. Thus,
with the use of drawings of a given notch on transparent
paper, a close check was run on the cuttinn tools and

notches. Consistent with the tools available, a blanket

tolerance of .001 inch was imposed on the test specimen.

III. . EXijilIxZB‘ITAL LZA‘I‘ERIAL

 

 

 

 

A. ‘ 1iariation of _Eperfiy 2f T‘urtzr '-.'7itjn.
Diameter 9f Test Se tion

I

”'—

With the problems of mass production of accurate test
specimens and the proper calibration of the torsion-impact
machine and resulting data satisfactorily overcome, thought
was then turned to a test problem. At that time, it was of
interest to determine what machining tolerance should be
impose on the specimen, and how this tolerance would affect
the resultinp measurements of energy and torque of rupture.
The initial tests were run partly to answer this question and
partly to systematize the total operation, including the
test run.

The first material investigated was a leaded, free

machininp brass chosen for its excellent machineabilitv,

16

high ductility and moderate strength level which seemed
to indicate a fair amount of energy absorption in ruptur-
ing. The physical preperties of this material were taken
as follows:

chemical composition — cepper, 61.5 percent;

zinc, 35.5 percent; lead, 3.0 percent;

mechanical properties - tensile strength,

58,000 psi; percent elongation, 25 percent;

elastic modulus, 1h X 106 psi; hardness,

Rockwell "B" 78.

The recommended top rake of cutting tools for machining
this type of brass is zero or negative rake, which simpli-
fied grinding the compound angle cutting tools.

Approximately thirty specimens were machined at random
with respect to root diameter, both over and under the nom-
inal root diameter of .3125 inches. An eighty degree tool-
bit was used with a width of .0625 inches and a nose radius
of .005 inches. The root diameters ranged from .30h7 to
.3181 inches and the resultant energy and torque measurements
are shown in Figure 6.

Although the scattering of data was rather large, a
definite trend was indicated. As the root diameter increased
over a small range, the cerr Spendinp energy and torque in-
creased, but apparently not in a straight-line manner. The
scattering of values was lessened in later experiments by
stiffening the cutting tool to reduce chatter and improving

the actual test conditions. The imposed tolerance of + 001

17

$53 .. 23 one RE .6 Eu mews u see a g a

 

093

00“

SON/70d - HOIW - 38/71 dflb’ JO JDOHOJ

 

fiwsh. 0‘xn. Qfikfi Obfim.
...-:in i . ..

a
.

O o o . ‘ o L l U . l
i

.l

t

c
A._. —-

v

u

.

4
. - .
, . . .
- — *ve—o—q
- . .
o
1
4- _.
0
o
...-v
c
,
o

 

'I III;

 

‘s—ov..q__.- _14.
n
Y

r
v
o
-
I
o
l
o
n u
o
o
--- -a
a
- »
0 u
v
u
v
4
v n
o
. n u
. ......h ”4 ..
. l
A
. .—.._—.. _*

W
i
9
I
I
i
;+.
9.
—<—<o—-—‘.—+—o—o——O—~——J
i
U
i
..
I
'
1
TV
|
4
i
”1-9
7?'
. .§.
1
i
I
i
.T-

 

0)?

 

 

 

 

 

e

a J
. .5
.
.
4-9.0?- ”#HL—w
O
O n
.
v
. .
—-—*——v 45
. a A
‘ o . -
l

 

   

 

gs:
as,
in,
“if
1
' i

t

     

 

i
f
”.---in
i
._
1
l
i
E.
. ‘.
if“
1 i i.
1--
t:
. i.

 

 

0'9

 

. H .

H o o a o u ~ 0 A
_

a - J 4 o - 1.....- A . 41

_ , M _. _

. t c g V I o< u I 4 e r 4 V o d [it I o

k v 4 4 . L: A

.

c f .

 

 

 

  

—- ——— o’< —'——1-

i

 

Na '1

 

 

 

 

 

 

 

7+ . . 1
_
+.. . ...
v ... .
l.--lAIrY.tlL
.

. . u I A

 

 

H... -4-
o

 

0'3

 

 

 

.4_i__-4..4 .-

o
A
b

- -..... A.“ _

 

 

 

 

. . .
- ‘ .v v . . . - . . .
. » . . t . . .
- i
‘.-——‘—o— '- , . ‘u .
l
. . 1» . v . . .
. . . I . o . ¢ .
v u - - 1 . . ‘ .
l
. 4L__.-._, ”.-.—4-. ... . .
. . . i , , .
l
0 iv o T . 0 c : - . -~
. . . -- . f 1 .
t .- . . § . '
L -

 

 

6. A o o o 0 n 4 a o A . V A v o __ - . A .
._ V
. _ o - 4, . u . n v . a e c c I v 4. L a o . a . .
_ .
i _ _ .
T'- 9 1 1.0.. 0 9 IOIIILYII’JI‘. ' .."V ‘li'l‘v'l t'-’ Ch. [1’1 It‘ll 1’10" T

 

‘

v
t.
T.
O
l
v
0

u
5
v
o

F
v
.
.—-~.._.
o
o

<
e
.

 

 

o
o
o
9
I

.....

t —
g o
o
- O
o

v

o
4444—- .4».

.

.

.

.

.

 

 

 

 

 

 

 

 

 

i n . o n! o n u v o. r o . p g H 9 0 . Q . o
. , i . _ . m .-
r . u v. . I . . . 1. _
. . . o ' Y
4 u o * o a c u w . v i .
7 U L Y N u A
:6 l 4 o . c o a h H c . .
II»! D) I. Ilcl| -II irollrlll I.L’~ill'

 

 

 

 

 

inch on the root diameter appeared to have a negligible
effect on energy and torque, based on the results of the
test run.

The results of the test shown in Figure 6 posed the
question of how the energy of rupture in torsion-impact
would vary with the root diameter of he est section.
The effect of impact on such a determination is one which
cannot be easily predicted, thus restricting the logic to
a static condition. A ductile material, such as leaded
brass, would exhibit a torque-twist curve that could be

idealized as follows:

AVERAGE TONGUE, T

TORQUE

 

 

 

 

I [HOLE or rwsr, 9-

an exact expression for the energy of rupture would be
the area enclosed under the torque-twist curve, but could
be approximated by

energy, E 3 TGR. ...........(l)
Assuminp an elemental area in the sectional view of the
notch root, and a stress distribution across the radius as
shown below, an equation may be written for the average

torque on the section of diameter ”d".

1?

 

 

 

 

  

 

 

 

 

 

 

 

 

as
v:
V
’
I F areas
013 "tour/on
Q:
o:
- t -
Lanna: sun mm:
was: star/om rTs mess

For static torsion, the shearing force on the elemental
area is shearing stress times area, and the torque exer-

ted is the force times the "lever arm".

1.8., F 3 ASS and T = Fr
where F = shearing force

A = area in shear
and SS = shearing stress.
Thus, dT = (33 dA) (P).
Since, ds = rd9 ,

dA-rdrde ‘
and ‘ . urn
T = SS r2 dr d9 .
00'
Performing the integration and substituting d/2 for R

yields T = IEEs.§3. . ...........(2)'
12
It now remains to obtain an expression for angle of
twist in terms of test diameter. For a rather ductile

material, failure should occur after a certain critical

2O

angular deformation has taken place, as Opposed to a crit-
ical torque since the torque increases only slightly after

the elastic limit of the material is reached.

 

U)
II

or 9

...—Q
n
are m
w
I
I
w

 

 

 

If s is the arc length to failure, the product of GR may be
a constant for a ductile material. The results of the
following experiments were rather inconclusive on this point,
but indicate that such an equivalence may be possible. Fur-
ther instrumentation is needed to clarify the situation.
Substituting enuations (3) and (2) in equation (1) and sim-

plifyinp yields,

Cd2 ...........(b.)'

where C = TTSs k
——.tT-.—

E

To substantiate the diameter-energy variation, an
expression was written
3:06P.
Taking the logarithm of both sides of the equation,
log E = log C + n log d.
This expression corresponds to the general equation of a

straight line where n is the slope and log C is the energy

21

intercept. A test run was then devised for two relatively
ductile materials using a full radius notch and a range of
root diameters from .200 to .320 inches as the independent
variable. The first material investipated was leaded brass,
previously described, and the second was l7S-TL aluminum
with physical properties as follows:

chemical composition - copper h.0 percent;

‘

manganese, 0.5 percent; magnesium, 0.5 percent;
aluminum, remainder.

mechanical properties - tensile strength,
55,000 psi; percent elongation, 16 percent;
elastic modulus, 10.5 x 106 psi; hardness,
Rockwell "B" 68.

The results of this test run were as shown in Figure
7. The slope of the log-log plot was approximately two and
one-half for the brass and two and three-fourths for the
l7S-Th aluminum. This constitutes a rather larpe divergence
from the previously calculated slope of two and indicates
that energy—diameter variations are not quite as simple as
reasoned.

There are several contributing factors. The equation
as derived was based on a two-dimensional system, where
actually a three-dimensional system is more suitable since
the stress at the extreme root of the notch differs from
the stress in a section immediately adjacent due to dis-
continuities imposed by the notch. Rupture probably begins

at the surface of the test section and proceeds rapidly

22

LOG -LOO PLOT

VS. ROOT DIAMETER

FOR TWO OUOTILE MATERIALS.

ENERGY OF RURTURE

TORSION - IMPA OT, FULL RAD/US NOTOH.

 

ease - BE _ Lessee- me- teas..- ,-

-3: 3.--‘ {raw

 

'lth'- . ‘ny‘vlv

 

 

-+:spiéix1311li111:11:lusza

a

_ . .f
_ . a
. a .. W

,E a w . g
271111-151: iii?! it 1| , 2.
. _

-
7.—.._..__.~_.—4.-_~‘ .<—_.- . -,.

ROOT DIAMETER - INCHES

Fig.

23

inward. Since the diameter of the root is a variable in
this experiment the surface speed, and hence, rupture

6

speed is also varying. William H. Hoppmann presented
data for hard cepper in tension—impact showinr a sizeable
variation of rupture energy and elonration with impact
velocity. The results of this work tended to substantiate
the theoretical calculation of a ”critical velocity" as
derived by Theodore von Karman23. At impact velocities
greater and less than the critical, there was a pronounced
drop in energy absorption from impact. If the logic based
on tension-impact is extrapolated to torsion, where a fifty
percent change in surface speed is encountered throuph the
range of root diameters investigated, approximately twenty
percent of variation in rupture enerpy could be accounted
for. The elimination of this variable would be possible if
an adjusted initial test speed of the flywheel were used.
Another source of variation in data is the effect of chang-

1

ing reometric stress concentration imposed by the ranre of
root diameters. Stress concentration factors for static

and fatigue torsion are available.8’ 13 The stress concen-
tration factor may be defined as the ratio of actual stress
in a section to the nominal stress as calculated by simple
theory. If this variable is accepted as applying to impact-

torsion, a decrease in root diameter would be accompanied

by an increase in stress concentration, and hence, a decrease

in energy of rupture. Taken literally, the energies obtained

at the lower root diameters would be lower than nominal with

a resulting increase in slope of the lop-low plot. A
factor of indeterminable magnitude is impact. M. Ithi-

18 reported dynamic torques and twist angles consid-

hara
erably higher than static values, and in general, the
maximum torque increased as the velocity of deformation
increased for materials investigated in torsion-impact.

An attempt was made at this point to evaluate the
shearing stress as it appears in equation (h). Assuming
a point on the log—log plot for brass, the angle of twist
was approximated from equation (1). The subsequent calcu-
lation of stress yielded a value of h2,000 psi. The
shearing strength of this material in static torsion is
about 30,000 psi showing a forty percent increase. If
the slope of the plot was two, as calculated, the experi-
mental value for stress in impact would more closely
approach the static value.

A more sensitive instrumentation of torque values

would perhaps shed some lirht on the variation of twist

with root diameter. The product 0d was assumed to be

2
constant. If the variation was exponential, i.e.,

st

...—J.

9 (d/2)n = a constant, careful determinations of tw
and diameter should yield a more exact expression with
the use of a log-log plot and augmenting equation (h)

to fit the results.

D.)
W

B. Variation of Energy 0f Rupture With

Notch Ancle For Two Ductile Materials

 

If the torsion-impact machine is to be accepted as
a materials testing machine, some effort should be made to
standardize a notch and determine the peneral sensitivity
that could be expected for a range of notches. In the ini—
tial tests, it became evident that size would impose
definite limitations on a notch variable, since the notch
is only .0625 inches wide. For this reason, the notch anple
was chosen as an independent variable in the subsequent
tests, with notch width, nose radius, and root diameter re-
maining constant. The full radius notch was well established
as a control, due to its rather high energy absorption
level in the materials investigated and low effect of tool
wear.

A range of notches from 600 to 1600 was investipated
for two ductile materials, as shown in Figure 8. The tool
bits were pround and honed in 200 increments and carefully
checked on the comparator for size and shape. 0f the two
materials, the brass has been previously described and the
properties of the 2LS-Th aluminum were taken as follows:

chemical composition - copper, h.5 percent;
manganese, 0.6 percent; magnesium, 1.5 percent;
aluminum, remainder.

physical properties - tensile strength, 65,000 psi;
percent elongation, 10 percent; elastic modulus,

10.5 x 106 psi; hardness, Rockwell "B" 75.

26

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o
I
'

 

 

 

 

 

'm‘i ' 1 pay

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

VI ll. +AIT;O-I' .. iii
_ .
c ‘v w T ‘Y o t 0 . Av . tn o v; w o . . u t“ ‘ n.“ I; .
, . . l
a . 0| vir v + n o I on . :o A o o . ¢ . . o u I l s a c u
0 1‘ ¢ 19 u no! *‘6 . 1 o t O ¥f . a o r o 4 . o .. 7|“.
_ . .
cl . . It 0 1’ r all 1 _ t i. . L I < w c t v . I» +2 [30 n. K A
l W Li . b L! '1
4 14] f i < 4 J W
_ _ , T I _ _
'.A1|w-||§u.uv.9 .- a YIP‘ i9 '0 L :I V A. o. I ¢ 0 V 4+ «l? but Tub O *1
. . _ , _ J. , .
v ‘A?1 IAVI 4. O 0'15 ~ ’ 0 A 5.0 I. O I. 0‘ . I . «I t L ‘5 . . r +0.] v 5 y
. , . . _ _
aI t» . T c 9 . A o a I, ’L o 1. OI ... .. . C V! v. . ITI e p ad L
. . 4
a . , _
l. .‘O V. . 'l . t ..l. o a o .. t . v 1*‘_V
i Y‘W‘ I! r b P

 

 

 

 

 

 

 

 

 

 

 

 

11‘4“
v 0‘ H v . Y vi a '6 n Y‘»‘ . p A
. _
I19: I. e . T .9 lo q . o v . o . .‘ . c? o! A
. _ _
lolliie.. «tnsut’AT‘u cl. o'¢ ,_ . . V c i w o‘1o|v A
, . r _
o .1)- .e‘ , .1 Ut , . . c o, L
O 4 ' ‘ I O ' D 4 . U i c Q 9 “‘ I4
r r p r a
l , 1 < 1 +
. - .
IO .0 v leuusltllvt -4ulvl‘lhv I . . a I. I. a . ;u v . 9 . l. »
u . ‘
ol‘ Q ‘Yioio 9'6 4' I n a '0 o 4 .>1 1 v s v
, . , , . _ e
A .v . «.Y . «JV o O :4 w e 0 o . l . . v o I. O . T U? 0 q
. w P 9 I. . u. . f _ L aw it .

mm

.
‘4. 0 .1’4
.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

« . . . . 4
.. . ,Tv.l . "+1.-.. v .. , .-.WL
_ . . i V _. . I . . i . _ . A.
,mfiealoé “ST;
+ . . 11 Q 5.!» 141
.. i. . ._|.. -.i . . . . J» 91?.
kg. Excellence- .. {Tl
. +-.. 41. hr . .fl+1.+ 9 i 1 . A . 9 .p F}? “x. -.
. ?-ola . .lTo fi'aw . A l . i *1. o. . F. . v . a . L o
i I _ v . - .fi --Iiilit 11.1:
. fut .-. . . ..H . t . . . . r. . . ,. “LL . ..
e 411. .r. . . * .I. .4 k. . -4 . . . v o H 49"; e T o
. ,7. fi-. olo .wn . . c. . v w. . . o “.1, q *A + ”if!
. . _ I i . v T i— . . . . i 3 v . . . . ..-l - .Ylow o. L»

 

 

 

 

 

 

 

 

 

I):-

[at

Juana—

Fig. 8

27

Eight specimens were prepared for each of the six points
in Figure 8, or Q8 specimens per material. In addition

to the notches shown, a square notch with .005 inch corner
radii was also investigated for both brass and ZLS-Th
aluminum and gave enerries of rupture of lh.30 and 8.78.
foot-pounds respectively. This represents a substantial
increase in enerry absorption over the vee-notches tested,

but is lorical since there was a definite test lenrth.

0

’.J

o A- ‘ g H n ‘5 L1,“ - ‘ 3 a 'v‘.‘
in torsion aliects MAC iiwal anile

t”

The lenpth of mater
twist and hence the area under the torque-twist curve
or ens-gy‘of‘rnl‘ture.

In general, as the notch anrle increased, the energy
of rupture increased, although apparently not in linear
fashion. The curve for the two ductile materials investi-
gated seemed to have a characteristic shape Vith a ninimum

energy level at a notch angle of approx‘uately 000 The

curve for aluminum is shifted slightly to the left with
respect to the brass. To eliminate tool chatter in nachininr,
it was necessary to grind a top rake of 120 on the tools
cutting aluminum. Effectively, this compounded the notch
angle on the tool-bit, and increased the actual notch angle
in the specimen by a few fiegrees. Thus, the points obtained
are nominal and to be truly accurate should be moved slight—
ly to the right.

The energy absorption level for the leaded brass w-s

shown to be higher than the ZhS-Th aluminum. Although the

aluminum nas a 12 percent hifher tensile strength, the
brass is 150 percent more ductile, as :ndioateo oy the
percent elongation. The aluminur appeared to be more
sensitive to a chance in notch anple by virtue of the
larger ranre of enerry values encountered. This notion

is fairly well eseaolished metallursieally, that is, hisher
strength materials are more notch—sensitive than lower
strength, more ductile materials.

The full ralius notch, used to establish a control
level, had energy values comparable to the 1000 notch in
both cases. It would be possible to calculate stress-
concentration factors_based on this control notch for the
ranre investipated. An analogy could then be made to
static torsion energies of rUpture if a testing machine

could be built with extremely low torque ranges, say hC

D
.L

to 50 foot-pounds, with very accurate determinations o
anple of twist. Thus a specimen of the same si7e and
shape as impact could be rupturer slowly and eneraies cal-
culated. The net effect would be an evaluation of the

f‘

factor of impact in torsion, and the relation oi chanrinc

notch anples to a control notch.

C. Variation of Energy of Rupture With

 

Notch Angle For Two Irons

 

\

It was of interest at this point to investigate the

variation of energy absorption in torsion-impact of two

29
brittle materials. Two samples of iron were available
of almost identical chemical composition, but different
physical characteristics by virtue of their graphite
distributions. The first or "normal" iron had carbon and
silicon contents of 2.87 percent and 2.26 percent respec-
tively and had been innoculated in the ladle with calcium—
silicon to assure proper graphite distribution. Normal
graphite distribution in gray iron is a random, unoriented,
distribution of graphite whorls in a matrix of pearlite.
The second or "abnormal" iron had 2.89 percent carbon and
2.26 percent silicon, and showed a patterned, dendritic,
graphite distribution with resulting planes of weakness.

Beth samples of iron were cast into round bars, 1.2
inches in diameter and 20 inches long, suitable for a
transverse bending test. The bars were mounted as simple
beams in a testing machine and center-loaned over an 18
inch span to failure. Rupture occurred in the normal iron
under a load of 3010 pounds and a mcximum center deflection
of O.h08 inches. The abnormal iron failed under a load of
2250 pounds with a maximum center deflection of 0.185 inches.
These results illustrate the generally inferior physical
properties of the abnormal irons.

Bending stresses have a similarity to torsional stress-
es in that both are proportional to the distance from a
neutral axis; thus the maximum stress occurs at the outer
fiber under both types of loads. A series of tests was

devised to demonstrate the difference in strength—deflection

3O

characteristics of the two irons in torsion-impact and
investigate the variation of rupture energy with notch
angle. Five Specimens were machined for each condition
and gave results as shown in Firure 0. Again, as the
notch anple increased, the energy of rupture also in-
creased in both irons, although not over such a large
range as was encountered in the more ductile materials
investigated. The normal iron was superior in enerry ab-

to rive

sorption for all the notches tested and tenaec
more uniform results. The erratic behavior of the abnormal
iron might be attributed to the random orientation of the
planes of weakness, induced by the graphite formation,
with the twistinr axis of torsion. Thus, failure could
occur in a random manne- across the test section and give
varying results, not necessarily representative. A check
of the test section after failure showed an irrerular
fracture for the abnormal iron. All the other materials
investirated, including the rormal iron, showed a strairht

rather clean, fracture across the base of the root diameter.

IV. DISCUSSICW’IdI)(lITCLUSICYb

The DeKoning torsion-impact machine respresents a
new unique instrument in the field of materials testing.
The initial quantitative investigations with this machine
as presented in this study are designed primarily to ex-

pedite Specimen production, testing,.and standardization,

31

—” --‘~ - ..v— i..-._'-.‘_ .... .... .

jut! I!:lul -1411 AI lllliiifil .III Illiqll‘l‘l 1"

~....
_.I;

7
'

 

 

42!. -... 11

r

...—4

 

 

 

 

 

- . . v

v . i i u +.
M
.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-w—ph.‘—.—.*._.
0

 

 

    

 

-.fl-‘ _— .M.. -1.-.“ ~—'.--.- V—‘W’ “_vqr—-—.-.h _- ,— -~—--—— -_ .Tq _. __'_..—sT._—.. ...-.....-
i . . . . - I . - . . i .
,

 

 

 

\’ i‘ C O 1 ' n i ‘ C A v 0
i
.A or . . e . 9 ‘fi. . _ L. . . * . c
. o . . _ ., c . l . v , w r I
.
[‘lll’lllvlli I ,‘ll‘u‘ OI I. lvlLlOllll .l t A | HI. I IV++ I 1 - .6l‘1.+ Y
. , . .
o u a . a . o O: A i v o 0 *r .u 1. v n v p 1 v + . . .
. . .
p o a o v 5 O _ .I u 0 v u _ . r o v
. N , _ . _
. r . . . . 4 * . . . t» . . . ~ . i i
_ i m
— u . c a . . . w o . Q a .
+1 if r .

 

 

 

 

 

 

 

I -.. .....-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. ‘
.
. i
3 t: . . - i. . . - . .
ill-‘14-!“ --ri-.“1l . :9l. .Ill.ur#
o T V bl. fl I. a c all Y 4 . + v u c u .l
. . . ,. v r 7 . . p . t . . . . r i. w . +y . _
m _ . . u
a . . 7 . 9 v 1 c e L I w o . ._\¢ .4 vl
.. to, ushmoziitgz. “ 9.17....
T > 4 LT ll 11 ¥ Lfi {J Ar *1
.. W .. ‘. .i.. .H. M- .i..,..*..|ru++ii.l ... w... H...”
. . ¢ . . . . _. . . . . . . . W , . e . . . . . . . . .. .
_ . . .
E mxmqaoa; . : .
w . 4 h , c I . o * V c a u 4 v v fil ' l 0 Vi“. fit A .. m 6 fl . u . .9 n 5 9"
_ l_ L ,. _ _
T0 tt 0 i 0”. Illnl IIIQII.‘ m .1 Ili‘OOIIi o 0 Al 1 Al w Til 1r fl . 'Lj’tel‘ftfll‘. ol‘l flln‘uloillfil Ji‘lli‘llfli
O o ~ - v a 0 u o . — o n .. v u . a _w t o n t * O A PI. 7 H u I 5.? Lv r Q19 > iiutle V I u- d o . v o o t
i. . - + . v a w c. v . i u 6 A O W . + . I a \‘ D 7 4 1 e _ o v 0 o f 0:. 0 V a v n u . o A O m g b v A A
a m o r . o u u e . . . ¢ 9 V . ¢ . fi . o v o - rlw o c H . Oil. .- T u n u a _ l. 20 o t 4 I : o . s 0 ¢ . c . 9
v 0 I6, 7 u ’ o p A 4? v p . A to v. # O u 9 v. p w. . » Y: a A v . O i c 51'. A I61 o H . e lo a o c . a a v a fill ~ t‘J
. > P . Ii lr

 

 

Fig. 9

and indicate the path of further experimentation in tor-
sion—impact.

The machine—tool setup that was designed and built
specifically for the mass production of test specimens,
has successfully turned out hundreds of accurate pieces.
The optical comparator proved to be a valuable device for
maintaining proper dimenSional limits on cutting tools and
specimens. Both instruments are storel and available for
future work in torsion-impact.

The testing technique is vitally important for con-
sistent quantitative results. A system for analyiing the
data obtained on an oscillopraph tape was devised and the
suitable energy curve drawn. General accuracy of one
hundredth of a foot-pound of energy and five inch-pounds
of torque was maintained in the experiments run.

An equation was derived relatina the variation of rup-
ture energy with diameter of test section and investigated
for two rather ductile materials. The energy of rupture in

torsion-impact varied to the 2.§ power of the diameter for

/

1

a typical leaded brass and to the 2.75 power of the diameter
for a l7S-Th aluminum. The possible effects of variable
stress concentration and rupture speed along with the fac-
tors of impact and three dimensional stress were offered as
explanations for the lack of complete agreement between
calculated and experimental results. Nevertheless, the

order of magnitude was indicated, and perhaps most important,

the diameter of test section was shown to vary exponentially

33

vith rupture energy since the log-log plot was linear for
the two materials tested. Further more delicate instru-
mentation is indicated for this study to ultimately yield

an expression relating the exact variation of energy and

(1'

diameter calculations of members subjected o impact-torsion.

A study of the notch sensitivity of four materials in

torsion-impact was devised and run with the following

results:

1. Energy of rupture increased with an increase in
notch angle for all materials investigated

2. A leaded brass showed a higher level of energy
absorption than a stronger, less ductile, aluminum

3. An iron with normal graphite distribution showed

a hipher level of energy absorption than an abnormal iron

of the same composition

4

h. A 2bS-Th aluminum showed more notch sensitirity,
or spread of energy absorption, than any of the materials
tested

5. The two irons tested showed less notch sensitivity
than either of theatre more ductile materials

6. Generally, the abnormal iron pave more erratic
results, attributed to a patterned gra;hite distribution

7. All specimens tested ruptured at the base of the
root with the exception of the abnormal iron which failed
randomly across the notch

8. A critical notch angle was indicated for the two

ductile materials investigated at approximately 900

O. The full radius notch was established as a con-
trol and is recommended as a standard notch for future
investigations.

An interesting future study with th DeKoninn machine
would be a plot of the variation of rupture energy vith
tempering temperature of alloy steels. Luerssen and Greenell
detected an Optimum tempering temierature for steel with a
definite test length. Perhaps the additional factor of
notch sensitivity would yield valuable information for
notched steels in torsion-impact. A higher energy of rUp—
ture is required and thus a more massive flywheel with the
resulting recalibration of the machine. Other problems of
heat-treatment and notch production would also be encountered.

The DeKoninr Machine with its related instruments and
specimen production machine and tools are available for fu-
ture research on a material's behavior in notched, torsion-

impact.

10.

11.

k.)
\Jl

L IST OF hEr‘LfihEII LIES

Caine, J. B. What is Strength? The l‘Y‘eundry,
July, lets.

Clark, D. 5., and Datwyler, G. Stress Strain dela—
tions Under Tension Impact Lgadinp. Proceedings of
the AoboToi‘io V01. 38, 1938. P. 98.

DeKoning, Paul J. The Design and Construction of a
Torsion Impact Machine. Unpublished M.S. thesis,
Michigan State College, 1950. L6 numbered leaves,
10 figures.

Fredenthal, Alfred M. The Inelastic Behavior of En-
ggneering_Materials and Structures. Wiley and Sons,
Inc., New York. 1950. P. 555.

Greene, 0. V., and Stout, R. D. A Study of the Influence
of Speed on The Torsion-Impact Test. Proceedings of the
AJSITTM. Vol. 39, 1939. P. 1292:—

Hoppmann, Wm. H. The Velocity Aspect of Tension-Impact
Testin . Proceedings of the A.S.T.h. Vol. L7, 19h7.
P. 533.

Jaeobsen, L. S. Torsional Stress Concentration In
Shafts of Circular Cross-Section And Variable Diameter.
Transactions of the A.S.M.E. Vol. H7, 1925. P. 619.

Lipson, C., Noll, G. D., Clock, L. 8. Stress and
Strength of Manufactured Parts. McGraw—Hill Book 00.,
New York, 1950. Pp. 65, 110, 121, 157.

 

 

Luerssen G. V. and Greene 0. V. Carpenter Torsion-
: a , :__ . ..e
Impact Machine. Baldwin Southwark bulletin # 11h.

Luerssen, G. V., and Greene, 0. V., The Torsion Impact
Test. Proceedings of the A.S.T.M. vol. 33, ParthI,
1933. P. 315.

 

Luerssen, G. V., and Greene, 0. V. The Torsion Impact
Properties of Tool Steel. Transactions of the A.S.T.M.,
Vol. 22, 193A. P} 311} .

 

 

12.

13.

it.

15.

16.

17.

19.

20.

21.
22.

23.

2t.

36

Luerssen, G. V., and Greene, 0. V., Interpretation of
Torsion Impact Properties of Carbon Tool Steel. /
Transactions of the A.S.T.h. Vol. 23, 935. P. Sol.

Kanjoine, M. and Nadai, A. High Speed Tension Tests
At Elevated Temperatures. Proceedings of the A.S.T.h.

 

Vol. Lo, l9h0. r. 822.

Mann, H. C. High Velocity Tension Impact Tests.
Proceedings of the A.S.T.m. Vol. 36, 1936. P. 85.

Margerum, C. E. A Test For Shock Strength of Hardened
Steel. Proceedings of the A.S.T.h. Vol. 21, 1921.

 

Marin, Joseph. Mechanical Properties of Materials and
Design. McGraw-Hill Book Co. 19h2. Pp. lgh-202.

 

Marin, Joseph. Engineering_Material. Prentiss-Hall,
Inc., New York, 1952. Pp. 169—212.

 

Mititosi, Ithihara. Impact Torsion Tests. Technology
Reports. Tohoku Imp. Univ., Sendav, Japan, 1933.
Vol. II. Pp. 16-72.

 

Petersen, R. E., Stress Concentration Factors in
Practical Design. Machine Design. Vol. 15, July le3.
P.‘105.

 

 

Peterson, R. E., Design Factors For Stress Concentr:-
tion -- Notches and Grooves in TenSien’And Torsion.
Machine Design. Vol. 23, No. 3, larch 1951. P. 161.

 

 

Sanneur, Albert. The Torsion Test. Proceedings of
the AoSoToiiio V01. 38: 193?. P. 3.

Symposium on Impact Testing. Proceedings of the A.S.T.M.

Vol. 38, 1938.

von Karman, Theodore. On the Propagation of Plastic
Deformation in Solids. National Defense Research
Council Report, No. A-29, 19h3.

Windenburg, D. F. Significance of Impact-Test Data
in Design of Engineering_Structures. Baldwin South-
wark Bulletin #'IOO7.

 

HICHIGRN STRTE UNIV. LIBRQRIES

 

31293010931156