Proceedings American Society of Civil Engineers 1953-12: Vol 79

SEPARATE No.

ENGINEERS
DECEMBER, 1953

OF ROWERS
y S. J. Johnson, _M. ASCE,

wW.G. A. M. ASCE

SOIL MECHANICS [AND FOUNDATIONS

{Discussion until April L 1

1953 by the Society oF CiviL

Printed in the United States of America

Headquarters of the ‘Society
New York 18, N. Y.

ER COPY

a || = iim. | LS
lh
PRICE $0.50 P —
Ti =) —

--represents ar an effort by the Society to deliver
technical data a direct from the author to ue
reader with the greatest possible speed. To this
it has had none e of the usual editing

Readers are invited submit discussion apply-

the front cover.

‘Shans on are planning papers or
for “Proceedings” will expedite Division and
“a Committee action measurably by first studying
“Publication Procedure for Technical Papers”
(Proceedings — _Separate No. 290). For
copies of this Separate—describing con-
tent, and format—address the

‘Reprints tee: this publication : may be made on
condition that the full title of paper, name of
author, page reference, and date of publication

made expressed in its

This paper was published at 1745 S. . State Street, —
Ann Arbor, ect the American n Society « of 7

|
— 8
q —
ln
—

FIELD PENETRATION TESTS FOR
SELECTION OF SHEEPSFOOT ROLLERS

_Johnson® and W. G. Shockley**

excess ap 750 psi was s accelerated during the last war r by the airfield c con-
_ struction program. It has now become common practice to specify these
_ large rollers on earth dam construction. It was considered that the heavy |
roll ers would make possible the attainment of densities in compacted soil =
which could not be obtained with smaller rollers e xerting foot pressures 0 -
(eee
& During the later part of the war, and since that time, tests ee ere “0
performed to determine factors influencing the results obtained by the large —

rollers. A detailed discussion of the large sheepsfoot rollers was given ina >
; paper “Factors Influencing the | Compaction of Soils. Taal As indicated in that
paper and by the results of field compaction tests, corroborated by oo
tions during field usage of the rollers, it became apparent that the use of

_ large, heavily-loaded sheepsfoot rollers involved many factors about which |
— little was known. Two of these important factors were the effect —
of roller foot pressure and the size of the feet. As) pointed out in the above-

"referenced paper, in order for a roller to “ “walk out” of a soil layer the bear-
ing capacity of the compacted soil must exceed the load on the feet. If this is
not the case the feet shear the soil and the roller does not walk out. =

The results” of field usage of compaction tests demonstrated that the

considered that the most efficient means of compacting soil would be praearn
if the large rollers were used with feet having sufficient end area so that the —
soilwasnotoverloaded,
The need for additional information suggested that some simple pilot tests
were necessary to aid in selection of the allowable contact pressure for the —
large rollers and the proper size of feet so as to furnish a basis for planning 7
_ tests for the selection of rollers for practical construction. Since no means |
were available to aid in selection of the size of feet, the writers proposed and >
_ executed simple loading or penetration tests on compacted soil using model
feet covering a range of sizes which would be of practical interest. These
-_ simple tests, which were small loading tests, were performed onthree |
#Chief, Embankment and Foundation Branch, Soils D Division, eames
Assistant Chief, Embankment and Foundation Branch, Soils Division,
_ Waterways Experiment Station, CE, Vicksburg, Miss. 7

***Figures in parentheses refer to bibliography | at end of —_

—

ae! 4
a... size of sheepsfoot rollers in common use has increased very sub- eee
_ stantially during the last decade. The use of large sheepsfoot rollers having a
i
4
—

lected, and the top few inches of loose, dried material were removed. At each

-iaee soils with the hope of obtaining preliminary information which would

indicate whether an ‘might have

used in earth construction. Additional testing is desirable but the usefulness _

of this approach is considered sufficient at this time to review the results ob- | -

ideas the desirable ‘contact pressures for sheepsfoot rollers

7 “a sseten | areas ranging from 6 to 24 sqin. The ‘majority of these tests were
_ conducted on the compacted embankment at Grenada Dam, Miss. andoncom-
‘_ pacted soil in a weather-protected test section at the Waterways Experiment >
Station. A few tests were made at Texarkana Dam, Texas. Soil characteris- 7
tics (classification, moisture content, density, and Proctor needle penetration —
resistance) were determined at each test site.

A photograph of the device used in the field penetration tests is shown “
figure 1. Model feet having end areas of 6, 9, 12, 18, and 24 sq in. were used.
The hydraulic ram of a truck-mounted drill rig was used to apply load to the

7 ; model foot; the apparatus was so adjusted that a constant rate of penetration J
7 would be obtained. The applied load was measured by means of a proving ‘ring cs

as shown. The depth of penetration of the model foot was measured by a _
pointer attached to the upper part of the _ and extending to a reference _
Stake driven into the ground,
“a The areas selected for testing at each sit site had been compacted by sheeps- os
_ foot rollers, _ Areas that appeared to have uniform soil conditions were se-
location a minimum of two penetration tests were made with each size of -
model foot. Locations of individual penetrations were spaced so that succes -

- sive points were outside the disturbed zone of the previous penetration point.

Moisture content and density samples and Proctor needle penetrations were
taken just cutside the area disturbed by the foot penetrations.
ee Seven areas were tested at Grenada Dam. Areas 1, 2, 3, 4, and 7 are

_ Classified as lean clays of low plasticity. Areas 5 and 6 are classified ts
awe sands. _ Mechanical analyses and Atterberg limits data are shown in
figure 2. Averages of field in-place densities and moisture contents and av- —
erage Proctor needle penetration resistance are listed in table 1. - Five sepa-_
rate areas in a compaction test section were tested at the Waterways Experi-—

ment Station. The soil was a lean clay, see figure 2, and had been thoroughly
little variation in material type. Averages of

shown i in table was tested at Texarkana Dam; the

_ mechanical padi and Atterberg limits are shown on figure 2. Density and

_ Discussion of Test Results
- Typical individual load- penetration ’n curves are shown on figures 3 and 4.
| Most of the curves have the same general characteristics; that is, a fairly —
rapid increase in penetration resistance up to a penetration of 1 to 3 in. and _
‘then very little further increase up to the full depth of penetration (7 in.). In

some i instances there v was consistent ins increase re-

bik

44
it
|
|
&g
—
4

4 the drier soils tested. . Certain « of the curves showed a a sharp increase in 1

oad

_ the last few inches of penetration. This effect is believed to be the result

_ of the model foot penetrating into a more resistant underlying layer. | Such ©

_ phenomena | cannot be avoided in normal lift type of construction where layer

thicknesses are irregular and density and moisture content characteristics
of the soil may vary between adjacent layers. =

One important factor that should be considered in the selection of sheeps-_
- - foot rollers is the size of the tamping foot. The bearing capacity of certain
soil types is dependent upon the area of load in contact. It follows, therefore, 7
that the proper operation of sheepsfoot rollers in fill construction may depend —
not only on the load intensity but also upon the foot size. The unit load at -
_ penetration of 2 in. was ‘selected to investigate this effect on the soils tested.
_A 2-in. penetration was considered sufficiently deep to obviate surface irregu-
Soe and, generally, at this depth the penetration resistance had built up to

a substantial part of the final value at maximum penetvation, The load (aver-

against foot area on figure 5. average resistance was about
_ the same for all model foot sizes in areas where the penetration resistance >
was low (1, 2, and 3 at Grenada; 1 and 4 at the Waterways Experiment ell
and 1 at Texarkana). = ‘It should be noted that the water contents of these test
areas were higher for each site than the water contents of the other test areas.
‘There was a fairly consistent tendency for penetration resistance to be greater
for the small foot areas than for the larger foot areas in the other test areas
where penetration resistances were higher (soils drier). This fact may have
_ some bearing on the selection of proper roller suet size, as will be discussed

Effect of Water com

‘It was implied in the preceding paragraph that there was a relationship _
_ between the penetration resistance of the model foot and the moisture content

: “of the soil tested. A study of this feature was made by means of the | plot on

figure 6. The loads at 2-in. penetration for all model foot sizes were averaged
for each test area and plotted against the average water content for that area. |
Average density values for each area are shown beside the plotted points on r

- the figure. A definite tendency for the penetration resistance to decrease -
rapidly with increasing water content will be noted. The curves shown on c

: figure 6 for the lean clay soils at Grenada Dam and at the Waterways Experi-

7 ment Station have the same general shape as similar plots of penetration re-
sistance for the Proctor needle in tests on laboratory-compacted soils. | This -
similarity is not unexpected, as the model-foot tests are comparable to —

Proctor needle tests except for the size of the penetration device. Bayi ®, a

—: The average values of Proctor needle readings for each test area are plotted

against model foot loads at 2-in. penetration on figure 7. There is adefinite
trend for the needle penetration resistance to increase with increasing model -
foot load for each soil type. The lack of close correlation is to be expected, ae

as the Proctor needle is of very small diameter and would be quite
to minor irregularities in soil characteristics, which would tend to be masked
out bythe larger model foot, —_

7
—
7.
Avie
je
oan
2m
—
4
—
4
|
tm:
&g
j

Certain of the should be mentioned before any
to the prediction of sheepsfoot roller behavior is made. One of these is the .
_ rate of penetration of the model feet. A constant rate of penetration was
‘maintained for the full depth of penetration in any one test. However, rates
of penetration varying from 0.004 to 0.04 ft per sec were ie in different
tests. In normal sheepsfoot roller operation, the vertical component of ve-
locity at the extremity of the foot is greatest (about 3 ft per sec for a typical
ie at the instant of contact with the ground and decreases to zero at full |
penetration. There were no discernible differences in penetration resistance _
Of the mode} foot for the ranges tested; nevertheless, these rates of penetra-
a tion are | different from those of an actual roller foot. " Since the rate of pene-—
tration of the actual sheepsfoot is relatively slow at the time the maximum cd
_ load on the foot is developed, it is believed that the difference in rates of ~<a :
: etration of the model foot and the actual sheepsfoot would have little effect on
the penetration resistances. , The actual sheepsfoot pressures probably are
not significantly different from those developed by the model foot. —_— 7
Another factor which should be considered is the pressure exerted by the
roller feet.(3) ‘The usual method of computing the foot pressure on a sheeps-
7 foot roller is to assume the entire weight of the loaded roller as acting on one ©
of feet. it is well known that the of roller feet in contact

is pvtenctloe with additional passes of the roller, the number of feet in contact _
with the soil at the same time decreases if the roller tends to walk out.* The _
pressure intensity on the individual feet should then increase. It is obvious _

that the actual pressure distribution ont the roller feet could vary w widely from 7
fe A few tests have been run at the Waterways Experiment Station replacing
a sheepsfoot roller by a specially conmeee ted foot which measured actual ©
_ pressures during ccmpaction operations. The device consisted of a steel — =
: cylinder fastened to the roller drum and neving a foot attached to the outer —
end; the size of the various components corresponded closely to similar parts _
3 the actual roller feet. Baldwin-Southwark SR-4 strain gages were attached —
7 | to the inner surface of the cylinder and electrical leads were carried to a 8
Brush recording oscillograph. Thus it was possible to obtain a continuous
_ record of pressures exerted by the foot during rolling operations. A device 7
_ was attached to the roller drum to indicate the orientation of the foot with —
_ respect to the recorded pressures. Two passes of the roller were made over -
a test section in which lean clay soils had been placed at different water con-
tents. The roller had a foot area of 14 sq in. and exerted a nominal pressure
A plot of measured roller foot pressure versus average moisture content
of the soil is shown on figure 8. The test foot in the first run had a slightly |
& greater area (16 sq in.) than the actual roller feet (14 sq in). However,no
significant differences in pressure between the two runs were noted which
could be attributed to tod differences in test foot size. The e data aaah somewhat _
‘It is generally accepted that in proper operation of a sheepsfoot roller the
‘Fae of the feet into the soil will be less with succeeding passes of — _
7 the equipment. |The roller feet usually will ponstwate 1 to 2 in. into the soil
on the last pass. This is known as “walking out.” ee er —

.
i
| ty
A
i
|
4
x
|
i

: scattered but they show a definite trend for less pressure to be exerted by oe

_ ‘the roller as the moisture content of the soil increases. . The wide variation
_ in pressures at the lower moisture content may be attributed to uneven dis-
- _ tribution of moisture in the soil with resulting hard and soft spots. It will be
_ noted that at the two higher moisture contents the actual roller pressure =e
_ considerably less than the nominal computed pressure. Furthermore, the 7
_ roller walked out on the 20 per cent moisture content soil and yet was not oo =
exerting the full nominal computed pressure on the feet. _ These data are very —

4 limited but should be borne in mind in the succeeding discussions on roller
behavior.

: Application of of Test Results

between the load-penetration curves of the model feet and the nominal foot —

pressure on the sheepsfoot rollers as related to general roller behavior ~~
_ ing compaction. The individual load-penetration curves for any one test area
q were grouped together regardless of model foot size, in order to simplify — ae
analysis. This was considered permissible for the areas of wetter soils, since
little difference in penetration resistance was noted with the different model

foot sizes (see figure 5). Curves for the areas with drier soils were also ze

combined, even though foot size had an effect on penetration resistances. This |

feature is discussed separately i in the succeeding paragraphs. 7 Furthermore, -
; the load-penetration curves for some of the areas were grouped together based

on the similarity of water contents, penetration resistance values, and roller _
behavior. Pairs of load penetration curves which envelop all the individual ‘

- curves for the groups of test areas are shown on figure 9 for the Were

Experiment Station data, on figures 10 and 11 for the Grenada Dam data, od
on figure 12 for the Texarkana

In the Waterways Experiment Station compaction studies, the

roller used had a foot area of 21 sq in. and exerted a nominal foot pressure of Z
- 250 lb per sq in. Soils were compacted in 6-in. lifts with six passes of the _

roller. Field observations showed that the roller did not walk out i iné areas 1

and 4 but did walk out in areas 2, 3, and 5. x ‘Yhe nominal foot pressure cf the
roller is plotted with the load- -penetration curves on figure 9.

It is immediately apparent from inspection of figure 9 that the nominal |

ee pressure exceeded the penetration resistance of the soil in areas 1 and
and did | not exceed the values: for | areas 2, 3, and 5. the load- ~penetration

: then it may be deduced that areas 1 and 4 could not support a roller exerting
7 pressures greater than about 150 psi and the roller probably would not walk yp
out unless the pressures were somewhat less, say inthe order of 100 psi. __
— This deduction is consistent with the measurements of actual roller pressures :
0 on the same ‘soil, ‘shown in figure 8, which indicates that the roller was exert -
a ing a pressure of 100 to 170 psi in the range of water contents of areas 1 and 7
= 4. Similarly, it may be reasoned that the soil in areas 2, 3, and 5 is capable ©
_ of supporting greater nominal roller pressures, in the range of 400 to 700 psi,
7 still permitting the roller to walk out during compaction, and that roller pres- :
7 sures greater than 600 psi for area 2 and 800 psi for areas 3 and 5 might be
_ too great and prevent the roller from operating satisfactorily. The foregoing —
values should be modified, however, to account for the size of roller foot, __
since, as shown on figure y this factor influences the penetration resistance
ink in areas 2, 3, and 5 because of their lo lower water contents. For example, an

—
as!
g
| —
he test data ained in th nvectivation were studied | aricon
il
ais
.
—
a
>
4:
oa
|
il:
Lf
—

a roller pressure of only 500 psi would be a 14- -in.
foot in areas 3 and 5. This effect is less pronounced as the water content in-
creases; in area 2 for the two foot sizes listed above, a change in roller sone
of only 40 psi is indicated.
‘The sheepsfoot roller used at Grenada Dam had a tamping foot area of _
6.25 sq in. and a computed nominal pressure on the feet of 685 lb per sqin. |
The embankment was being compacted in 8-in. lifts with 12 passes of the rol- a a

Field observations indicated that the roller was not walking out

Inspection of figures 10 and 11 ‘shows that the nominal roller pressure « ex-
—- the penetration resistance or supporting capacity of the soils with the
7 exception of depths below 4 in. in area 4. This relationship is consistent with
_ the observed field behavior of the roller. It may be postulated that for effi-

cient ‘Sheepsfoot roller operation on these soils—that is, for the roller to walk

out during compaction — the nominal pressure on the roller feet should be i in
_ _ the order of 300 lb per sq in. or less at the water contents tested. Some vari-
ie in this value, depending on the size of foot used, would be indicated for _
aes 4,5, and 6, | based on the curves of figure 5. " Sufficient information was
not obtained to indicate the range of water contents which might have permit-

ted the roller used to walk out. However er, indications are that the upper end ©

_ Two different sheepsfoot rollers were used for compaction at Texarkana _
Dam. They were each loaded to a total load of 39,500 lb but the feet on one
roller were 8.25 sq in. in size and on the other roller were 13.68 sq in. in -
= . Computed nominal pressures on the feet were 595 and 360 lb per sq in.,
respectively. With both rollers the soil was compacted in 8-in. lifts for six
7 passes of the equipment. . It may be well to mentionthat both rollers achieved
- equivalent densities in compaction of the embankment. Observations in the
indicated that neither roller was walking out during however,

with the feet was said to be easier to operate.

“rollers is welll in excess of the maximum values obtained in the load- -penetra-
tion curves. This corresponds with the observed behavior of the rollers in the :
7 ‘field, inasmuch as neither walked out during compaction. It appears f from the
- curves that a a nominal roller pressure in the order of 100 lb per sq in. would —
_ permit the roller to walk out during compaction at the water contents tested. _
a _ The analyses presented in the preceding paragraphs indicate that for the
soils tested in this ro there is ageneral relationship between me

Son

"penetration resistances less than the nominal pressures of the
4 ler. This indicates that the soil did not have sufficient bearing capacity to
support the roller feet and therefore the roller sank until a

_ spreading the load o over a 1 sufficiently large a area ato reduce the contact pressure
- to a value which would support the roller. _ In three areas where the roller —
walked out, the load-penetration curves gave values" greater than the nominal —

_ Experiment Station data were also analyzed with respect to the moisture con-
i _ tents being wetter or drier than the field optimum values. Field compaction —
: curves were determined by plotting the field moisture content versus sed

|:
:
a
|

; and drawing a curve of best fit through the data. The field data for the Water-
Experiment Station tests provided a good field compaction curve. How-
ever, the data for Grenada were scattering and did not cover a sufficient a
= to develop a field compaction curve; in this case the laboratory com-
_ paction curves were used as a guide in drawing the field compaction curve. on
_ Average moisture contents for each test area were compared to the field com- 7
7 ‘paction curves to determine the percentage | wet or rdry of field optimum. a.
These values were plotted against the range of model foot penetration resist-
ances at 2-in. penetration for an foot sizes for each area and pairs of smooth

= It c can n be seen ¢ on figure. 13 that for moisture contents wet of field optimum —
- the allowable contact pressure as inferred from the penetration tests is very
_ much lower than for moisture contents at and dry of optimum. It appears con-
7 ceivable, therefore, that it might be possible to so select a contact pressure —
- > for sheepsfoot rollers as to result in the roller walking out for moisture con -
de
_ tents up to a certain per cent wet of optimum water content and for the roller _
to not walk out for moisture contents wetter than this value. The advantage
7 7 of this would be that a better control of field moisture content could be main- a
tained, since the behavior of the roller itself wouid assist the inspector in 4
judging whether or not the moisture contents were sienna for rolling. —

j

al The following conclusions are considered applicable for the types of of soils

a Model foot tests on field- ‘compacted soils | may be e used in predict the

_ probable behavior of a given sheepsfoot rotler in so far as efficient —
compaction operations are concerned, i.e., the roller feet walking out.

be hn approximate value of nominal roller foot pressures for satisfactory
oper ration may be obtained from model foot tests on a given soil. ales
© . The roller foot size had little or no effect on the penetration resistance
values when the water content of the soils tested was high. However,

_ lower water contents the smaller model feet | tended to we higher

. It ‘may be possible | to correlate the resist-
“— with model foot tests in the laboratory and with sheepsfoot roller

behavior, of the ‘method of prediction.

BIBLIOGRAPHY

(2) ) Compaction Studie: s
(3) Factors Influencing | the Compaction of Soils, w.

and A. A. Maxwell, Highway Research Board Bulletin No. 23, Nov. 1949, |

an
4 if
4
%
|
|
—
i
il
i
7
=

AAN

aed
~

Ty tm xn “Say

g0QD0ag oSeasay

q 5

«6°66
wed

wes]
(0S) pues
(oS) pues AafeTO
ABTS wee]
(19) ABT.

LeLot 6 *60T

JON

ted qT

STO}

uy Ard

:

ts"

UN\O

ag

Bone

(Phd hb

a
tm
— | g
|

—

ssemb

net

quipme

netration

Pe

ic

—-
—

AZAV15 €! O€ VOVN3YS
©

=

++ ——

hq

?

os
or

q

———+—--+ -+ 4-44

:

Lp

~ —
ears puepues

|

—
—
— 4 ad
— Sion
4 3 ia
—
|

00

oO

10902000
IN,FFOOT

AREA |

=

WwW
a
@

a

WwW
U
4
<

2

ww
WwW

a

-SQ q

S3HONI - NOILWYL3N3d S3HDNI -

—— 200 il
| —

ke

—
|
“aa -WATERWAYS EXPERIME NT STATION.

if |

PENETRATION -INCHES

‘LEGEND

6-SQ-IN FOOT
9-SQ-IN FOOT

4 IN FOOT

PENE TRATION-INCHES

x

r~

4

9 FIGURE 4 LOAD- PENETRATION |

AREA - WATERWAYS KPE RIMES
>

|

|
—

WES ROLLER 250 PS/ du 6-

=TEXARKANA ROLLER de PSI

5- GRENADA

GRENADA

bal FIGURE 5 LOAD AT 2-IN

15

EL Foot - $Q. IN.

MODEL FOOT AREA

—
—
—

| @ LEAN CLAY, GRENADA DAM

a SAND, GRENADA DAM

CLAY, TEXARKANA DAM
LEAN CLAY, wes

LEAN CLAY

(103.8

GRENADA

waren CONTENT-% OF DRY WEIGHT

GURE 6 L 6 LOAD AT 2-IN PENETRATI

WATER CONTENT

363-14

—
a
ON |

@ LEAN CLAY , GRENADA DAM

ia) CLAYEY SAND, GRENADA DAM 7
fe) LEAN CLAY,

wo

FIGURE 7 PROCTOR (NEEDLE PENETRATION

RESISTANCE LOAD ON MODEL FOOT

=

—
4 - ave
|

|
— FIELD OPTIMUM

LB IN.

PRESSURE
—SHEEPSFOOT ROLLER,
= 250 LB/SQ IN. ROLLER

E ON ROLLER FOOT-

a

ROLLER WALKED OUT

RUN NO. I-TEST FOOT, 16 SQ IN. AREA.
RUN NO. 2-TEST FOOT, SQIN. AREA.

ACTUAL

"363-1 6

~ '
|

|

—

2

4
=¥

A 4

>

- PRESSURE
SHEEPSFOOT ROLL ER, ‘a

AREA 2,17 TOI9%
WATER CON TEN ro

WwW

‘ IN

AREAS 38 5.13 TO/6 %
WATER CONTENT-—,\

‘ive

FIGURE COMBINED LOAD - ~PENE TRATION © CURVES

da

—
|

Ul

—

PENETRATION RESISTANCE - LB PER SQ IN. ee
200 + + £400 600 800 1000

—

PRESSURE_OF
SHEEPSFOOT ROLLER,
685 SQ IN.

NCH

AREA 4, 10 705%

WwW
WwW
a.

AREA 7, /0

q

WA CONTENT \

| AREAS 1,283,
{—— “i670 24%

NETRATION | CURVES

LEAN

— we, FIGURE 10 COMBINED LOAD -PE
— GRENADA DAM
—

‘PRESSURE ~
ROLLER, 685 LB/SQ IN.
IND 6

SHEEPSFOOT

PER SQ
—AREAS 5A

NOMINA

OF

363-19

—

°
o

LATION RESISTANCE-LB

S3HINI

a
| —
| | | |
FIGURE COMBINED LOA NE TRATION CURVES
GRENADA CLAYEY SANO

PENETRATION RESISTANCE -

NOMINAL PRES
SHEEPSFOOT ROLLERS,
960

TRATION-(\NCHES

WATER CONTENT

PENE

COMBINE CURVES

—
itr f

—

—

_ GRENADA LEAN CLAY:

MOISTURE CONTE PTIUM

RELATED TO FIELD OPTIMUM

4

AMERICAN SOCIETY OF CIVIL ENGINEERS

(OFFICERS FOR

PRESIDENT
DANIEL VOIERS TERRELL

af, on
_ Term expires October, 1954: a Term expires October, 1955:
EDMUND FRIEDMAN ENOCH R. NEEDI
G. BROOKS E ARNEST MASON G. L ockWooD:

"DIRECTORS

Term expires October, 1954 Term expires October, 1955: Term expires October, 1956: -

WALTER D. BINGER im CHARLES B. MOLINE AUX WILLIAM S. LaLONDE, IR.
FRANK A. MARSTON MERCEL J. SHELTON OLIVER W. HARTWELL

=

GEORGE W. McALPIN A. K. BOOTH THOMAS C. SHEDD-
A G. PAULSEN EL B. MORRIS.
he D. KNAPP ERNEST W. CARL

WARREN W. PARKS: GLENN W. “HOLCOMB RAYMOND F. DAWSON

PRESIDENTS

Members of the Board
—

TREASURER)” EXECUTIVE ‘SECRETARY

E
CHARLES E. TROU WILLIAM N. CAREY

ASSISTANT TREASURER ASSISTANT SECRETARY

GEORGE W. BURPFE- HANDLER

PROCEEDINGS THE SOCIE TY

HAROLD T. LARSEN

Manager of Technical

DEFOREST A. MATTESON,

Editor of Technical Publications

A. PARIS

Asst. Ex litor of Technical Publications

GLENN | HOLCOMB”

ERNEST W. CARLTON W. HARTWELL

SAMUEL ‘EL B. MORRIS

'
—
=
a
we
-
—
tm
—
—
=