Dew Point/Temperature Curves for

Selected Metal/Metal Oxide Systems

in Hydrogen Atmospheres

A new dew point/temperature diagram

is developed for process engineering purposes
in metallizing, brazing and sintering

BY M. C. REY, D. P. KRAMER, W. R. HENDERSON, AND L. D. ABNEY

ABSTRACT. As an aid in determining ture's dew point is easily done and pro- condensed and gaseous states, respec-
metallizing, brazing and sintering condi- vides a convenient, if indirect, measure of tively. The equilibrium constant for this
tions, a new dew point/temperature dia- the hydrogen-to-water ratio. reaction is given by the expression,
gram is developed for selected metal/ In order to provide a direct correlation
metal oxide equilibria in hydrogen atmo- between dew point and metal/metal K = (PH2o/PH2)n (2)
spheres. The new diagram constitutes an oxide equilibria, Chang (Ref. 1) devel-
where PH2O and PH2 are the partial pres-
expanded and more accurate version of oped a diagram based on widely known
sures of water vapor and hydrogen,
the original set of curves developed ear- thermodynamic data for several metals
respectively, and the activities of the
lier by W. H. Chang (Ref. 1). important to metallizing and brazing
condensed phases are assumed to be
Experiments carried out to check the operations —Fig. 1. The authors exam-
unity. The net free energy change associ-
calculated curves show that chemical ined Chang's calculations with the inten-
ated with the reaction is obtained from
kinetics, rather than equilibrium thermo- tion of extending the range of the dia-
the relation,
dynamics, determine the temperatures of gram, but they arrived at results different
reaction initiation for the majority of the from Chang's for several of the metal/ AG n e l = -RT 1nK (3a)
systems investigated. In these cases the metal oxide systems represented in Fig. 1.
empirically determined dew point/tem- The authors proceeded to recalculate where R is the gas constant and T is the
perature curves supercede the calculated and to extend all of Chang's curves and absolute temperature in Kelvin. If calories
ones for process engineering purposes. undertook a program of measurements are the units specified for AGnet and
to provide an experimental check of the logarithms to the base ten are employed,
calculated results. equation (3a) assumes the form:
Introduction
AG n e t (calories) = -4.576 nT Logio
Metallizing a ceramic surface and braz- Method of Calculating Dew Points (PH 2 O/PH 2 ) (3b)
ing a ceramic to a metal must be carried
out in atmospheres sufficiently reducing The reaction between a metal oxide The total pressure in the system is
to prevent oxidation of the system's met- and hydrogen can be represented by an assumed to be one atmosphere, so that
al components. Sintering of an oxide/ equation of the form: PH 2 = 1-PH2O-
metal composite further requires that the M m O n (c) + nH2(g) = Making this substitution in equation (3b):
atmosphere be sufficiently oxidizing to mM(c) 4- n H 2 0 (g) (1)
prevent reduction of the desired oxide 10-Q
PH2o(atm) = (4)
and to facilitate the removal of any where M is the metal and c and g denote 1 4- 10_Q
organic binders present. The necessary
atmospheric conditions are commonly where
M. C REY. D. P. KRAMER, W. R. HENDERSON,
provided by gas mixtures consisting of ANDL D. ABNEY are with Monsanto Research Q = AG net /4.576nT (5)
hydrogen and water vapor in appropri- Corporation, Mound Facility, Miamisburg,
ate ratios. Measurement of a gas mix- Ohio. To calculate AG n e t , the reaction (1) is

162-s | M A Y 1984
 80 t h e literature (Ref. 2b) and solving f o r A
FeO^^^ a n d B.
60 Selection of d i f f e r e n t t a b u l a t e d values
f o r substitution into e q u a t i o n s (9a) a n d
u_ 40 (9b) will result in slightly d i f f e r e n t values
FE 3
o

11i
DC
?n V^ f o r the t w o constants. A judicious choice
o f n e a r - e x t r e m a f r o m t h e tables will m i n -
LU
nz 1 imize the e r r o r inherent in t h e a p p r o x i -
D_
0
CO
O
1—
<
-20 ' / 2 m a t e relation given b y equations (9a) a n d
(9b). T h e values selected f o r A a n d B in
c o m p u t i n g the equilibrium d e w points
w o 2 ^ ^^1/lo02
X -40 -J- 7N$i021/ in this p a p e r are: A = -2387.718, B = -
11 6.401.
o
h- -60 T h e H 2 0 partial pressures calculated
zn(
} / vinoy / / f o r the various m e t a l / m e t a l o x i d e equilib-
ria are such that m a x i m u m d e v i a t i o n o f
/Cr2037 - t h e c o m p u t e d d e w p o i n t values f r o m
/ /'' tables in the literature (Ref. 2b) is less than
-100
^ /
L-Li...
1 /Ti0 2 >
/ / / I
4 ° C (7.2°F). T h e discrepancy is greatest
/
-120 f o r equilibria at the l o w e r t e m p e r a t u r e s

-140 o o o o o
o o o o o
//'/....,_T o o o
o o o
( < 5 0 0 ° C , i.e., 932°F). For m e t a l / m e t a l
oxide equilibria at t e m p e r a t u r e s higher
CD N cocno CO L O CD than 5 0 0 ° C (932 C F), t h e discrepancy
co co ^r b e t w e e n calculated a n d t a b u l a t e d d e w
TEMPERATURE. °F
points is less t h a n 2 ° C (3.6°F) a n d , in
Fig. I — After Chang (Ret. I). Used with permission m a n y cases, less than 1 ° C (1.8°F). Thus,
the a g r e e m e n t is best in the t e m p e r a t u r e
range applicable t o t h e processes o f
metallizing, brazing a n d sintering.
w r i t t e n as t h e sum of t h e t w o chemical pressure w h i c h is essentially the Clausius-
reactions; C l a p e y r o n e q u a t i o n . Values o f the c o n -
Discussion of Calculations
stants A a n d B are o b t a i n e d b y selecting
M m O n = m M 4- V 2 n 0 2 (-mAG-t) (6)
t w o n e a r - e x t r e m e values f r o m the H 2 0 Results of t h e a b o v e calculations are
nH2 + Vi n 0 2 = nH 2 0 (+nAC2) (7) v a p o r pressure vs. d e w p o i n t tables in s h o w n in Fig. 2. T h e r e g i o n o f the g r a p h

T h e n AG n e t is equal t o t h e e n e r g y differ- TEMPERATURE, °C

ence given by the expression, o CO
o Oo O
oO o
o o o o o CD O ooo
n A G 2 - mAG-i. o o o o CD O ooo CNJ -sr c o c o
C\J oo o LO CD r— X>CT>T—
T h e t w o f r e e energy changes !\G^ a n d 160 i l i 'i 1
-JM r 1 1 1 i- 70
A G 2 are o b t a i n e d f o r t h e a p p r o p r i a t e
o x i d a t i o n reaction f r o m t h e e q u a t i o n 140 60
(Ref. 2a):
120 50
A G = H 0 + 2.303a T log 1 0 T + 40
b X 10~ 3 T 2 4- c X l O ^ " 1 + IT (8)
100
- 30 o
w h e r e H D is the standard enthalpy o f 80 0
f o r m a t i o n o f t h e o x i d e ( M m O n , o r H 2 OJ 20 UJ
and T is t h e absolute t e m p e r a t u r e in 60 rr
UJ

Kelvin. T h e f a c t o r s multiplying t h e t e m - - 10 Q_

p e r a t u r e in e q u a t i o n (8) are t a b u l a t e d f o r
m a n y m e t a l / m e t a l oxide reactions in the
literature (Ref. 2a). T h e e q u i l i b r i u m partial
tn
o
40
20
F e O ^
f
Fe 3 (\A . . . -ll\/lr |0:
(1
-10
cn
o
:>

pressure of w a t e r v a p o r is t h e n o b t a i n e d 0
f o r a g i v e n reaction t e m p e r a t u r e b y " -20
Fe203 y
11

-20 o
applying t h e c o m p u t e d value of A G n e t t o -30 1
—
e q u a t i o n (5) a n d substituting t h e resulting 7»
value of Q into e q u a t i o n (4). o -40 2 10 i -40 n
CL /si0 2 L^

The d e w point corresponding to the

-60 ^ M o O , y -50 •5
111
calculated P H 2 O is o b t a i n e d f r o m t h e Q
equation: -80 -60

A -100
^W02 J--70
D e w point = L r2 3 IJ -A2U3
-80
logio (PH2O) +B
-120
273, in ° C (9a) - -90
-140
A o o CO CO O O OOOO o o o co CO
= 32 + - (- o o o O CO CO o o o o o CO CO o o
CNJ CO LO CD r^oocno LD CO LO
5 log10(PH2o)+B CO CNJ CO CO
(9b) CNJ

-273), in °F TEMPERATURE. °F
Equation 9 is an a p p r o x i m a t e relation Fig. 2 —Newly calculated curves for selected metal/metal oxide equilibria in hydrogen atmo-
b e t w e e n d e w point and water vapor spheres

W E L D I N G RESEARCH S U P P L E M E N T 1163-s
above each curve is the region of oxidiz- diffraction. All reactions were carried out ARGON SOURCE
(PURGE!
ing conditions for the metal affiliated with using a DuPontt, model 990 thermal ana-
the curve, while the region below a given lyzer control module in conjunction with
curve comprises the region of reducing either a Dupont model 951 thermogravi- GAS
SATUHATOR
conditions for the corresponding oxide. HYDROGEN
metric analysis (TGA) cell or a DuPont SOURCE =Bn 3S7 DEWAND HYGROMETER
POINT
Figure 2 does not include curves for
copper, nickel, cobalt or silver, because
high temperature differential thermal CONTROL
J]
n
analysis (DTA) cell connected to a model
DEW POINT p
these metals are too noble. Their curves 910 differential scanning calorimetry MEASURING! CHART
RECORDER
would lie well outside the low tempera- (DSC) cell base. For reactions carried out HEAD

ture/high dew point limits of the graph. at temperatures lower than 1050°C
In the high temperature/low dew (1922°F), the TGA apparatus was used
point region of Fig. 2, several of the and the specimen weight was continu- INERT

curves are in good agreement with the ously monitored. UU REFERENCE

MATERIAL

corresponding ones in Chang's diagram, Typically, a specimen was heated in

particularly for Al 2 03, Si0 2 , and Cr2C>3. the TGA cell at a rate of 5°C/min (9°F/ GAS EXHAUST -
The curves for M n O and ZnO match min) until a significant weight change was
Chang's at very low dew points, but observed, then cooled several hundred X-Y CONTROL
diverge with increasing equilibrium tem- degrees below reaction temperature, RECORDER CONSOLE
perature. For the Zn-ZnO equilibrium at and heated again. This process would be
1500°F (2732°F), Chang's curve gives a repeated three to six times in the course Fig. 3 —Schematic of experimental apparatus
dew point of —5°F (—21 °C), lower by of a heat treatment. used for differential thermal analysis (DTA)
40°F than the corresponding dew point Isothermal reactions were also carried
shown in Fig. 2. out using the TGA apparatus. Oxides
The disparity between Fig. 2 and were rapidly heated to set point temper- high temperature DTA cell was used to
Chang's diagram is most striking for those atures in dry hydrogen and isothermally carry out isothermal reactions at temper-
equilibria which are represented in the reduced. With temperature held con- atures higher than 1050°C (1922°F). The
low temperature/high dew point region stant, the dew point was then increased DTA traces could not be used to deter-
of the graph. For example, the dew point in order to re-oxidize the specimen, and mine reaction initiation temperatures,
for W - W 0 2 equilibrium at 100°F decreased again. In this way a calculated since deflections were also caused by
(37.8°C) is 90°F (32.2°C) according to curve of Fig. 2 could be bracketed. The melting, volatilization, annealing and sin-
Fig. 2, but - 2 8 ° F (-33.3°C) according to tering. This circumstance necessitated the
Chang's diagram; and the difference fE.I. DuPont De Nemours & Co., Inc., Instru- heating of numerous specimens of the
becomes larger with increasing equilibri- ment Products Div., Wilmington, Del. same species in order to obtain data.
um temperature. Generally, for the sys-
tems which fall in the low temperature/
high dew point region, Chang's curves
are shifted to higher dew point values TEMPERATURE. °C
than the corresponding curves of Fig. 2. CO
CD O O O O
o o o
CD
CD
O
CD
O O OCDiO
CD CD C D O C D
CD
OJ
CD O O
"St" COCO
The source of the discrepancies o CO
CNJ
CO "^ LTO ( A N COODi— 1— 1—

between Chang's curves and the newly 160 1 1 r 1 1 1 r 1 i 1 1 1 \\ \- 70

calculated curves is not known. The 140 60
present authors have checked their
results against Richardson plots (Ref. 3), 120 50
and agreement with these has been satis- 40
factory. 100
The curves of both diagrams are based
30 CO
0
80
on the assumption that the metal and its 20 LU

oxide are completely immiscible. The 60 cc

LU

validity of this assumption extends to only CL 10 Q_

a limited number of systems, necessitat-
CO 40 - - • —

"Td cn
o 0
ing experimental confirmation of the cal- 0 o>
20
culated curves.
r Z 1 -10 <c
0 1 -20 3_
Experiments 1J
O
-20 -30
High purity powders of the metals or
their oxides were heated in flowing, con- o -40 -40 o
trolled dew point mixtures of hydrogen CL

and water vapor. Dew points were mon- -60 -50

itored and controlled to ± 1°C ( ± 1.8°F)
. -60
-80
using a Panametrics* model 700 hygrom- Cr 2 0 3 / - -70
eter and Panametrics model 101 moisture -100 >
generator. The powder specimens were - -80
weighed both before and after heating, -120
and were contained during heat treat- A - -90
ment in either platinum or alumina cups. o CO CO CD CD CD O O O O co co co co co
CO CO CD co co 0 0 0 0 CO CO CO CO CO
CO CO
Phase identification of the reacted CNJ uo CD r~-oocx>o LO
T—
CO
CNI
LO
CNJ
CO LD
CO CO
material was made by means of x-ray
TEMPERATURE. °F
*Panametrics, Inc., Waltham, Massachusetts Fig. 4 — Experimental data for Cr-Cr203 system superimposed on calculated curve

164-slMAY 1984
 600 A Fe203
Calculated /
• FeO
• Fe304

"of 5 5 0
y ^
3

Measured^,^-"^
i 500
t-
- /

450

i i i i i i
-20 -10 -10 0
Dew point, ' C Dew point, °C

Fig. 5 —Reduction initiation temperatures vs. dew points for ZnO in Fig. 6—Measured reduction initiation temperatures for Fe2Oj, FejO^
H2/H20 gas mixtures and FeO in H2/H20 gas mixtures

700 •
• First heating
my
/%
650 • Subsequent heatings •
^m
600 - JL-——"•" * ***
m
£
| 550
a. x^***
E

500
-

450 " *"

i i i
-30 -JO -10 0 -20 -10 0
Dew point, C Dew point, C

Fig. 7-Measured reduction initiation temperatures for Mo02 in H2/ Fig. 8-Measured reduction initiation temperatures for W02 in H2/H20
H20 gas mixtures gas mixtures

X-ray analysis alone was used to deter- that equilibrium thermodynamics ade- unity, the resulting water-to-hydrogen
mine if reaction had occurred during a quately describe the behavior of this ratio consistent with the equilibrium is
DTA run. A schematic of the DTA exper- system. greater than it would be if K were given
imental configuration is shown in Fig. 3. The Zn-ZnO system exhibits marked by equation (2). Therefore, the calculated
The TGA configuration used was, except deviation from ideal behavior —Fig. 5. dew points for the Zn-ZnO equilibria are
for the furnace module, the same in all Zinc oxide is reduced at temperatures too low, as Fig. 5 shows. Efforts to detect
respects. much lower than the thermodynamic oxidation of zinc in wet hydrogen were
calculations predicted. Zinc has a low unsuccessful; the zinc specimens vola-
melting temperature and the liquid has a tized when heated. The empirical Zn-
Results and Discussion
high vapor pressure, boiling at 907 °C ZnO curve in Fig. 5 was generated by
Figure 4 shows the results for the (1665°F) (Ref. 2c). The effects of having reducing zinc oxide powders.
Cr-Cr 2 03 system. The solid curve is taken both liquid and gaseous phases of the Figures 6, 7 and 8 show oxide reduc-
from Fig. 2. The error bars bracketing the zinc present is to reduce the activity tion temperature vs. dew point data for
curve represent the limits of deviation of ac(Zn) of the condensed phase to some oxides of iron, molybdenum and tung-
the empirical data from the predicted value less than unity, and to introduce the sten, respectively. The observed reaction
results. The data confirm that the thermo- activity ag(Zn), also less than unity, of the initiation temperatures are, in all cases,
dynamic model described previously is gaseous zinc phase into the expression considerably higher than the correspond-
valid for the Cr-Cr 2 03 system. The for the mass action constant. Conse- ing metal/metal oxide equilibria tempera-
assumption made in the calculation that quently, equations (1) and (2) are more tures shown in Fig. 2. The phase diagram
the t w o condensed phases, chromium correctly written as follows: for the systems Fe-O (Ref. 4b), M o - O
and chromic oxide, are completely immis- (Ref. 4c), and W - O (Ref. 4d) indicate that
cible and, therefore, have unit activity — ZnO(c) 4- H2(g) = xZn(c) + none of these metals is soluble in its
equation (2) — is consistent with the Cr-O oxides. Furthermore, reaction tempera-
(1-x)Zn(g) + H 2 0 (10)
phase diagram (Ref. 4a). In addition, for tures are well below the melting temper-
dew points higher than - 4 0 ° C (-40°F) where 0 < x < 1 , and atures of all three metals. Therefore, the
the equilibria temperatures for the Cr- assumption of unit activity for the con-
Cr 2 03 system in hydrogen atmospheres K = (P H 2 o/PH 2 )-a c (Zn).a g (Zn) (11) densed phases participating in the reac-
exceed 1000°C (1832°F). At these high Since the mass action constant K is a tions should be valid.
temperatures the reaction kinetics appar- function only of temperature, and the It remains that reaction kinetics are too
ently proceed at rates sufficient to ensure product of ac(Zn) and ag(Zn) is less than slow for equilibrium to be attained over a

WELDING RESEARCH SUPPLEMENT 1165-s

period of several hours for the systems or delaminated products which permit and their accuracy improved. A program
that lie in the low temperature/high dew the passage of gaseous reactants and of measurements carried out to check
point region of Fig. 2. It is to the domi- products to and from the specimen sur- the validity of the calculated curves has
nance of chemical kinetics over equilibri- face. shown that chemical kinetics dominate
um thermodynamics that we attribute 2. Partial reaction of a crystalline spec- equilibrium thermodynamics in the low
our inability to oxidize tungsten, molyb- imen can greatly increase the degree of temperature/high dew point regime. For
denum or iron in wet hydrogen during disorder in the crystal by an increase in the metal/metal oxide systems thus
heats of up to seven hours duration. the number of vacancies or interstitial affected, the empirically determined tem-
Figures 7 and 8 show that, for M o 0 2 ions, or the formation of non-equilibrium perature/dew point curves may be
and W 0 2 , the reduction initiation tem- phases; increased disorder serves to referred to for process engineering pur-
perature observed during the first heating make the specimen material more reac- poses.
of a specimen was higher than for the tive, effectively lowering the tempera-
subsequent reheatings. The reaction start ture at which the specimen will further
temperatures observed in the second react in a given chemical environment. References
and following cycles were repeatable, so The process engineer should be aware 1. Chang, W. H. 1956. A dew point-
that t w o distinct reaction start tempera- of these possibilities and of those cases temperature diagram for the metal-metal
tures were to be considered. for which their effects are significant. oxide equilibria in hydrogen atmospheres.
The authors attribute the observed Figures 7 and 8 clearly demonstrate this Welding Journal 35(12):622-s to 624-s.
differences between initial and subse- significance for the reactions of M o 0 2 2. Weast, R. C , ed. 1978. Handbook of
quent reduction initiation temperatures and of W 0 2 with hydrogen/water gas Chemistry and Physics, 59th ed., (a) —pp. D45
to one or both of t w o unrelated mixtures. to D50, ( b ) - p p . D231 to D233, ( c ) - p . B65.
Cleveland: The Chemical Rubber Company.
causes:
3. Darken, t. D., and Gurry, R. A. 1953.
1. Contaminants adsorbed onto speci- Summary Physical Chemistry of Metals. N e w York:
men surfaces can form protective barri- McGraw-Hill Book Company, Inc.
ers between the specimen particles and The equilibria curves for metal/metal 4. Samsonov, G V. 1973. The Oxide Hand-
the surrounding atmosphere until the oxide systems in H 2 / H 2 0 atmospheres book, ( a ) - p . 4 7 2 , ( b ) - p . 4 7 4 , ( c ) - p . 4 7 6 ,
contaminants are thermally desorbed or originally constructed by Chang have (d) - p.492. New York: IFI Plenum Data Corpo-
are themselves reacted to form porous been recalculated, their range extended ration.

WRC Bulletin 289

November, 1983

Hot Cracking Susceptibility of Austenitic Stainless Steel Weld Metals

by C. D. Lundin and C. P. D. Chou

This comprehensive report describes the influence of the alloying elements and multiple t h e r m a l
cycling on the hot cracking resistance of austenitic stainless steel weld metals. In addition, the effect of
composition and multiple t h e r m a l cycling on the hot ductility is also determined and correlated with the
hot cracking tendency.

Publication of this paper was sponsored by the Subcommittee on Welding Stainless Steel of the High
Alloys Committee of the Welding Research Council.

The price of WRC Bulletin 289 is $16.50 per copy, plus $5.00 for postage and handling. Orders should
be sent with payment to the Welding Research Council, Room 1 3 0 1 , 345 East 4 7 t h St., New York, NY
10017.

166-s I M A Y 1984
 COMPLETE BOTH SIDES AND
AUTHOR APPLICATION FORM MAIL WITH 500-1000 WORD
-FOR- SUMMARY TO:

66th Annual AWS Convention American Welding Society

P.O. Box 351040
Las Vegas, Nevada, April 29-May 3,1985 Miami, Florida 33135

Date Mailed
Author's Name Check how addressed: Mr. D Ms. •
Title or Position Dr. • Other
Company or Organization
Mailing Address
City State Zip Telephone: Area Number

If there are to be joint Name

Authors, give name(s) Address
of other author(s) Name
Address
PROPOSED TITLE (10 words or less):

SUBJECT CLASSIFICATION:
Classify your paper by placing a check mark in the t w o appropriate boxes:
Applied Technology • Research Oriented • Other D / A n Original Contribution • A Review D Progress Report •

SUMMARY:
• Type, double spaced, a summary of not less than 500—but preferably not more than 1000 — words and attach to this
form.
• Be sure to give sufficient information to enable the Technical Papers Committee to obtain a clear idea of content of the
proposed paper; confine background to about 100 words. Also be sure to emphasize Results and Conclusions since this
material, together with information on what is NEW, will have a very important bearing on the decision of the Technical
Papers Committee.
• If complete manuscript is available, in addition to summary, please attach t w o copies of this form.
• Application Form and summary must be postmarked not later than August 15, 1984, to ensure consideration for inclusion
on the 66th Annual AWS Convention Professional Program.

MANUSCRIPT DEADLINES:
• All manuscripts should be in the hands of the Technical Papers Committee not later than March 1,1985. If received prior to
November 30, 1984, every effort will be made to publish them in advance of meeting.
• It is expected that the Committee's selections will be announced sometime in October or November 1984.
• If your paper is made a part of the program, which of the following manuscript deadlines will you be able to meet?
November 30, 1984 • January 3 1 , 1985 • March 1, 1985 •

PRESENTATION A N D PUBLICATION O F PAPERS:

• Has material in this paper been previously presented in meeting or published?
YesD NoD When?
Where?
Following presentation at the 66th Annual AWS Convention, would you accept invitations to present this paper before
AWS sections? Y e s • NoD

Papers accepted for presentation become the property of the Society with original publication rights assigned to the
Welding Journal.

RETURN TO AWS HEADQUARTERS. MUST BE POSTMARKEDMOT

LATER THAN AUGUST 15,1984, TO ENSURE CONSIDERA TION.

Author's Signature
Author(s) of Proposed Paper:
SUMMARY CODE
(For AWS Headquarters Use Only)

Title of Paper (may be abbreviated)

TO AUTHOR(S): Please briefly answer questions in spaces provided below—also

please note that answers given below are required in addition to 500
to 1,000 word Summary. Please use typewriter.

(1) WHY WAS THE WORK DONE?

(2) WHAT WAS DONE?

(3) WHAT WAS FOUND?

(4) WHAT ARE YOUR MOST IMPORTANT CONCLUSIONS?

(5) TO W H O M D O YOU THINK YOUR PAPER WILL BE IMPORTANT? HOW?

OTHER COMMENT (Optional).

RETURN TO A WS HEADQUARTERS. MUST BE POSTMARKED NOT Be sure to attach 500-1,000 word Summary when
LATER THAN AUGUST 15, 1984, TO ENSURE CONSIDERATION. mailing completed Author's Application Form to AWS
Headquarters.