,'

i-

NSWC MP-83-444

-I
GRIT BLASTING A DISTRIBUTED ROUGHNESS
BASED ON A 309/ PROBABILITY OF EXCEEDENCE
BY E. J. BECrER,

M. 0. JOSE

STRATEGIC SYSTEMS DEPARTMENT

,,

OCTOBER 1983

)A.proved for pualic

oeam, distribution unicnited.

DTIC
ELECTE
MAR 8

4../

NAVAL SURFACE WEAPONS CENTER

SINDahlgrn,

Virginia 22448 * Silver Spring, Maryland 20910

84

-'-"

03 08

UNCLASSIFIED
ECU.ITY CLASSIFICATION OF TrIS PAGE ("On Date Entered)

R
BEFORE
COMPLETING FORM

REPORT DOCUMENTATION PAGE

b'..%

*REPZ~ NUMB ER

NSWC

12. GOVT ACCESSION NO.

A~~

P83-444

BEFOREIZT'
CAOEG
RCPENSCTLGNME

S. TYPE OF REPORT & PERIOD COVERED

4. TIVLE (and Subtitle)

Final, Jan-Mar 1981

GRIT BLASTING A DISTRIBUTED ROUGINESS BASED ON

OF EXCEEDENCE
A 30% PROBABILITY
."6.

P.-

7.

FOMMRM

PERFORMING ORG. REPORT NUMBER

AUTOR(,)

CONTRACT OR GRANT NUMBER()

9.

E. J. BECKER
M. D. JOBE
3.

10.

PERFORMING ORGANIZATION NAME AND ADDRESS

Naval Surface Weapons Center (Code K22)

White Oak
Silver Spring, MD
II.

PROGRAM ELEMENT, PROJECT, TASK

AREA

& WORK

UNIT

NUMMERS

FY7653-81003-1

20910
12.

CONTROLLING OFFICE NAME AND ADDRESS

REPORT DATP

October 1983
NUMBER OP PAGES

13.

31
I4.

MONITORING AGENCY NAME & AODRESS(II dllerent from Controlting Office)

BMO/MNRTE
Norton AFB
Los Angeles, CA
16.

IS.

SECURITY CLASS. (of this

epef)

UNCLASSIFIED
IS&. OECLASSIICATION/DOWNGRAOING
SCH EDULE

DISTRIBUTION STATEMENT (of this Report)

Approved for public release; distribution unlimited.

(of the abestted

entered In Block 20. It different from Report)

17.

DISTRIBUTION STATEMENT

IS.

SUPPLEMENTARY NOTES

19.

KEY WORDS (Continue on reverse aid* It necessary and Identify by block nuber)

Surface Roughness
Roughness Characterization
Grit Blasting
Photomicrograph
Probability it Exceedence
1.

ABSTRACT (Continue on rever.e aide It neceeeamy amd identify by block number)

) The objective of this study was to obtain a prescribed roughness on a

metal surface through control of grit blasting techniques. It includes a
description of the equipment used to grit blast the metal samples, the grit*

blasting technique, the photomicrographic technique, and the procedure used

to characterize the roughness. As a result, calibration curves were obtained

which gave surface roughness as a function of grit blasting pressure setting,

grit size, and surface material.

DD I CAN7 1473

EDiTION oP I NOV 65
S/W 0102"LF"014'6601

ISOSNETE

UCLASSIFIED
SECURITY CLASSIFICATION Of THIS PAGE (mier Data
Xnt44

".

--.

NSWC MP 83-444
0-"

FOREWORD

This report documents a surface roughness study conducted at the Naval

Surface Weapons Center/White Oak (NSWC/WO) in February 1981. The objective of
the study was to obtain a prescribed roughness on a metal surface through
control of a grit-blasting technique. The grit-blasting technique was used to

generate a desired surface roughness height and distribution on wind tunnel

models.
The report includes a description of the equipment used to grit blast the
metal samples, the grit-blasting technique, and the photomicrographic technique
used to record the surface roughness. It also outlines the procedure used to
,*

characterize the roughness. Calibration curves for the grit-blasting apparatus

located in the Aerodynamics Model Shop (Bldg. 402-117) were obtained which give

surface roughness as a function of grit-blasting pressure setting, grit size,

and surface material.
This work was performed in support of the Maneuvering Thermodynamics and
Shape Change Technology Program conducted by AVCO Systems Division, Wilmington,
Massachusetts in the NSWC/WO Tunnel 8 (WTR 1346). A special thanks is expressed
to F. W. Rider (R32) for preparing and photomicrographing the metal samples.
Approved by:

C. A. FISHER

II

.;

Availability Codes

i
!
i'I.

I U2I

...... Avail and/or

Dist

Sp ee,1,l

...

NSWC MP 83-444

CONTENTS

Page
INTRODUCTION ..

.............. ............................

ee

..

GRIT BLASTING EQUIPMENT AND TECHNIQUE........................................

P HOTOHICROGRAPHIC TECHNIQUE ....................................

ROUGHNESS CHARACTERIZATION TECHNIQUE .................................

RESULTS AND CONCLUSIONS........

APPENDIX A

-RECOMMENDATIONS

.............

AND DATA.....................

...

*e

.................A-1

NSWC MP 83-444
,'S..

ILLUSTRATIONS

Page

Figure

GRITS USED IN STUDY (4X MAGNIFICATION) .............................. 5

SAMPLE BEFORE AND AFTER GRIT BLASTING, SETTING IN
6
ACRYLIC, AND SAMPLE HOLDER .......................................
7
.............................
SURFACE
AT
50X
MAGNIFICATION
3
ROUGHENED
8
SURFACE ROUGHNESS CHARACTERIZATION AND DEFINITION ..................
4
9
5
STAINLESS STEEL 17-4PH CALIBRATION CURVES ..........................
10
6
NICKEL 200 CALIBRATION CURVES ....................................
A-i
STAINLESS STEEL 17-4PH, 20PSI, #12 GRIT ........................... A-2
STAINLESS STEEL 17-4PH, 60PSI, #12 GRIT ........................... A-3
A-2
A-3
STAINLESS STEEL 17-4PH, 92PI, #12 GRIT ........................... A-4
A-4
STAINLESS STEEL 17-4PH, HEAT TREATED TO 1150 0 F,
A-5
20PSI, #20 GRIT ...............................................
A-5
STAINLESS STEEL 17-4PH, HEAT TREATED TO 1150 0 F,
................................... A-6
60PSI, #20 GRIT
A-6
STAINLESS STEEL 17-&PH, HEAT TREATED TO 1150 0 F,
10OPSI, #20 GRIT ............................................... A-7
.A-7 NICKEL 200, 20PSI, #12 GRIT ...................................... A-8
A-8
NICKEL 200, 60PSI, #12 GRIT ...................................... A-9
I
2

A-9

A-10

NICKEL 200, 96PSI, #12 GRIT ......................................

TABLES

Page

Table
1
2

STUDY SAMPLE DESCRIPTIONS ...........................................

COMPARISONS OF AVERAGE, ROOT MEAN SQUARED, AND 0.3 PROBABILITY

METAL HARDNESS NUMBERS ...........................................

12

........................

ROUGHNESS VALUES

. " -"- .

...''' ''-,*,-,"

.; ""',',..,-.-.-.-

. . -

-. .

13

r-w

iii

<'

11

" " . '.:

.5,., ,-

", ;

,'

NSWC MP 83-444

INTRODUCTION

Surface roughness is used on wind tunnel models for a number of reasons.

First, it is commonly used as a boundary layer trip to produce turbulent flow
over models which are tested at low Reynolds number conditions. Secondly, it is
used to generate aerothermodynamic effects such as augmented heat transfer or
skin friction. In each case, the exact character of the roughness must be known
in order to compare results from different tests and facilities.
The current study centers on a grit-blasted roughness and its characterization. Grit-blasting parameters under study included:

FI.

ofis, stud
mlkued
itposibledrayrti
toot
1. il
supply
pressure
to the grit-blasting

presrdsurfce
apparatus

controdln grit-baestg
tnqe lon
heys gritbraontionprs.
2. size of grit
used
thed Aeodyeneaiste
l Seohrodg.i
efecssuh2sau1ntd1et7rasero

byo

rugess
lcaedl

ini

3. type of metal being roughened

The above parameters were varied (see Table 1) and the resulting surface
roughness characterized using a photomicrographic technique and statistical
analysis based on a 30% probability of exceedence. Both the characterization
and the photomicrographic techniques are outlined in this report. The results

GRIT BLASTING EQUIPMENT AND TECHNIQUE

The grit-blasting equipment used for this study was a commercially
available sandblasting gun distributed by Montgomery Ward (model 6352) that had
been converted to a grit-blasting gun. This included fitting the gun with a

siphon feed and a larger nozzle (3/8" inside diameter).

The nozzle size was

increased to allow free passage of the larger grit. The gun was used with a
maximum pressure of l2epsi. The pressure was controlled through a regulator
valve that could be read to + lpsi.
th

erdnais
oe

So

(lg.42-1)

NSWC MP 83-444
The grit sizes used were #12 chilled iron grit and #20 silicon dioxide
grit. The #12 grit had an average size of 56 mils and was nonuniform in shape.
The #20 grit had an average size of 39 mils and also was nonuniform in shape
(see Figure 1).
After grit blasting a 1-inch-diameter sample for more than 3
minutes, it was noticed that the #20 grit started to break down into a powder.
It was felt that this happened because the #20 grit was a nonmetallic grit.
Metal test samples used in this study were cut from 1-inch-diameter stock
with a 0.10-inch thickness and a #32 finish before grit blasting. A metal
holder (see Figure 2) was used to hold the samples during the blasting process.
The holder supported the sample so that it would not become domed from grit
blasting and shielded an area of surface which was later used as a smooth
reference edge. The nozzle of the gun was held perpendicular to the sample at a
distance of 2 to 5 inches during grit blasting. Samples were exposed to the
blasting for a period of time which was sufficient to provide uniform and
complete coverage as determined by visual inspection.
It was found that
roughness uniformity did not improve after a finite grit-blasting time
(typically 2 to 3 minutes per square inch).
Prolonged grit-blasting produced no
mass loss, but simply resulted in a redistribution of the surface roughness.

PHOTOMICROGRAPHIC TECHNIQUE
The photomicrographics were produced at NSWC/WO by the Metallic Materials
Branch, R32.
First the roughened sample was cross sectioned using a diamond
edge blade.
Then half of the sample was encased in an acrylic plastic (see
Figure 2).
The sample was set in a form mold and the plastic allowed to harden
without applying any pressure or heat.
The edge of the sample was finished by sanding with progressively finer
abrasives. The process was started with 180-grit sandpaper on a belt sander.
The next sandpaper used was 240-grit with the sample being moved across the
sandpaper by hand.
This polishing process was continued using progressively
finer sandpaper, including 320-, 400-, and 600-grit.
Finally, the sample was
polished with 6-micron and 1-micron diamond paste on a sanding wheel.
The photomicrographs of a sectioned sample were taken using a Metaligraph
with #52 Polariod film. Pictures were taken at 50X or 10OX, depending on the
surface roughness. After each picture, the sample was moved for the next
picture while being careful to overlap the preceeding edge. A complete set of
photographs were taken for each surface and the edges of the photographs were
matched to form a continuous strip of photomicrographs.
Figure 3 shows a
portion of a continuous strip of photomicrographs at a 50X magnification.

ROUGHNESS CHARACTERIZATION TECHNIQUE

There are many different ways to define surface roughness.
Values can be
based on average peak-to-valley, root mean squared (overall and height only), or
significant peak-to-valley dimensions. The surface roughness value obtained
from these different methods vary (see Table 2) and depend on whether a

NSWC MP 83-444
profilometer or photomicrographic technique was used. The photomicrographic
technique was chosen as the more accurate method of measuring surface roughness
in this study because of mechanical limitations exhibited by profilometers.

*"

Each photomicrograph was characterized by first drawing an arbitrary

reference line (yo) on the photomicrograph. This line is chosen to be
representative of the "valley" locations. The yo line should be drawn
parallel to the shielded (i.e., smooth) portion of the sample and generally a
few millimeters below the original surface.
Once the yo line is drawn, height measurements are obtained at even
divisions along the reference line. There should be enough divisions so as to
result in 100 to 200 data points for statistical analysis. The distance between
the Yo line and the surface of the sample (y) is measured at each division
using a Bausch-Lomb 7x magnifer with graduated scales.
The y-yo values are plotted in histogram form and a probability of
exceedence vs. y-yo curve is generated. This curve is generated by
calculating the percent of the total data that exceeds the value of a certain
roughness size group. From this curve, a least squares fit straight line
between the 0.1 and 0.9 probability points is calculated. Then, using the slope
of this line, a straight line is drawn tangent to the curve at the 0.5
probability point. The y value at the intersection of this tangent line and a
probability of exceedence value equal to 1.0 produces a y correction value, ho
(see Figures A-la and A-lb).

With this ho value, a new reference line, called the optically apparent
surface, is drawn. Now only 'significant peaks' are measured from the new
reference line. 'Significant peaks' are the roughness elements which are
primarily responsible for triggering transitional flow. Again, a table of
h-ho values are obtained and a histogram and a probability of exceedence curve
are generated (see Figures A-lc and A-ld). Both of these are done in the same
manner as outlined for the y-yo data. From this probability of exceedence vs.
h-ho curve, the y-value where the probability of exceedence is equal to 0.3
(h3 0 ) is determined to be the roughness value for the sample. Since the
plane of measurement does not pass through the middle of each peak, the h30
value is multiplied by a shape factor of 4/w (hemispherical shaped elements)
to arrive at the final K30 value. Figure 4 summarizes this roughness
characterization technique.
RESULTS AND CONCLUSIONS

'.

Calibration curves for the grit-blasting apparatus were obtained for #12
grit and #20 grit and are shown in Figures 5 and 6. Three types of metal
(stainless steel 17-4 PH annealed, stainless steel 17-4 PH heat treated to
1150F, and nickel 200 annealed) were grit blasted at various pressure
settings. The mechanical properties of stainless steel 17-4 PH annealed and
heat treated to 1150 F were essentially the same (see Table 3) and no
noticeable differences were noted in the calibration curves. For a given
pressure setting, a larger roughness value was obtained on the nickel samples.

NSWC MP 83-444
Over the range of pressure settings that were tested (20psi to lOOpsi),

nearly linear correlations resulted. However, it is felt that there is a

threshold pressure above which the sample must be grit blasted before a surface
roughness can be measured. Additionally, there is a maximum pressure above
which no increase in surface roughness will be found.
The error associated with reading the photomicrographs is inversely
proportional to the magnification of the pictures. Error sources included
resolution of the photomicrographs, reading the seven-power calibrated eye-

*piece,

and data reduction procedures.

Appendix.

*"

These items are addressed in the

Using the error sources, an error bound was calculated depending on

the magnification factor. The error bounds are + 0.6 mils and + 0.3 mils for
50X and 10OX magnification respectively.
The trends that are seen make it possible to obtain a prescribed surface
roughness by using the calibration curves. These curves are intended as a
guide, and with the results of future roughness characterizations, the curves
will become more accurate. For now, these curves will give a good approximation
as to what pressure setting and grit size will be needed to obtain a prescribed

surface roughness.
.4

'P.

p.

U
i

,.

- .

. .

".

'

NSWC MP 83-444

,i

Ab

eow

Ij .
5.."

do.

:i

.. ;
.,',

9*

Sz

I.,4
,.

zz
I-L

40

ago

54b

im

..

NSWC MP 83-444

z
LuJ
z

0*

.j

00

z2

P&
LU

V)-

zz
a

LU

&

LU
-Ja
a

1~'

a.
2

OD

''

0
LL)
LU

:M a

coU

NIWC W U-4M

K
I
S

NSWC MP 83-444

'ARBITRARY' REFERENCE LINE (yo)

OPTICALLY APPARENT REFERENCE LINE (h0

(a)

Lu

1.0

Lz
Lu
Lu

L)
x

0.8

0.6

Lu

S0.4
0
cc
a-

0.2I

0
0

0.001

0.002

0.003

0.004

(v) INCHES
o)

1.0

LU

0.8

zLU
LU

x
K30

0.6

LL

h 3 0 x4/7r

0
S0.4
S0.2
0
0

0.001

0.002
(h-h 0 ) INCHES

0.003

0.004

(b)
FIGURE 4. SURFACE ROUGHNESS CHARACTERIZATION AND DEFINITION

NSWC M4P 83-444

I--

I-

C.D
00

ko

V)

LU

e~uj
CDJ

0=
4%

0)
'0

0D

CD

I
.

LA;,

0
.

wU

I--

Cdn

C\J

LL.

NSWC MP 83-444

I-L
C.DD

V)-

0.

*U

0
Lii

LU

L/)o
Lii

1-

I I.

-;

-4-L

I-X:c
0D
C)

CL

0.

.7. & -

-L

'.

.'

",".

*-

*.-

"-

"

"

"

"

"

'

"

"

V ".

NSWC MP 83-444

TABLE 1.
*

Sample
#

Metal
Type

STUDY SAMPLE DESCRIPTIONS

Pressure
(psi)

Grit
Size

K3 0 Roughness
Value (mils)

Magnification
Factor

SS-17-4PH

20

12

2.88

SS-17-4PH

60

12

4.10

50

SS-17-4PH

92

12

4.98

50

SS-1150

20

20

1.60

100

SS-1150

60

20

2.13

100

SS-1150

100

20

2.75

100

N-200

20

12

3.32

50

N-200

60

12

5.23

50

N-200

96

12

6.74

50

Metal Type Abbreviations:

Stainless Steel 17-4PH-------------------- SS-17-4PH
(annealed)
Stainless Steel 17-4PH-------------------- S-1150
(heat treated to 1150 0 F)
Nickel 200-------------------------------- N-200
(annealed)

i1
.......................

100

NSWC MP 83-444

TABLE 2.

COMPARISONS OF AVERAGE, ROOT MEAN SQUARED,

AND 0.3 PROBABILITY ROUGHNESS VALUES

Sample

Metal
Type*

Roughness Value (mils)**

Average
Root Mean Square
0.3 Probability

SS-17-4PH

20

1.79

1.91

2.26

SS-17-4PH

60

2.53

2.80

3.22

SS-17-4PH

92

3.40

3.64

3.91

SS-1150

20

0.97

1.09

1.26

SS-1150

60

1.35

1.46

1.67

SS-1150

100

1.73

1.90

2.16

N-200

20

2.03

2.19

2.61

N-200

60

3.60

4.11

:.

4..

Pressure
(psi)

N20603
N-200

96

3.31

4.321304

4.67

5.29

*See Table 1 for Metal Type abbreviations

**Values not multiplied by 4/w shape factor

U
1

12

..

0.

.--.

.,.
S5"

...

.1.*

NSWC MP 83-444

TABLE 3.

METAL HARDNESS NUMBERS

Metal Type

Hardness Number
(Brinell)

Stainless steel
17-4PH annealed
Stainless steel
17-4PH heat
treated to 1150OF
Nickel 200

332*

311*
i0**

annealed

-'.

*Properties given by Republic Steel Corporation, Cleveland, Ohio

44101

**Materials Engineering, Materials Selector Issue, Vol. 66, No. 5, Mid-October

1967, p. 156 and Standard Handbook for Mechanical Engineers, Seventh Edition,
McGraw-Hill Book Company, pp. 5-17.

f*

'..................................................................................
*.

13

..-

NSWC RP 83-444

APPENDIX A
RECOMMENDATIONS AND DATA

This section discusses improvements to the method used to characterize the

surface roughness. These recommendations are included in order to reduce error
sources in the future.

"

The magnitude of the error found depends largely on the magnification at

which the photomicrographs are taken. The greater the magnification, the easier
it is to measure surface heights, select significant peaks, and obtain a more
accurate surface roughness value. When comparing the cost of analysis and
preparing the photomicrographs to the accuracy desired, a maximum magnification
of 250X is suggested.

Another consideration is edge retention. When a harder metal is surface

roughened, it is suggested that a Bakelite method be used to encase the sample
edge. This includes using pressure and heat when preparing the sample to be
.polished.

When drawing the yo line, it has been found that a razor blade line is
much finer and allows the zero on the eyepiece to be located more accurately.
This procedure was implemented prior to the AVCO MTSCT Tunnel 8 test (WTR 1340)
and accounts for the additional data points in Figure 6. The bias between the
two calibration curves for the #12 grit in Figure 6 was due to the method of

*placing

the yo line on the photomicrographs. Note that the later calibration

curves have been extrapolated to larger roughness values. The criteria to use
for determining interval spacing on a set of photomicrographs is that a minimum
of 100 data points be used for determining both the h o and h3 0 values.

-"
.

By implementing these recommendations, it is felt that the error bound for

and OOX magnification can be reduced to + 0.4 mils and + 0.2 mils
respectively. The use of 250X magnification further reduces-the error bound to

*50X

+ 0.08 mils.
The figures that follow (Figures A-l through A-9) are the results of the
-

work documented earlier in this report. Each figure is comprised of four graphs
designated as A, B, C, or D. In each case graph A is the y-yo histogram plot,
graph B is the probability of exceedence vs. the y-yo curve generated for each

test sample, graph C is the corresponding h-ho histogram plot, and graph D is
the probability of exceedence vs. the h-ho curve.

A-i

NSWC MP 83-444

!-

tn~e

m~~jEc

~E
0

s'u d

0I-

~E

A CJ

?- C%

-.
awn

V)

00

-Wu
if

If.

C)i
If

It

CD)

C)
0

V)

0,-

II I-

1I 7I

%0J

)
0

Sl.U!.d 40 JaqwnN

A-2-

NSWC le 83-444

cvle'oj
11

11

'cv)Cj
If

0CC%!J4

EC

LOm

11Vm

If

c~cr

c..

czJ

03 0 0;

co
CA

"- -40

>

0~

co

()

-4

'c-J
CD

CD

A-3A

NSWC MP 83-444

r-

40

co

co

r-;

4*m-r

-4

o~~C

C\)O~~(VO

to

-C~

C%J

-4

cro

Tr

o%

00

0o

U-

A-4-

NSWC MP 83-444

- E

EC

C)C

'0

*0

CDi

liii

0
SJL~

qwnN

0o

.C
93a-8aX

0o
0

;o k;

vi

0)

SULOd ;o JaqwnN

aoupaSx
0

CD'0~

0 o 0X4Lq~

A-5-

Iq.d

NSWC MP 83-444

E
04 LO)

C4 14 4 V

LOl~
r%

-4%0
4

LC
U.-

CY)

0
J_

0n

m
I

-CD

0
LO

cmJ

-oL

-i-

CD
0

0o CD

-own

(AC

0"

L9%
O4 O

-4Ln~

0=

'CD

.-

-I

CD

0m

L-Dw
aO

%*

-60X

10RI

qe

-A

(V-2

il

NSWC MP 83-444

C)

I0

If

O'

Cr-4

O)Lfl

ON k

CO
C9

C)
-4)

(A)

If1

tCIf

(w6-

Cj

L.

CD1
C-

7-

00

SJUOd J.0jaqwnN

0m

a~a0ax 10RMIq~

U)

U)

A-7

0D

NSWC MP' 83-444

-4

-4

: -

to.

EA E
IA

cc CI

clC\E~

flC)

CD.C%

01

rl

CD

C)

CD

-E

51m

0t

ko

lt-z

ch

0.

r-q

cr

000)

S~~~~.UI.Od~~~%.o
~
UB~ZX
Jc

~~L.~Odu

zw~

C)~

C)

CD

CD

co

1t0

C;
S;UL~

10

aqw0
a~ua~93x3jo
k
L~qq~d0

A-8J

.,J

NSWC MP 83-444

CD.

c'J

C)

'I

CD

CD~*

CDJ

cc-

'.0

'.~e~jC.~j

CC

CCJ
I-*C

I-

A-94

NSWC xP 83-444

-'

~%0

00u,

%0
.C-

CD

%00

-04:

jo

'.0

-t

~ LO

IawNa.aa33

I4.~

D C. D C
j

-c

-~

-~

-W

S4U!.0d ;0 iaqwnN

0X-qq~

93uaPaaDx3 40 iX4tMqeqOJd

A-10

NSWC MP 83-444

DISTRIBUTION
Copies
Defense Technical Information
Center
Cameron Station
Alexandria, VA 22314

--

Library of Congress
Attn: Gift and Exchange
Division
Washington, DC 20540

(Yanta, W. J.)1

R32

(Rider, F. W.)1

E35 (GIDEP Office)I

E419
E432
4

25
1

K23 (Cornett, R. H.)

J. P.)
(Kavetsky, R. A.)
(Marshall, R. J.)
(Metzger, M. A.)
(Rorabaugh, A. D.)
(Rozanski, C. S.)
(Schlie, R.)
(Truitt, R. J.)
(Waldo, J. F.)
(Waser, R. H.)

1
1
1
1
1
1
1
1
1
1
1

K24 (Ausherman, D. M.)

(Baltakis, F. P.)
(Cothran, S., Jr.)
(Driftmyer, R. T.)
(Hedlund, E. R.)
(Hill, J. A. F.)
(Jobe, M. D.)
(Knott, J.)1
(Krumins, M. V.)1
(Prats, B. D.)1
W. C.)
*(Ragsdale,
M. M.)
*(Roberts,
(Smith, T. S.)
(Voisinet, R. L. P.)1
(Watts, E. B., Jr.)1
(Westermeyer, R. H., Jr.)I

1
1
1
1
1
1
1

.(Johnson,

K24

12

Internal Distribution:
K22 (Becker, E. J.)
(Morrison, A. M.)
-

Copies

1
1
1