leyel~ - dtic · may report 1979date attn: drdar-clj-r -is. number ofpages aberdeen proving ground,...
TRANSCRIPT
LEYEL~CONTRACTOR REPORT ARCSL-CR-79020
SURVEY OF TECHNIQUES FOR CLEARING MILITARY SMOKE CLOUDS
byZ
Josef Podzimek
May 1979
UNIVERSITY OF MISSOURI - ROLLAC6. Space Sciences Research Center1Lu.., Rolla, Missouri 66401
'LU-D DC
Contract No. DAAK-11,78-C-0O -71 Q rf7DA AK -11-78 .C -00 11-i•.. _..• ..j..
D U 2 19
D
US ARMY ARMAMENT RESEARCH AND DEVELOPMENT COMMANDV. Chem"ical Systems LaboratoryAberdee Proving Ground, Maryland 21010
SREPRODU 16 14FO
REPRDUCED FROM
Disclaimer
The views, opinions and/or findings contained in this report are those of the authorsand should not be construed as a.,i official Department of the Army position, policy or decisionunless so designated by other documentation.
Disposition
Destroy this report when it is no longer needed. Do not return it to the originator.
REPRODUCED FROMBEST AVAILABLE COPY
_ _ _ _ _ _ _I
'14
UNC LAS SI F IEDSECURITY CLASIPtCATIIN OF THIS PAGE MWe" Data Efitff0 ________________
REPOT DCUMNTATON AGEREAD INSTRUCTIONSREPOT DM ENTAION AGEBEFORE COMPLETING FORM
I. REPORT NUIMUER 1 -2. GOVT ACCESSION NO. 3. RECIPIENT'S CATALOG NUMBER
ARCSL-CR-79020 / '~-I____________
14. TITLE (ad Stawitle) S. TYPE OF REPORT & PERIOD COVERED
SURVEY OF TECHNIQUES FOR CLEARING Final ReportMILITARY SMOKE CLOUDS September 1977 to September 1978
S. PERFOR111N11 noR. REPORT NUMBER
7. AUTHON(e) S. CONTRACT OR GRANT NUMqR()
Josef Podzimek DAAK 11- 7 8 - Q' 1'a
9. PERFORMING ORGANIZATION NAMIE AND ADORESS 10. PROGRAM ELEMENT, PROJECT. TAS-KUnivrsit ofMissuriIoll" eAREA & WORK UNIT NUMBERS
Space Sciences Research 'Center <-Rolla, Missouri 6540111. CONTROLLING OFIC NAME AND ADDRESS 12. REPORT DATE
Commader/irecto Chemical Systems Lab. May 1979ATTN: DRDAR-CLJ-R -is. NUMBER OFPAGES
Aberdeen Proving Ground, Maryland 21010 20414. MONITORING AGENCY NAME & ADORESS(II different from Controlling Office) It. SECURITY CLASS. (of this e port)
Commander/Director Chemical Systems Lab.ATTN: DRDAR-CLB-PS UNCLASSIFIEDAberdeen Proving Ground, Maryland 21010 SCEDULE~z~rO/OWGADN(CPO Mr. R. Frickel 671-2808) NA
1S. DISTRIBUTION STATEMENT (of this Report)
Approved for public release; distribution unlimited.
17. DISTRIBUTION STATEMENT (of the abstract entered let Stock I.ft different hose Report)
IS. SUPPLEMENTARY NOTES
This study was sponsored by the Army Smoke Research Program,Chemical Systems Laboratory, Aberdeen Proving Ground, MD.
It. KEY WORDS (Continue on reverse aide It necessay and Identify by block num~ber)
(U) Smoke CoagulationSmoke clearing Visual rangeFog dispersal Cloud seedingAerosol stability Aerosol scavenging
M A ýT"ACr M; -ý aw -ees hsNnssaWm dwrb lc nsinbov)
ý(U)q A literature search and critical review on the subject of thec 'learing of fogs and smokes was conducted. The results, whichinclude descriptions and evaluations of a variety of possible smokeclearing techniques, along with hundreds of citations and abstractsof articles on pertinent subjects, are presented.
DD I Pam MIS b .GoSMOrSSI NOV 111Is .eTa UNCLASSIFIED 'SECU~RIY CLASSIFICATION OF THIS PAGE (fwte Date Entered)
~ F, N
FOREWORD
At the onset of this study the author was asked to "conducta survey of all past and present unclassificd work in areasrelated or applicable to the clearing of military obscuration
agents including, but not limited to, natural fog dispersionweather modification, condensation trail suppression, vapor andaerosol cloud dilution and dissipation, aerosol agglomerationby turbulence, electrostatic precipitation, and any other relatedwork".
This study began one year ago with the collection of unclas-sified literature and information. During this time the author,a cloud physicist, was introduced to the problems related toobscuration techniques by Dr. E. Stuebing and has profited muchfrom short visits to the laboratories at Edgewood (MD) and thefacilities at China Lake (CA), where he had contact with Mr. F.Davis from the Naval Weapons Center.
With regard to this broad area, which indeed should becovered by the literary survey, the author has chosen to-dividethe main subject into many subareas, each having a special number.The references are sectioned so that the reader can detach themto produce his own future documentation. The key words on thequotation cards are written subjectively; however, the' authorhas attempted to inform the reader briefly about the points froma specific article or book which are related to the goal of0.thisstudy. The reader will complete many of these notes, rearrange
them, and thus achieve, in the author's opinion, the mission ofany literary survey: to inform the reader about the potentialimportance of any written document and to indicate the relation-
ship of specific works.
The author has tried to approach the subject as broadly andas deeply as possible. In spite of his good intentions, he knowsthat on many points the study should be more specific and system-atic and, as a matter of fact, would like to expand some areas'of this study in the future. Also, he feels that there is noguideline from what point the historical notes on the subject's
3
-j
~A'~'{'~
evolution should begin to be mentioned. The author has inten-tionally made several historical notes because he is convinced
that in the era of data processing and huge modeling efforts,one should not forget that new ideas, even if yet not mathemat-ically well-founded and not supported by modern and expensivetechnology, still can help find ways to solve new and little-known problems such as clearing of military smokes.
Josef Podzimek
"> Rolla, Missouri
October 8, 1978
NTIS G1r"IDC TA3
Justifioatto .
.!-igability Codes&Avail and/or
.:t. special
4
-... ~' i
PREFACE
The work described in this report was conducted under pro-
.icct/task DAAK-1I-78-C-0001, entitled "Survey of Techniques for
Clearing Military Smoke Clouds". It was carried out from Scp-
tember 1977 to September 1978.
Reproduction of this document in whole or in part is pro-hibited except with permission of Commander/Director, Chemical
Systems Laboratory, Attn: DRDAR-CLJ-R, Aberdeen Proving Ground,
MD 21010. However, the Defense Documentation Center and the NationalTechnical Information Service are authorized to reproduce the documentfor United States Government purposes.
ACKNOWLEDGEMENT
The author is indebted to Dr. E. Stuebing of the U. S. ArmyArmament Research and Development Command Research Division,Aberdeen Proving Ground, and to his staff for much valuable in-formation. During this study the author visited Naval WeaponsCenter at China Lake where he had very useful discussions withMr. F. Davis and Dr. E. E. Hindman, II and the Naval Environ-mental Prediction Research Facility at Monterey (Dr. A. Weinsteinand A. Goroch). He is obliged to all of them for the friendly
exchange of opinion and information.Many friends and colleagues at home and abroad helped the
author formulate some conclusions and collect the literature andinformation about rare reports and patents. He is expeciallygrateful to Prof. Dr. Bricard (Paris), Prof. Dr. E. Hesstvedt(Oslo), Prof. Dr. H. Hinzpeter (Hamburg), Prof. Dr. Ch. Junge(Mainz), Dr. G. Madeleine (Paris), Prof. Dr. 0. Preining (Vienna),
Dr. R. Reiter (Garmish-Partenkirchen), Dr. Roach (Bracknell),Dr. P. Ryder (Bracknell), Dr. R. Serpolay (Clermont-Ferrand),Prof. Dr. Schumann (Heidelberg) and Prof. Dr. R. G. Soulage(Clermont-Ferrand).
The author enjoyed very much the friendly atmosphere and
5
Ui.
the very valuable discussions with Drs. Clipson, Jarvis, andJones at the Chemical Defense Experimental Establishment, Porton,and at the Amt fur Wehrgeophysik at Traber-Trabach with Drs.Uhlig, Reiss, Aufm Kampe and others.
Drs. B. Vonnegut and J. Jiusto from the AtmosphericSciences Research Center, SUNY at Albany, supplied the authorwith several rare reports and information on the extensive workin warm-fog seeding in our country.
Last but not least the author is obliged to his fellowworkers, Drs. J. Schmitt and P. Yue for reviewing the narrativeportion of this report and for valuable advice. Mrs. V. Maplesand Mrs. C. Turek, secretaries, and Mr. G. Frick, graduatestudent, of the Graduate Center for Cloud Physics Research,ably supported the author in preparation of this report.
6
CONTENTS
Page
I. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . .. . 9
2. PHYSICAL AND CHEMICAL PROPERTIES OF MILITARY SMOKE CLOUDS 12
3. SMOKE CLOUDS AND THE ENVRIONMENT .......... ............... 20
4. PRINCIPLES OF CLEARING FOG AND SMOKE CLOUDS ...... ........... 27
5. FIELD EXPERIMENTS ............... ......................... 62
6. RECOMMENDATIONS FOR FUTURE RESEARCH ..... .............. ... 65
7. SUMMARY ............... .............................. ... 67
APPENDIX, References, Articles, Reports ........ .................. ... 71
DISTRIBUTION LIbT .............. ......................... ... 199
-I7
*
i7
Ii '
4 . .\ ,
SURVEY OF TECHNIQUES FOR CLEARING MILITARY SMOKE CLOUDS
1. INTRODUCTION
The problem of clearing military smokes is only briefly
treated in the open literature. It has much in common with
dispersion of warm fogs or clearing of industrial smoke cloudsby the methodologies of increasing the visual range by droplet
evaporation, removal of particulates by impaction, coagulation
or scavenging. However, it has some specific features which
makes the task difficult. For example, in attempting to clear
a smoke cloud of highly hygroscopic substances in the free
atmosphere with its large water vapor content, the heat and
mass exchange will depend largely on the state of the terrain.
The clearing of military smokes above the terrain has many
facets which are related to specific questions such as how
large is the smoke-covered area, what is the vertical extent of
the cloud, and what are the important meteorological parameters?
In addition, one needs information about the physico-chemical
nature of smoke particulates and about the smoke microstructure.
Disregarding for the moment these specific questions, one can
attempt a systematic investigation by analyzing the necessary
conditions for improving the visual range in a cloud.
Visual range (Z) is usually calculated with the formula
lnS- (1)
7rn of(r)r2F(-2r-)dr
where c is the sensitivity threshold of the sensor (for the eye
e •0.02), n is the total droplet concentration per cm' in the
cloud, f(r) is the droplet size distribution function, r is the
droplet radius and F is the characteristic oscillating function
of the intensity of light scattered by the fog droplets and is
a function of 2- where X is the wavelength of the light. Usinga mean value FT• for the integral and the value for liquid water
9
content of a cloud (in droplet form) q = ' pp n r3 is the
density of fog elements) eq. (1) can be written
4p rT Inp= L (2)3q FT
Assuming an approximate value of F = 2.0, which is used forobservations in visible light by the naked eye (c = 0.02), the
visual range for a monodispersed cloud (F = r and =p 1.0 g cm 3 )
is
.= 3.4685 n-1/3q-2/3 • (3)
All quantities are in cgs units. The droplet concentration nand the liquid water content q are quantities usually measured
before an interference into the fog development is considered.In the case of a polydispersed fog (cloud) one notes that a fewlarge drops change q considerably but make no significant change
in n.
Formula (3) shows clearly that in the case of a constantlight source (characterized by a specific wavelength X) and afixed sensor (e.g., c = 0.02) the improvement in visibility canbe achieved by decreasing the droplet concentration or by decreas-ing the droplet size (which affects the liquid water content q).
This simple one-dimensional model requires, however, refinement
if one deals with a polydispersed aerosol, i.e., the integral in
the denominator of eq. (1) has to be evaluated rigorously (with
a different light scattering function F(+) for each size inter-val). Specifically, for a long wave light source one would needto include absorption in the treatment of the radiation trans-
versing the fog. In addition, in the transition region betweenlong and short wavelengths special treatment will be required.
S , This treatise will exclude all qUestions related to theimprovement of visibility in fog by optical means. However, it
will consider all other factors in eq. (1), including the changeof droplet spectrum, which is of potential importance. Thesefactors should be measured, calculated or estimated from a suitable
*10
-', • .. ... / . . . . .. . ... . . . . . . . . . . ., . . ••
model. Also, the simplified eq. (3) helps us categorize many
of the laboratory experiments and field trials to be reviewed
in the following chapters.
The simplest way to increase the visual range is to remove
particulates by air filtration and other means. In addition,
improvements in visibility can be achieved by smoke or fog cloud
dilution (mixing with clear air), by droplet evaporation, by
particle coagulation and by settling of large particle,.
Several of the methods applied in the past represent complexprocesses in which several elementary processes play importantroles. For example, fog droplet evaporation by heating of the
foggy air parcel inevitnbly stimulates air circulation, upward
transport of fog elements, and dilution of ground fog. The
treatment of fog droplets by methods which promote coagulationand fallout of particulates generates a slight downdraft. Aparametrization of individual terms in equations describing com-plex processes is a very expedient, but often tedious task.Serious difficulties are often encountered in the specificphysico-chemical nature of the smoke particles.
In enumerating the factors influencing the behavior and/orclearing of military smoke clouds, one has to stress the impor-tance of the environment. The state of the atmosphere and ofthe ground may play a decisive role in fog development. However,one has the impression while reading the many reports on fieldexperiments, that this fact has been disregarded, very often foreconomical reasons.
Realizing the shortness of the report and the broad fieldof endeavor that it must survey, the author feels that it is mostuseful to describe first the state of the subject based onavailable and unclassified articles and reports, and then to
present a survey of the literature including only some key words
mainly in the foreign reports to help the reader orient himself.Some of the references are without key words which means thatthe author has not had access to the paper or its abstract, buthe thinks that it might be important for other investigators.
The state of the subject represents an interpretation of
11
2..[ r
other investigators' work by the author, sometimes accompanied
by simple calculations and other checking of conclusions. Inthe following chapters we will discuss the Physical and Chemical
Properties of Military Smoke Clouds (Chapter 2), Smoke Cloudsand the Environment (Chapter 3), Principles of Clearing Fog andSmoke Clouds (Chapter 4), Field Experiments (Chapter 5) andRecommendations for Future Research (Chapter 6).
2. PHYSICAL AND CHEMICAL PROPERTIES OF MILITARY SMOKE CLOUDS
By physical properties of a military fog or smoke cloud,one refers mainly to its microstructure. This includes the con-centration of particulates, the size distribution, and the opti-cal and electrical properties. Of very great importance is thephysico-chemical nature of particulates: composition, surfaceproperties, solubility and behavior in the variable humidity ofthe atmospheric ground layer. Many of these parameters are notwell known even for the most common screening smokes or fogs.
The nature of natural fogs will be included for comparisonand because one cannot exclude the influence of warm or super-cooled water droplets in producing military fogs. It is wellknown that fogs in different localities and under differentmeteorological conditions have different microstructures (e.g.,Mason, 1957 or Podzimek, 1959) according to collected data fromdifferent authors on the mean size of fog droplets measured onthe seashore and inland. Hougton and Radford (1938) found largemean drop radii (5 to 10 um) on the seashore and small concen-trations (often below 50 cm- 3 ). Hagemann (1936) found that themean radius of fog drops collected in Germany varied from 4.5
to 17.0 Um. Smaller mean radii of fog drops were measured byKrasikov and Chikirova (1956) in the USSR (often r < 3 pm).However, the latter found much larger mean radii for dropletsin fogs generated by the evaporation of water from large waterbodies. Very small mean radii were found in fogs in the Oslo
(Norway) area (r < 1.0 pm).The liquid water content of fogs has been found to be very
12
,% V ,
small in comparison with clouds. Radford (1938) found values
between 0.1 and 0.22 gm"'. In Japan, Kuroiwa and Kinosita (1953)
measured values around 0.4 gmn3. Corin et al (1974) workedwith new data collected in the U.S.A. These data are based on
the measurements performed by Pili6 (1966), Kunkel (1970) and
Rogers et al (1972). Taken together, these authors found that
droplet diameters are .close to 10.0 pm in a radiation fog, 13.0-20.0.pm in an advection fog and 32.0 pm in a dense marine fog.However, large fluctuations in the mean droplet size, dropletconcentration and liquid water content were observed. Concen-
tration was much higher for a radiation fog (200 cm- 3) than for
an advection fog (100-150 cM- 3 ). A dense marine fog featured
an extremely low drop concentration (20-40 cm-') and relatively
high liquid water content (0.5 gm 3 ). The mean liquid watercontent of a radiation fog was 0.11 gm-' and that of an advection
fog was between 0.17 and 0.30 gM" 3. The thickness of a radiationfog varied from several meters to 300 m and in the case of an
advection fog it was up to 600 m.Recent measurements at a television transmitter tower in
the San Francisco area (Goodman, 1976) indicated smaller meandiameters of fog droplets (4.5-8.3 um) and concentrations rang-ing from 126 to 260 ci"3 . Due to the limited capability of the'sampler in collecting large fog droplets, the measured liquidwater contents in this series of measurements were much lower
than those communicated by other authors (5.3 x 10-3 gm-i to
6.07 x 10-2 gm- 3 ). Relatively low liquid water contents were
also measured by the Calspan Corp. group in a "pure" marine fogoff the coast of Nova Scotia (Mack et al, 1975). Liquid water
contents ranged from 0.005 gm 3 (at a visibility of 5,000 m) upto 0.489 gm- 3 (at a visibility of 50 m). At the same time the
mean droplet radius varied between 4.1 to 10.7 um.Many of the physico-chemical properties of the military
screening smokes and fogs are largely unknown, or the reportsare classified and are not available to this author. However,
general classification of screening smokes is found in theThorpes Applied Dictionary of Chemistry, Vol. X, pp. 781-791.
13
_._ . .-_
The author (K. F. Sawyer) divides screening smokes into Carbon
smokes, Oil smokes, and Hygroscopic smokes. The first kind,
which is generated by an incomplete combustion of fuel oil, is
now obsolete and has been superseded by other techniques which
produce better screening. Oil smokes are widely used in small
scale experiments (e.g., smoke generators in aerodynamic wind
tunnels for tracing air motion in the boundary layer of the
atmosphere, etc.) and for large area screens. The oil isusually burned with a deficiency of air and the generated smoke
consists of two different components: micron sized oil droplets
and larger carbon particles in the ratio 5-20 parts to 1. Better
screening performance is reached by completely eliminating the
carbon fraction, thus producing a white smoke which is a mixture
of oil and water vapor condensation products. The mean size of
particulates is between 0.5 pm and several Um. The size spectrum
is relatively narrow. Large generators burn 100 gallons of oil
per hour and successful screenings have been made by spraying
oil into the exhaust pipes of aircraft engines.
Hygroscopic smokes have been extensively studied for several
decades for their efficiency as screening smokes and due to their
operational and economical advantages. The dispersed hygroscopic
particulates become condensation nuclei and the ratio of the
final drop mass to the initial nucleus mass (yield factor) amounts
to several units depending on the environmental humidity. Athumidities larger than 90% the yield factor usually surpasses
10. Phosphorus, oleum of sulfuric acid, oleum combined withI chlorosulfonic acid, chlorosulfonic acid with dimethylsulfate,
several metallic chlorides, "Berger Mixture" (zinc dust and
carbontetrachloride with zinc and kieselguhr as absorbents) and
several others of similar composition to the Berger Mixture
(e.g., HCE or CTC) have been widely used. Also, several types
of alkyl metal compounds are often mentioned. Some of these
materials are described in monographs on particulate clouds
H• (e.g., Green and Lane, 1957) in handbooks on applied chemistry
(e.g., Green: Smoke; chapter in Hermans, 1953, 344-381), in
Army manuals (e.g., TM3-215 AFM 355-7, Military Chemistry and
14C 1€
Chemical Agents, August 1956; TM3-300, DATM, Ground ChemicalMunitions, August; NWIP 1-2 Smoke Screen Manual; NAVWEPS OP3142,
Characteristic of Biological and Chemical Munitions and DeliverySystems, White Oak, Mar. US NOL, January 1963; TM 3-500 Chemical
Corp• Equipment Data Sheets, DA TM, Headquarters, DA, April 1961)
and in several reports published from scientific meetings (e.g.,ACC Symposium on Aerosols, Symp. VIII., Vol. 1, C.W.L. Spec.
Publ. No. 2, July 1958, 158 pp.). Several studies and reports
have been published on different aspects of the application ofscreening smokes of hygroscopic particles (e.g., Rodebush et al,
O.S.R.D. Report No. 940, 1942; Kabrich, 194S; Axford et al, 1958;
Defense Document Center for Scient. & Tech. Inform. DSA, Cameron
Station, Alexandria, 1970; Rubel, 1978).
The most important factor is the dependence of the screen-ing smoke particle size on the nature of the agent and the en-
vironmental humidity. Sawyer (1942) investigated the behavior
of titanium chloride smoke which in the presence of water vaporproduces a dense, stable smoke. The structure of the embryonicnuclei is largely unknown, especially at low environmental
moisture. The author assumes that the stable aerosol has a
composition of hydrates (presumably TiC1 4 .5H 2 0); however, its
structure changes at relative humidities higher than 50%.A very detailed model of the growth of phosphorus smoke
particles was presented recently by Rubel (1978). It is basedon the "classical" approach to the description of the phosphoric
acid droplet growth which includes the Kelvin and solute term
and assumes equilibrium water vapor pressure between the growing
droplet and its environment. The quasi-steady approach is
justified by calculating the rclaxation times of the growing
droplets containing phosphoric acid (which holds strictly for
the environment close to equilibrium on the drop growth curve).
Also, the calculation of the growth rates of individual droplets
seems to be justified except in the case of a highly polydis-
perse aerosol with a high droplet concentration. The authorfound a simple relationship between the size of growing phos-
phoric acid droplets and the environmental humidity which
15
'1~*Al
varied between 10% and 98%. The model phosphorus smoke was
characterized by the initial radii of condensation nuclei
ranging from 0.30 to 2.5 um (0.30 x 10"15 to 0,14 x 10-12 moles).
Yield factors were strongly dependent on the relative humidityso that a relative humidity of 10% corresponded to 3.89 and 90%
corresponded to 16.29. This suggests a strong change in the
visibility range close to the ground where the humidity gradientand its time variation are often very large.
There are not many data relating to the number of mass con-
centration of screening smoke particulates in the atmosphere.
Usually several tens to several hundreds of particulates per
cm are measured shortly after the particulates reached thermal
and mass growth equilibrium with the environment.
Optical properties of the fog or smoke elements represent
a field of activity where many important questions remain
unanswered. The situation seems to be simpler in the case of
particulates with permanently homogeneous structure such as
pure water droplets or metallic powders (e.g., Al) used as
screening smokes. In the latter case, the nonspherical shape
and the helicoidal motion of larger particulates present prob-
lems for theoretical and experimental treatment. The hygroscopic
particles in the atmosphere, on the other hand, were recently
investigated not only for their importance in cloud and fog
forming processes, but also for their role in the propagation
of light signals both in the visible and infrared domain (e.g.,
Junge, 1963, p. 141-146; Wells et al, 1976). Some of the most
important parameters include an appropriate model size distri-
bution of particulates and the knowledge of the influence of
environmental (meteorological) factors on the deformation of
the size distribution function and on the change in light
scattering or extinction in the fog or smoke cloud.
Several authors prefer Junge's distribution for atmospheric
aerosoL (e.g., Junge, 1963, p. 118) mainly for its simplicity.
For the same reason many authors have used log-normal size
distribution. The gamma distribution, in its general form
(e.g., Podzimek and Saad, 1974) or Deirmendjian's (e.g., Wells
16
i_____I ._ ..
• •. • : • : • .• .- ' -.•7 - -T ii • • r i VL.• ± • •.- -•, -. ~-- . .. ., ... -. •
et al, 1976) functions also have great potential. This type ofdistribution seems to be very suitable for screening-smoke
particle distribution in spite of the fact that during thescreening smoke operation the ground layer of the atmosphere
will be pl.luted by other particulates both soluble and insolublein water. Different sources of pollutants and contaminants would
suggest a multimode distribution rather than the single mode.
Every effort should be made to establish model situations fora typical terrain and mean meteorological parameters.
The changing of the size distribution spectrum of smoke
p-rticulates is caused mainly by particle settling, coagulation
ind, in the case of hygroscopic nuclei, by the condensation of
water vapor. Excepting the previously mentioned study of
phosphorus aerosol by Rubel (1978), several investigators paid
attention to the size change of atmospheric nuclei (e.g., sea
salt particles) at various environmental humidities. The
results of many of these studies can be applied to the clearing
of screening-smoke clouds in which the smoke particles are
hygroscopic. Fitzgerald (1975) found a one-to-one correspondencebetween "wet" and "dry" salt particle radii except in the domain
of known "hysteresis" in the drop growth curve. The different
behavior of the growing and diminishing salt particle could
explain the findings of several authors (e.g., Podzimek, 1977)
and why many of the giant sea salt nuclei exist in the form of
solution drops even at a relative humidity of 55%. There is,
however, a discrepancy between these findings and the statementby Sawyer (Thorpes Applied Dictionary in Chemistry, p. 783).
He found that "the same quantity of water" is absorbed by a unit
weight of titanium tetrachloride particles whatever the relative
humidity.
The propagation of optical signals in haze, fog or smokedepends on the growth rate of particulates and is dependent on
relative humidity. Attempts to calculate the visual range infog have been made by several investigators (Dickson and Hales,
1963; Kasten, 1969, H~nel, 1968; Prishivalko and Astafyeva, 1974;
Zuev et al, 1973). H~nel's study (1976) showed that the
17I!
hysteresis in hygroscopic particle growth curve, and the corre-
sponding change in the complex refractive index depend not only
on the nature of a particle, but also on the whole history of
particle growth. Fischer and H~nel (1972) found the real part
of the index of refraction of atmospheric aerosol is between1.55 and 1.35 and the imaginary part is between 0.047 and 0.003
for X - 0.5 Wm. Fischer (1970, 1976) concluded from his
measurements that under normal atmospheric conditions one canignore the absorption property of atmospheric aerosol for the
wavelength used. An exact calculation of the refractive index
was replaced later by an approximate procedure (H~nel andDlugi, 1976) which might be useful for some of the screening
smoke models. 5
Much attention has been paid during the last two decadesto the optical properties of the aerosol particles in the infra-
red. Measurements of the absorptive power of atmosphericaerosol (imaginary part of the refractive index) were performed
by Irving and Pollak (1968). Remsberg (1971) and Volz (1972)Jfound that the extinction of natural aerosol has a minimum
around 8 pm followed by a strong maximum near 9 Vm. The same
author also analyzed the optical properties of composite aerosol(e.g.. incomplete dissolved salt). Several important contribu-
tions in the study of the optical properties of aqueous solutionsof electrolytes in the infrared domain were made by Querry (1972)0
Hale and Querry (1973), Querry et al (1974) and Rhine et al
(1974). The use of infrared transmitting and light scattering
(1i techniques for droplet size distribution measurement in a fogor smoke cloud seems to be promising (e.g., Eldridge, 1957,
1961, 1966).In summary, one concludes that visible light scattering
plays a much more important role in screening smokes than light
absorption. For this reason detailed information about the size
distribution, the composition, and eventually the shape of the
particulates is necessary. In spite of the unsymmetrical dis-
tribution of the scattered light about the particle the approx-
imation F(-4 ) ' 2.0 (see Introduction) for visible light is
18
I - .-. ,j• -
appropriate in many cases. A detailed investigation of the
effective scattering cross section, which varies between wr2
and 2wr 2 , was made by Sinclair (1947). Many physical and
physiological factors, however, are involved in calculating or
estimating the obscuring effect of a screening smoke might not
be related to the smoke cloud microstructure, such as the color
of the target, the background, etc. (e.g., Horvath and Presle,
1978). Similar problems will not be discussed in this study
in spite of their great importance.
Very little has been found on the electrical charge of the
generated screening smoke particles and its influence on the
colloidal stability of the cloud. Sinclair (1950) reported an
almost neutral nature of a homogeneous oleic acid fog produced
by electric sparks. Only 5% of the particles were charged
(mainly positively). However, more than 99% of the droplets
can be charged by a direct current corona discharge producing
droplets charged with 25 to 50 electrons. Dilution of the
aerosol cloud with clean air and an increase in air humidity
have no effect on the charge distribution. One expects that
in the free atmosphere, electrical charges on smoke particulates
will not considerably influence the colloidal stability of a
smoke cloud due to the relatively high concentration of ions
(>1,000 cm') in the ground layer over continents.
The physico-chemical properties of screening smoke particles
deduced from the laboratory experiments will be quite different
when investigated in the real atmosphere which is characterized
by a large temperature lapse rate, by a steep gradient of humidity
and by strong wind shear close to the ground. The mean values
and variances of particle concentration and particle size reflect
very often the turbulent nature of the atmosphere. Because the
persistence or clearing of a smoke cloud is largely dependent
on these elements, an analysis of the interaction of the meteo-
rological parameters with fog or smoke is appropriate.
19
1 .. . .. . .F-I _ LZ -" : - ,
3. SMOKE CLOUDS AND THE ENVIRONMENT
The behavior in the environment of a military smoke cloud
has much in common with natural fog. Common to both cases is
the dependence on temperature field and air stability, on pros-
sure or air parcel density (related to the pressure, tempera-
ture and substance content in a unit volume) and on the wind
vector. The wind vector and its variability in space and withtime represents one of the most important factors in judgingthe deepening or the dissipation of a fog or smoke cloud. Tur-
bulent exchange in the boundary layer of momentum, mass and
energy influences the eddy diffusion of smog or fog parcels andis closely related to the radiative transfer 4n the atmospheric
boundary layer and to the heat exchange with the soil. However,
if one attempts to change the microstructure of a smoke cloud
or to clear it from the environment, several points about its
behavior require special treatment.
Different, mainly in the initial stage, is the microstruc-
ture of a fog or smoke cloud (concentration, cize distribution,shape and composition of particulates). Different is the inter-action with atmospheric humidity (smoke particulates can be
hygrophobic, insoluble in water or highly hygroscopic) and in
general their physical properties are unlike (optical index ofrefraction, dielectric constant and surface properties). In
general, smoke particles located well in interior of the cloud
are less sensitive to the slight but very important changes ofatmospheric parameters sufficient for natural fog creation or
dissipation. One might expect that to clear military smokewould require a large amount of energy; however, it should be
remembered that the smoke is concentrated in a relatively small,well defined space and is a substance of known physico-chemical
properties. Different substances will require special techniques.Below will be discussed the interaction between a smoke cloud
and the environment, setting aside the very different properties
between a fog and a smoke cloud.
Much attention has been given during the last three decades
20
to air pollution and to the propagation of pollutants in the
atmospheric boundary layer. Of interest is the fact that a
strong impulse and physical basis for the development of math-
ematical models came from the studies of the behavior of mili-
tary smokes. The interpretation or misinterpretation of Sutton's
formulas (1949, 1953) together with several important studies
made the Chemical Defense Experimental Establishment at Porton
in England formed a backbone for the future models. The funda-
mental observations and measurements in the atmospheric boundary
layer were made by Lettau (1939, 1956) and Sutton (1949, 1953)
and later used in several books dealing with atmospheric bound-
ary layer as a subject of atmospheric physics (e.g., Pasquill,
1962; Pristley, 1959; Laichtman, 1970; Schmeter, 1972) or applied
in an environment where pollutants propagate (e.g., Berlyand,
1975; Scorer, 1968). Besides these monographs many reports and
articles on similar subjects appeared (a survey of many studies
can be found in Turbulent Diffusion in Environmental Pollution,Proc. Sympos. IUTAM & IUGG, Charlottesville, Virginia, April
1973, in Advances in Geophys., Acad. Press, New York, Vol. 18
A & 18 B). Only very few of these studies, however, deal with
the specific problems of fog stability and dissipation. Several
articles published in the USSR (Berlyand and Onikul, 1968 a;Berlyand et al, 1968 b; Berlyand and Kurebin, 1969) and in the
U.S.A. (e.g., Corrin et al, 1974) reveal the difficulties of
calculating the diffusion in a calm situation or in a fog.
Measurements of the diffusion at night and theoretical modelingof the diffusion of pollutants at night are still in a rudimen-
tary stage (see for instance Deardorff, 1978). In spite of these
difficulties progress has been made in investigating the behavior
of a smoke or fog cloud in the boundary layer during the last
three decades.
Of interest is the fact that the old measurements and
theoretical investigations of air stability by Richardson (1920 a,
1920 b) are still very useful for both stable and unstablestratifications (e.g., Hanna, 1978). However, the basic value
of the Richardson number Rio which characterizes the amplification
21
(Ri < Ri ) or damping (Ri > Ri ) of a small perturbation is
largely unknown. In a shear flow having a moan velocity u (z)
and at a temperature lapse rate the Richardson number is a
nondimensional parameteraT a7
Ri = •-(4)
T~ (ni/a Z)
For instance Taylor assumes Rio = 0.25, Richardson Rio = 1.00
and Prandtl Ri = 2.00. Petterssen and Swinbank suggest Ri0 0.6500
and Schlichting found that Rio varies between 0.029 and 0.041
which is supported by Paeschke's measurements. More experimental
work and theoretical refinements will be needed before Richardson's
number (Ri) will be fully applicable even for very low wind
velocities and for temperature gradients far from adiabatic.
Rossby and Montgomery (1935) applied in their models a mixing
length
,. kz (5)
/1 + aRi
where k is Karman's constant and a is a parameter which some
authors found to be close to 11. A slightly different expres-
sion was found by Holzman (1943)
2 - kz V1 - a Ri (6)
where ao = 7.0 in accordance with the measurements by Deacon
(1949). However, at large temperature gradients these models,
useful for the application of the mixing length hypothesis,
showed a poor agreement with the measurement. An interesting
investigation of a stable boundary layer was presented by
Businger and Arya (1974).
The mixing length hypothesis, once so useful for a simple
description of velocity field, power-law profiles and eddy
viscosity (K - = i s" (w2; w' = vertical component of
turbulent velocity) assumes a new face in the statistical theory
of turbulence. There the turbulent velocities (and their
anisotrophy) changing with time are related to the wind shear,
22
temperature gradient and roughness of the ground and are ex-
pressed in the form of statistical relationships valid for a
certain situation and time period (e.g., Monin and Yaglom, 1973.and 1975). This approach has proven to be very useful due to
the improved ability of instruments to measure the turbulent
fluctuations of meteorological parameters in the boundary layer.
The approach has been widely exploited in the modeling of thetransport of momentum, energy and mass and has been most pro-ductive in the treatment of isotropic turbulence. However, the
requirement of quasi-homogeneity and quasi-stationarity restrictsthe general application of many results to the boundary layerclose to the ground. For this reason, Pasquill (1974) suggestsdividing the polluted atmosphere into three main layers: 1) ashallow surface layer containing the pollution from nearbysources where the concentration of pollutants is strongly depend-
ent upon heating or cooling of the surface, 2) a layer of nearuniform vertical distribution of pollutants above the surfacelayer and 3) a "free atmosphere" containing the backgroundpollution from distant sources or from large scale ascendingmotion. The statistical parameters (e.g., Lagrangian correlationcoefficient R[4] related to the time lag ý) will be differentfor each layer and are expressed in the form
R(C) = exp (-p;) cos (qC). (7)
From this an integral time scale can be deduced
tL = + q2) (8)
which, also, enters the expression for eddy viscosity in the form
Kmat = a~ 2 1 A 1/3 X4/ (9)w = L 1 w Xm ,/ M/.
Where X m is the "equivalent wavelength" of the peak of the spec-
tral energy curve for the w velocity component and c is the rateof dissipation turbulent kinetic energy. aw is the standard devi-ation of the w-velocity component from its mean value. A slightly
23
different numerical factor was used by Hanna (1978) for eddy
diffusivity: Kz = 0.15 w Am' Much attention has been given
recently to the variation of the eddy viscosity and diffusivity
with the altitude and with stability conditions (e.g., Pasquill,
1974; Hanna, 1978). Pasquill published the dependence of Km on
altitude z for stable, neutral and unstable conditions. Hanna
mentioned an expression for Kz valid for the surface layer:
zzKz 0.35 ux Z¢ (z/L)] (10)
where ux is the friction speed, L is the Monin-Obukhov length
and *h is a universal function of z/L which was empirically
deduced. For unstable conditions
ýh (z/L) = 0.74 (1-9 z/L)-/2 (lla)
and for stable stratification
*h (z/L) 0.74 + 4.7 z/L (llb)
is recommended. Kaimal et al (1976) deduced formulas where mis a function of z/L
7•, "0 < z < /L/ (12)m 0.55 - 0.38 z/L
X = 5.9 z /L/ < z < 0.1 h (13)m2
X 1.5 h [I - ekp (-5 z/h)] 0.1 h < z < h. (14)m3
The usefulness of similar formulas has to be checked in the
future, but the importance of the measurement of the vertical
components of turbulent velocity and of the turbulent energy
spectrum is apparent. Recent measurements in different parts
of the U.S.A. (Hanna, 1978) show already that the depth of themixed layer h is an important factor which changes during the
daytime and when combined with the surface heat flux Q deter-mines the scaling velocity wx = (Qh)1/3. Scaling velocity wx
24
and the depth h determine the state of the boundary layer. These
new findings in boundary layer theory influence the diffusion
model of propagation of pollutants or smoke particles. Very
beneficial would be the knowledge of some of the microstructural
characteristics of the smoke. That information, combined with
the features of the terrain, could lead to a useful forecast of
the smoke behavior or prediction of its clearing. Very often
mesoscale meteorological processes will considerably influence
smoke microstructure (solar radiation, washout and fallout).
Some advice on these interactions can be found in manuals for
forecasting air pollution (e.g., WMO Technical Note No. 121).
Still unresolved are basic problems to the general applica-
tion of the statistical theory of turbulence to the diffusion of
pollutants and to the exchange in the boundary layer. Short-
comings of the gradient transport models in turbulence were dis-
cussed several years ago by Corrsin (1976) from the point of view
of random walk. The largest error seems to be related to the
inhomogeneities produced by the transporting mechanisms (scale
of turbulence and rms velocity).
There are basic difficulties in modeling air pollution. fromsources scattered over a large territory because of the different
density and intensity of sources and their orientation with re-.
gard to the wind direction and velocity. The classical (formal)
approach was to use a dense grid and to calculate for each point
* the concentration corresponding to main air trajectories (Bcrlyand,
1972).
Special treatment is required for the case of anomalousvariation of wind-velocity with height mainly for an elevated
temperature inversion and the dispersion of smoke particulates
in calm-wind conditions. Berlyand (1972) made several attempts
to describe similar anomalous propagation of pollutants with the
aid of empirical relationships and dimensional analysis. However,
the general validity of similar calculations is very limited.Several authors dealt with the propagation of pollutants over
an uneven terrain. In this case a new z-coordinate was usually
introduced z' z h(x) enabling one to describe the shape of
25
.0
the terrain (Berlyand, 1972). Berlyand et al (1968 b) used the
so called potential current method with a complex argument which
permitted conformal representation on the semiplane with curvi-
* linear boundary. This method was previously used by Stimke (1964,
1966) who, however, used constant coefficients in the diffusion
equation.
Another case requiring special treatment is that of atmo-
spheric diffusion in fog or a smoke cloud and the interaction of
* smoke (fog) elements with the environment. Pollutants (gases,
particulates) are absorbed or deposited on smoke (fog) particles,
the particle size is largely dependent on water vapor transport
and on droplet coagulation efficiency. On the other hand, all
parameters are dependent on the turbulent fluctuations of the
individual parameters. Unfortunately, very few measurements
have been made in this area. Typical data measured in fogs
(Corrin et al,. 1976) are: Shear of the mean wind 0 to
0.015 s-1, turbulent transport coefficient K - 0.1 - 10 2 m2 s'm
and the cooling rate for fog 2-T 1 - 30C h r-1. Very little
information is available about the turbulent intensity measured
in a fog or smoke cloud for a well defined meteorological
situation. Sedunov (1972, p. 92) assumed a situation in a layer
type cloud with a mean updraft velocity of w = 1 - 3 cm sec-
and estimated the influence of fluctuating velocity
[W-•T-]1/2 - 30 - 50 cm sec- 1 on the activation of hygroscopicand insoluble nuclei. For salt solution nuclei containing 20%
of solute by weight he found that the nuclei activation rate
was 4 sec- Cm 3 compared to l0-3 sec- 1 cm 3 for insoluble nuclei.
In both cases Jumiie's distribution of nuclei was assumed.
f New perspectives are opened in the author's opinion, by
the stochastic condensation theory. However, the mathematical
framework of this theory is still in the rudimentary stage (e.g.,
Sedunov, 1972; Clark and Hall, 1978). Clark and Hall calculated
e.g., the droplet size distribution from the equations including
the perturbation terms under the more genera' ' -ee-dimensional"deformation" term and assuming a n(n-varying a. ipation of
turbulent energy e - -100 cm 2 sec"1 corresponding to the values
26
/ ~<A
[TU1311'd-YT]ll 1/1 r *lr 1 46.6 and [(-r-r]1/2 - 69.1 cm sec".Another question which arises is how the microturbulence
or the fluctuations of microstructural parameters in the fog in-fluence the coagulation of fog droplets or smoke particulates.In a layer-type cloud Staffman and Turner (1956) found an insig-nificant effect due to small scale turbulence. However, Woods
et al (1972) and Jonas (1972) indi..ate that even normal shear inthe velocity field might contribute significantly to droplet col-
lision (experiments by Jonas and Goldsmith at shear greater than
" sec-1). In essence, this conclusion was supported by Tennekesand Woods (1973). The main problem is still the lack of system-atic measurements of meteorological parameters inside of a fog
close to the ground.Finally, it should be emphasized that the interaction of
the environment with a fog or smoke cloud is an extremely complexprocess. Successful clearing of military smoke requires a rapidestimate of the smoke extent, and its nature, and may require
a very fast measurement of the main meteorological elements suchas wind speed and direction, temperature and lapse rate, humidity,
visibility and possibly' solar radiation.
4. PRINCIPLES OF CLEARING FOG AND SMOKE CLOUDS
This survey of methods for clearing fog and smoke cloudswill be based mainly on the methodology developed for clearingnatural fogs. The reason is first, not many of the methods forclearing smoke clouds have been described in the open literatureand second, many of the principles applied to natural fogs aresuitable for clearing of smoke clouds.
Categorization by different techniques for natural fogdispersion has been made by several authors (e.g., Katchurin,1973). In the author's opinion, the most suitable seems to bedivisions based on the main physical processes leading to theimprovement of visibility in a smoke cloud. In this survey allmethods for changing the optical parameters of the light scatter-ing elements and the freezing of liquid elements will be omitted.
27
.- x'I
The following scheme is suggested;
Direct removal of smog and fog particles from the cloudby: a) filtration, b) sedimentation, c) phoretic forces,and d) condensation of vapors on nuclei.
Coagulation and subsequent sedimentation in: a) agravitational field, b) an electric field, and c) airacoustic field.
Evaporation of droplets through: a) heating of the fog(FIDO), b) combination of thermal and dynamical system
(TURBOCLAIR), c) mixing with dry air (dynamical method),d) absorption of solar radiation, and e) heating of thefoggy air by laser beam.
Dilution of the aerosol cloud through: a) mixing withclean air in horizontal direction, and b) mixing orparticle transport through artificial convection.Other techniques.
In the following the principles of different techniques
will be discussed and some estimates of their importance, based
on information in the available literature, will be mentioned.
4.1 Direct Removal of Smog and Fog Particles4.1.1 Filtration - The simplest method to remove the smog or fogparticles from the air is t" direct deposition of particulates
onto "eliminators". Devi _f this type consist of fine wire
meshes or air deflectors on which the foggy air is cleaned bythe higher kinetic energy of the droplets in comparison to theair parcel. Patents have been issued for the clearing of fogby blowing it through a set of fine rotating meshes (West GermanPat. 1135940, U. Smieschek), by depositing fog drops on deflec-tors (West German Pat. 1816733, U. Regehr) and by towing large
fine meshes behind a vehicle along the runways or highways (WestGerman Pat. 1909946, K. Wanders). A large number of similardevices have been suggested in the U.S.A. and many other coun-tries. However, the main problem seems to be the resistance to
the passage of air through the mesh or deflector. It has been
28
ii
estimated that 2,000 ml of air has to be moved through the system
per second (Hougton and Radford, 1938) for it to be effective
Also, most of the authors do not consider that in the case of
water drops the liquid or humidity must be removed from the col-
lector. For this reason, several inventors have suggested the
removal of fog droplets by cooling them down below the freezing
point in a system of jalousie-type deflectcrs (Austrian Pat.
166780, E. J. Millonig). One also notes that the efficiency ofmechanical separation of smoke particulates can be increased by
the incorporation of electrostatic filters (e.g., Austrian Pat.
305S48, Braun Aktien-Ges., Frankfurt a. M.).
The physical principles of the direct deposition of partic-
ulates on bodies of simple geometry are well known and a large
number of experiments on the filtration efficiency of different
materials have been described (e.g., books by Davies, 1973;
Davies, 1966, Fuchs, 1959). The practical disadvantages of
applying this method for field operations probably caused the
very pesimistic statement by Hougton and Radford (1938, p. 20):
"It must be concluded that although possible, thc method is
hardly applicable". However, this statement was not fully
supported by Prof. C. Junge during a private discussion with
the author. He believes that for small area fog clearing
experiments (such as were anticipated by one of the German
inventors) this method should not be rejected without careful
checking of its potential. The author of this report feels
that this might also be true with regard to clearing military
smokes obscuring small areas.
4.1.2 Sedimentation - A spherical particle 0.5 pm in diameter
and of unit density will settle in the atmosphere under calm
wind conditions at a rate of 6.8 x 10" cm sec" 1 . The settling
rate of a S.0 pm particle will be 5.0 x 10-1 cm sec". This
low settling rate can, however, be strongly influenced by air
turbulence in the ground layer and by the air advection. The
type of vegetation combined with the air turbulence plays a
decisive role in smoke particle deposition above the ground.
This was clearly demonstrated by the deposition of marine fog
29
droplets on the leaves of shrubs and trees (Oura and Horn, 19S3)
which generate an intense turbulent field in their wake. Another
interesting observation was made by Eichborn (1954) who observed
that the average behavior of aerosol particles above the terrain
is related to the time of day and to the solar radiation. He
found the sinking velocity of smoke particles in the early morn-
ing sunshine was 15 to 20 times greater than in cloudy weather
or at evening dusk.
Magono et al (1964) undertook with his fellow workers a
very interesting attempt at fog dispersion using the downward
air flow caused by the fall of water drops released from a heli-
copter 100 m above the fog layer. The size of water drops which
fell through the fog was selected carefully to attain the
maximum entrainment of the particulates in the wake of following
larger drops. The author of this report suggests that the use
of dry ice particles of several mm in diameter might enhance the
transport of smoke aerosol toward the ground. The wake effect
might be magnified in this case by the larger density of gaseous
CO2 .
The main importance of sedimentation processes is attributed,
however, to the increase of the settling rate of particulates
by condensation of vapors. This will be discussed later.
4.1.3 Phoretic Forces - Under phoretic forces one lists diffusio-
phoretic (a special case is Stefan flow), thermophoretic, photo-
phoretic and electrophoretic forces. Descriptions of the
mechanisms and estimates of the importance of individual cases
have been published, e.g., in the Davies book on Aerosol
Science (1966), by Hidy and Brock (1970), by Fuchs (1959) and
others. For application to clearing of military smokes one
has to consider two basic cases: 1) particles are collected
in a high gradient of a diffusing substance, a temperature
gradient or electric potential in a device which serves as a
precipitation zone through which the medium is passing, or
2) the high diffusive gradients promoting the smoke particle
deposition are generated around droplets which are growing by
vapor condensation process (or serving as heat sink).
30
S...1
F
As early as 1870 Tyndal discovered that aerosol particlesmove towards a body with a lower temperature and in 1887 Stefan
found that there is a "dust-free" zone around an evaporating
body. Since that time very few investigators have paid atten-
tion to this phenomenon. However, Stetter (1954) obtained a
patent for an arrangement whereby dust particles can be depos-
ited in a concentration--or temperature gradient. Facy (1955)
described a "dust-free" zone around an evaporating drop. The
theoretical explanation of the observed features started inde-
pendently in West Germany (the papers by Waldmann and Schmidt
are reviewed in Davies, 1966) and in the USSR. The Russian group
headed by Deryaguin published their results in several papers
(Deryaguin and Bakanov, 19S7; Bakanov and Deryaguin, 19S9;
Dukhin and Deryaguin, 1964; Deryaguin and Yalamov, 1971, and
Deryaguin et al, 1971). The most important result of these
studies was the difference of particle deposition rates for
ultrafine particulates (Molecular regime) and for low Kn numbers.
The other important finding is the difference in deposition
velocity between a moderately large volatile aerosol particle
and nonvolatile particle (Deryaguin and Yalamov, 1971; Deryaguin
et al, 1971). Schmidt's and later Goldsmith's experiments
(Davies, 1966) showed a reasonable agreement with the theories
predicted. The data indicate also the small deposition rate
of the particles subjected to the phoretical forces. Only inexceptional cases can one find a situation, when the phoretical
deposition might not be negligible in the atmosphere (e.g., in
mixed clouds as mentioned by Podzimek, 1965 and 1966 and during
the special case of particle scavenging reported by Slinn, 1976and 1968).
Further development of the theory of diffusiophoretical
forces covers the important transitional regime of particlesizes (e.g., Brock, 1968 or Anisis et al, 1973). However, a rough
calculation of the deposition velocity of particulates according
to Goldsmith's formula (Davies, 1966) given for diffusiophoresis
vd - -1.9 1 I0" [cm sec"1 ] (1)
4 '31
(p is the pressure in mb) or for thermophoresis
v AdT [ sec- 1 ] (16)
Note how small the deposition rates are if one considers thegradients existing in the atmosphere and if one assumes particulessmaller than 0.1 ur. For larger particulates one has to includea correction term. The formulas are based, however, on a quasi-steady situation around a growing or dimishing droplet whichmight not be applicable if the collector's size is changingrapidly. The influence of a fast growing collector should beinvestigated in more detail before a final judgement about theimportance of diffusiophoretic forces for the clearing of mili-tary smoke is made.
In accordance with some preliminary calculations of theinfluence of photophoretical forces on particle deposition oneconcludes that they are unimportant for the clearing of a fogor smoke cloud. For this estimate the author took the formulaspublished by Preining in Davies' book (1966) and assumed that afog is irradiated by a source of light or laser bean of nediumintensity. The collision increase among the droplets of 0.5and 5.0 us size is not significant. However, there is stillgreat uncertainty in the appropriateness of the formulas forphotophoretic forces.
The electrophoretic case will be discussed later in con-nection with electrostatic coagulation and electrical chargingof particulates. Strong limitations are imposed to the attemptsto reach a very high potential gradient in the atmosphere closeto the ground. Usually corona discharge above uneven terraincovered with vegetation makes the high charging of individualparticles impossible (private communication by Vonnegut).4.1.4 Condensation of Vapors on Nuclei - Basically one can dividethe physical processes used for clearing of fogs into two largegroups: Those operating at positive temperatures ('C) and thoseapplied at negative or freezing temperatures (below O°C). Thefirst division uses mainly the condensational growth of some
32
(usually highly hygroscopic) substances, the second uses direct
water vapor transport from supercooled water droplets onto ice
crystals (desublimation) or first enhances the freezing of
supercooled droplets by contact nucleation followed by sublima-
tion or coagulation process. This paragraph will discuss the
first process and only some of the basic ideas pertaining to
the second one.
Nucleation on hygroscopic nuclei with the resulting dropletgrowth has been discussed in many textbooks on cloud physics andphysical chemistry. Several interesting points related to the
application of the theory for clearing of natural fogs have been
made by Corrin et al (1974), and nucleation and droplet growth
on phosphorus smoke particles has been treated recently by
Rubel (1978).
Hygroscopic condensation nuclei such as NaCl, CaCl 2 , NH4 Cl
or droplets of solutions of H2S0 4 or H3 PO 4 are treated in thesame way as far as the progressive stages of growth are concerned.
They act, however, differently in the early stage of nucleation.One usually assumes the validity of Raoult's law for the whole
process of growing droplets, and one calculates the equilibrium
size of a solution droplet related to the environmental humidity.
A very challenging problem for a given nucleus is to calculate
the characteristic relaxation time needed to reach an equilibriumstate at a given humidity. This question has been analyzed in
detail by Sedunov (1972), be Carstens et al (1974) and for a
phosphoric acid droplet by Rubel (1978). Rubel used an accommo-
dation coefficient a w 1.0 and a condensation coefficient 8 - 0.5.He found relaxation times between 10-1 to 0.9 seconds for nuclei
sizes corresponding to those generated in phosphorus smokes andfor relative humidities ranging from 10% to 98%. Larger values
of relaxation time were obtained by Carstens for NaCl nucleiwith a condensation coefficient 8 * 0.035. In spite of the open
question of the value of the condensation coefficient it seems
reasonable to assume quasi-steady state for the model calculations.
With regard to the relationship between the smoke cloud or
fog microstructure and visibility mentioned in the Introduction,
33
--.-
one concludes that seeding with hygroscopic substances will con-tribute to the colloidal instability of a system. A few of thelarge drops will grow at the expense of the many tiny droplets.This effect combined with several changes of the surface proper-ties of the seeded drops and possibly combined with electricalcharge redistribution may contribute to the enhancement of
coagulation, faster drop growth and finally fallout of the large
drops. The visual range in a seeded fog is, however, stronglydependent not only on the droplet concentration but also on thedroplet size spectrum as a function of time (Saad et al, 1976).
In most of the studies outlining the methodology for hygro-scopic nuclei seeding (Nail, CaCl2) the authors (e.g., Hougtonand Radford, 1938; Stewart, 1958; Jiusto, 1964 a, 1964 b; Jiustoet al, 1968; Kocmond et al, 1968; Kocmond and Jiusto, 1968;
Kocmond and Pili6, 1969; Pili6, 1969; Kraght, 1969; Kornfeld,1970; Fedoseev, 1971; Serpolay and Andro, ]972) concluded thatthe amount of salt used for a successful experiment is not very
large. Hougton and Radford's calculation led to a spraying rateof 4 to S liters of saturated solution fCaCd 2 ) per second.Cornell Laboratory (Calspan) experiments indicated that 1.6 mgof dry NaCl per rn might cause a substantial improvement invisibility in a dense fog (Fig. 1). The recommended particlesizes for NaCl crystals varied between 5 to 35 um and the
perceivable effect was observed one to several minutes afterthe seeding. Katchurin (1973) reports a substantial improvementof the visibility in a large expansion chamber three minutesafter seeding the fog with hygroscopic nuclei of 1 to 5 um in
size. The improved visibility lasted more than ten minutes afterthe introduction of the seeding agent (NaCl). During systematicexperiments it was found that the change in humidity in the
600 m3 chamber was usually smaller than 1% R.H. (Jiusto et al,
1968), nevertheless it led to the evaporation of a large portionof droplets. The effect of seeding in nature is not always assatisfactory as it is in the laboratory. The only explanation
is that air mass exchange with the soil, cannot be simulated suc-
cessfully in a chamber. The results of some of the field
34
experiments with hygroscopic substances will be discussed later.Several other substances enhancing condensation of water
vapor have been mentioned such as the mixture of HC1 SO3 and
SO3 (McDonald et al, 1965), certain phosphates and polyelectro-
lytes (Kocmond, 1968). The latter substances (e.g., polyacry-
lamines) can swell enormously with the liquid water and a high
electric charge density on their surface could contribute to a
fast coagulation with other droplets and to the fast removal
of drops from the foggy air. However, preliminary tests of
polyelectrolytes on artificial fog clearing are not very
encouraging (Kocmond, 1969). Other active substances for warm-fog seeding have been investigated by Hindman and Clark (1972).
Warm-fog dispersal tests with glycerine did not yield
conclusive results (McDuff et al, 1973 a, b, c) and the use of
urea (Weinstein and Silverman, 1973) for the combination of
urea-bentonite (Depietri and Rosini, 1968) is still in rudimen-
tary stage. The use of water-absorbing ion-exchanging resins
for fog dispersion in the USSR was not recommended after the
laboratory experiments by Chikirova (1967) were completed.
Some positive effect was found using a powder of an alginic
acid as a warm-fog-seeding agent (Paugam and Serpolay, 1970;Maguet and Serpolay, 1973). The noncorrosive sodium alginate
powder is comparable with NaCl crystals as far as the activa-
tion threshold is concerned.
Another means to support the colloidal instability of a
warm fog is to use substances which are effective in preventing
the growth of drops. The idea of covering some droplets by
surfactants in order to enhance the growth of the uncovered
ones is, in the author's opinion, not yet well supported bytheoretical analysis (Deryaguin et al, 1960; Deryaguin and
Durgin, 1969; Deryaguin et al, 1971; Podzimek and Saad, 1975)
and Shiniaiev, 1968; Bigg et al, 1969; Leonov et al, 1969,Bakhanova ot al, 1969; Leonov and Prokhorov, 1967; Storozhilova,
1971; Kocmond et al, 1971, and Duguin and Stampfer, 1971) which
are inconclusive. The possible effect of surface active sub-
stances on droplet coagulation will be discussed later in more
35
S,*
104. . . ...I f
1%
lift)
10
,.', ~ ~102
0 10 20 30 t (min)
I-v
Fig. 1. Visibility £ [ft.] in an artificialfog as a function of time t [min.].Dashed curve is for the fog seededwith salt nuclei in a concentrationof 8 mg m", solid curve correspondsto the control expansion (withoutseeding). The upper figure was ob-tained from measurements at 15 ft.level and the lower curves wereplotted from the measurements abovethe bottom of the expansion chamber.Jiusto, et al (1968)
36
I
IIdetail. It appears that most of the authors support the ideathat the growth of a nucleus or a drop covered by surfactant is
only retarded by the surfactant layer and that after several
minutes the drop will reach the size of a drop which was not
covered by any surface active material. This result, however,is influenced greatly by the theoretical model which is usually
a quasi-steady approach with the assumption of a time-independent
diffusion coefficient. Also, the structural differences of
different surfactants some of which are soluble and some
insoluble in water should be stressed more in the author's
opinion.
The conclusion one draws for the clearing of military smoke
clouds from the warm-fog-seeding experiments is the following:
very useful relationships for the hygroscopic nuclei and solution
drop growth, applicable to atmospheric conditions, have been
obtained. However, a wide exploitation of the hygroscopic nuclei
seeding technique cannot be anticipated if it is not combined
with other mechanisms such as coalescence or phoretical forces.This is because military smokes consist mainly of highly hygro-
scopic substances. In the case of fine solid smoke particles
some chance exists for removal of them by hygroscopic particleseeding. The author favorably considers the possibility that
one can use some surfactants to cover part of the population of
the smoke particles (composed of highly hygroscopic substances)
in order to reach colloidal instability. However, a new theo-retical approach to this problem has to be developed andsystematic experiments mainly with insoluble surfactants will
be needed (Fig. 2).
The growth of ice crystals among supercooled water dropletsis very similar to the transfer of water vapor onto salt solution
droplets. The gradient of water vapor around the crystals is
maintained by the difference of vapor pressure over water and
ice which is maximum about -120C. Without going into manydetails one might note (aims with the possible importance of a
similar process in clearing military smokes) the fact that in
water vapor gradient particulates of insoluble and soluble
37
104
I ft 00
loa
0 20 40 60 t(min)
Fig. 2. Visibility t [ft.) as a functicn of time t (min.)for a control fog (full line) and for a fog seededwith 0.9 g of cetyl alcohol (dashed line). kocmond,et al (1972).
38
I t xI
substances are transported toward ice crystals (Stefan flow).
The model of water phase transition in a supercooled foguses the equation for water vapor balance inside of the fog,the growth rate of the ice crystal and evaporation of water
drops (e.g., Katchurin, 1973). From these equations three
unknowns (air humidity, size of the crystal and of the drops)
can be calculated as a function of time. Due to the number ofevaporating drops and large crystals which settle rapidly the
visibility of fog can considerably increase within 10 to 20
minutes depending on the concentration of ice nuclei or
nucleating agent. The main problem is to seed enough but not
to overseed the supercooled fog.
Several techniques have been introduced in field experiments
with the aim of transforming the supercooled fog into ice fog
(with better visibility). The most successful is dry ice seed-ing, used as early as 1931 by Veraart. Another technique is
the dispersion of liquid propane and the use of substances with
a high ice nucleating capability (AgI, PbT 2 , CuS, metaldehyde).
Supercooled fog seeding by dry ice was successfully per-
formed by many authors, e.g., by Aufm Kampe ct al (1957),
Gaivoronskii and Seregin (1962), Gaivoronskii et al (1965),
Beliaiev et al (1966), Rabbe (1969), Mfiller (1974), etc. An
analysis of the physical processes related to dry ice seeding.,was performed by Hindman (1966), Rabbe et al (1968) and Buikov
and Polovina (1969). ","
Several experiments with liquid propane spraying were per-formed in France (e.g., Olivier, 1956; Serpolay, 1959; Cot andSerpolay, 1961; Serpolay 1969; Andro and Serpolay, 1970), in
the U.S.A. (Gerdel, 1968; Kumai, 1969; Wise, 1975) and in other
countries. A detailed analysis of the physical processes relatedto the clearing of supercooled fogs by liquid propane spraying
Shas been made by Charry and Lininger (1975).
Since 1947 when Vonnegut introduced silver iodide intocloud physics as an ice nucleating agent, many investigatorshave applied it in supercooled fog seeding experiments. Surveys
of the physico-chemical properties ofAgI can be found in any
H 39
modern textbook on cloud physics. Balabanova (1960) showed that *1silver iodide acts mainly through contact nucleation in a super-
cooled fog because the number of generated ice crystals increased
only to a definite maximum concentration which is related to thenumber of supercooled droplets. Several experiments with silveriodide fog seeding were performed by Miller (1960), Nikandrov
(1962) and Sumin (1968). The latter analyzed the capability ofAgI and PbI 2 to convert a supercooled fog into an ice fog and.found a relationship between the rate of crystallization, the
temperature, the wind-speed and the propagation of the crystal-lization zone. The same author later published (Sumin, 1969) asimilar study on the use of CuS particles as ice nuclei. Rela-
tively few field applications are found for seeding withmetaldehyde (Fukuta, 1969).
In general, the seeding of supercooled fogs is one of thefew fields in which cloud physicists gained an economicallysignificant success. However, in the author's opinion, thepotential use of similar techniques for clearing of militarysmokes is very limited except in the case where ice crystalsare generated having articulated forms and thus have muchlarger scavenging efficiency than water drops.4.2 Coagulation4.2.1 Gravitational Coagulation (Scavenging)i - In the earlythirties the group headed by Wigand started to investigate thestability of fogs and the droplet coagulation in a gravitationalfield (Wigand and Frankenberger, 1930; Wigand, 1930; Frankenberger,1930, Findeisen, 1930). It was found that the fog droplet sizespectrum in a large chamber changes with time and that the ,coagulation of drops accelerates their growth and their removal
through settling.
Most of the important studies on gravitational coagulation
have been published after the memorable publication of Langmuir'sinvestigation into the collision efficiency of two fallingdroplets (Langmuir, 1948). They are treated in full details in
* the textbooks on cloud physics. For this reason, only fewremarks on this subject will be made.
40
.. ........_-------_--
Neiburger et al (1974) summarized the findings of manyauthors who calculated and measured the collision efficiency of
falling drops. He tried to explain the discrepancy in collision
efficiency for equally sized drops calculated by an analytical
formula and by the formula deduced by Shafrir and Neiburger which
postulates a zero collision efficiency for droplets of 30 um in
size. A similar result obtained originally by Hocking (1959),
was later corrected (Hocking and Jonas, 1970) and challenged by
Klett and Davis (1973), Lin and Lee (1975), de Almeida (1977)and others. Experimental verification of the theoretical
results was attempted by Woods and Mason (1965), Beard andPruppacher (1968), Abboth (1974) for drops of similar size and
by Beard and Pruppacher (1971), Tung and Beard (1978) for dis-
similar drops. The results show a good agreement with thetheoretical values of collision efficiencies, at least for the
drop size ratios p < 0.1 and 0.7 < p < 1.0. The lowest range
is the most important domain for smoke particle sc'vongingstudies.!
The conclusions from the theoretical and experimental
studies of gravitational coagulation were applied in simplemodels of raindrop scavenging of the tropospheric aerosol and
in experimental studies. Several articles of this nature were
published in the proceedings of the symposiums on Precipitation
Scavenging (e.g., Dana, 1970; Adam. and Semonin, 1970; McCormack
and Hilliard, 1970). Experimental verification on the theoret-
ical models of raindrop scavenging efficiencies for submicron
particles was performed also by McCully et al (19S6), Barth
(1959), Severynse (1963), by Starr and Mason (1966), Hagen (1967)
and others. The very interesting change in catching efficiency
of droplets for micron size aerosol particles around Re A 250
was explained by the change of the shape of the wake formed behindspherical drops (Toulcova and Podzimck, 1968).
Facy (1960), Podzimek (1966) and Sood and Jackson (1969)
called attention to the potentially important scavenging ofparticulates by falling snow crystals. There is, however, a
large scatter of collection efficiencies deduced from experiments
41
by different authors. In spite of the low catching efficiency
of ice crystals in scavenging particulates of 1.0 to 3.4 um in
size (according to Sood and Jackson I to 4%) the importance of
this process is obvious. The most important is the fact that
on stellar type ice crystals Podzimek (1970) found droplets ds
small as 0.5 um in size and that motion of ice crystals (e.g.,
Podzimek, 1967) greatly enhances their catching efficiency.
The highest efficiency was seen for dendritic ice crystals (also
due to their slow, oscillatory motion) and rimed ice crystals.
The laboratory study by Yue and Podzimek (1975) of the catchingefficiency of the crystals of simple forms and its comparison
with the experiments by Sasyo (1971) and theoretical calculations
by Pitter et al (1973) indicate a potential use of this technique
for clearing of military smokes.
In the author's opinion this simple technique can use anycollectors with fine crystallinic structure and high catching
efficiency. The crystals can be artificially generated and
used at positive temperatures in a smoke cloud. Also, it would
be worthwhile to investigate theoretically and experimentally
the stability of motion of crystal collectors and their aggregates,
4.2.2 Coagulation Due to Electric Forces - One of the first
attempts to influence the evolution of a fog is described in the
article by Van de Vyver (1901). More systematic studies wereundertaken by Wigand (1926) and Wigand and Frankenberger (1931)with the aim to stabilize or to disperse the fog. Wigand also
obtained a German patent on a procedure for fog dispersion by
using unipolarly charged droplets.
In the following years two techniques were developed for
fog dispersion based on enhanced coagulation dua to highly
charged individual collectors (drops) or due to induced charges
on drops in a strong electric field. The first technique wassupported mainly by the theoretical study by Pauthenier Pnd
Cochet (1953) who considered a positively charged drop among
its neighboring cloud drops. Usually in nature many drops bearcharges amounting to several tens to several hunired elementary
charged. Introduction of unipolarly charged particles or
42
" -- . . . . . . . . . I I 1 11 " -- II IIIII. .._________
hipolarly charged particles into a fog might enhance the coagu-
lation postulated by Wigand. Vadoll (1961) undertook labor;tory
experiments in the dispersion of fogs. He came to the Qonclusion
that bipolarly charged drops had less effect on drop coagulation
than unipolarly charged ones. Sedimentation and droplet motion
in a steady and turbulent field present a problem which has
been partly solved in the past few decades in the laboratory
and by theoretical calculations (e.g., Semonin and Plumlee,
1966; Neiburger et al, 1974; Schlamp et al, 1976 and 1978).
One postulates artifical collectors immobile or falling
through oppositely charged smoke or fog elements. Assuming a
collector with radius 10-2 cm, with a charge corresponding to
a potential difference of 10,000 V and with a downward velocity
of 100 cm sec"I in a fog with drops of radius 2.0 Um and 200
elementary charges on each, a simple calculation shows that the
collection efficiency of the charged system is approximately 15
times larger than that of an uncharged one. This .;imple con-
sideration clearly shows that electrical charging can signifi-
cantly enhance the coagulation or removal of fog droplets.
However, other factors such as air stability and uneveness of
the terrain greatly limit the applicability of these results in
the field. The mechanism of effective droplet or aerosol charg-
ing as a result of capture of gas ions has been described by
Natanson (1960). Carroz and Keller (1976) later describe the
charging of sprayed water drops by corona and induction.
Practical applications of these calculations in the field and
in the laboratory were made by Wigand (1931) and Vadell (1961).
Recently a new technique of using charged bubbles (electro-
gasdynamic method) as collectors was reported by Wright and
Clark (1973) and by Chiang et al (197%,). However, the prelim-
inary results are inconclusive.Physical description of particle precipitation in a strong
electric field are found in most of the books on electrostatic
precipitation (e.g., White, 1963). To the author's knowledge,
the most comprehensive study on this subject was performed by
A. D. Little's research group and described by Vonnegut (Little,
43
II U
1953, 1954, 1955, 1956 and 1965). In a series of papers they
reported their investigations on dispersion of warm-fogs (espe-
cially using an electric field). The investigators performed
laboratory and field experiments with a "Fine-wire space-charge
generator" (800 ft. long and 0.005 inch in diameter stainless
steel wire) placed on poles 12 ft. high above the ground. The
wire was maintained at either positive or negative polarity at
potentials up to 35 kV. The electric field was 10 to 20 V cm"•
which appeared to be insufficient for any noticeable precipita-
tion of fog droplets. During a private discussion, Dr. Vonnegut
expressed the opinion that the main problem, due to corona
charge losses, is to maintain the high potential difference
above the uneven ground. Different aspects of electrostatic
fog precipitation are mentioned in the article by Phan-Cong and
Jordan (1969). Furthermore, the rather pessimistic outlook for
electrostatic fog precipitation is supported by Tag (1974) who
performed a numerical simulation of warm-fog dissipation by
electrically enhanced coallescence under conditions correspond-
ing to a real situation in the Panama-Canal zone. He found
that only an electric field as high as 3,000 V cm" 1 might cause
a significant improvement in visibility. However, this result
contradicts a more optimistic conclusion by Katchurin (1973)
who used a very simplified one-dimensional model (without
interaction with the environment) and an electric field of
1,000 V cm"1.
4.2.3 Coagulation in Acoustic Field - J. W. Hann and W. Kbppencommented in 1889 on an observation by Ch. E. Guillaume of fog
dispersion aftei cannon grenade explosions. The explanation
given was the accelerated precipitation of fog droplets, A
systematic investigation of this subject was performed by
Andrade da Costa (1936), by Brandt and Hiedemann (1936) and by
Branidt et al (1937). They investigated the coagulation of
smoke particles by acoustic waves. The latter found that for
cigarett smoke particles the most effective range is between
5 to 50 kHz. These ultrasonic frequencies were highly absorbed
by the medium. Andrade further mentioned an important observation
IL4"
that two small spheres in a vibrating medium repel each other
if the line of their centers is parrallel to the vibration
vector, and attract each other if normal to it.
These observatione later were exploited for fog dispersion
by Tverskoi (1960) and for many applications in science and
industry (see Mednikov's book, 1965). Except for several ex-
.periments made with powerful sirens at airports (e.g., in Israel)
most of the studies were confined to laboratories. Larca and
Capuz (1969) observed the artificial precipitation of fogs at
frequencies between 5,000 and 15,000 Hz and Viltsev (1969) re-
ported on much faster fog dispersion in an acoustic field
generated in a chamber of 500 m3 volume. However, Podzimek
(1971) was unable to detect any influence by sonic and ultra-
sonic waves on the phase transition in a supercooled artificial
fog. Recently, a systematic study of acoustic coagulation in
aerosols has been started under Prof. Shaw at the New York
State University, Buffalo, New York.
One can simplify the equation for particle motion in an
acoustic field in such a way that the gravitational term, the
term associated with acceleration of the medium by displacement
of the particle and the so-called Bassett's term can be
neglected (e.g., Fuchs, 1959). From the equation of particle
motion under an acoustic field characterized by its velocity
v - A sin (wt) and a particle velocity vr - Ar sin (wt-#) one
obtains a general solution (Katchurin, 1973):
t
Vr * (v) e' + A sin (wt arc tg Aw) , (17)• 2 Pr2
vhere X -" (p - particle density; r - particle radius;nn - dynamic viscosity Most important is the ratio of particle
to air amplitude K = [+ ÷ XI2] -/' and its derivation._* max*dK The maximum value of 3X indicates the most effective
coagulation and smoke or fog dispersion. For the case of acoustic
coagulation of water droplets in air, one finds the following
frequencies (corresponding to different dK/kra): 200 Hz for
r 1 10 vm; 3,800 Hz for r - 2 Um and 90,000 Hz for r - 1.0 Wm.
45
However, the strong absorption of sound waves strongly limitsthe application of sound waves for fog dispersion on a large
scale.
Recently several authors mention experiments that combineacoustic coagulation techniques with hygroscopic nuclei seeding
or with thermal methods (e.g., Katchurin, 1973). No results of
similar experiments have been reported yet.
4.3 Evapgration
Evaporation of fog drops or a substantial change in thedroplet size distribution can be accomplished in several ways;unfortunately only a few of them have been applied in the atmo-
sphere or in large laboratory simulation chambers. They can
be divided into methods using direct heating of the air, heating
combined with artificial air motion, mixing of air masses,absorption of solar radiation and heating by laser beam.
4.3.1 Heating of Foggy Air - This technique has been known forseveral decades and is known under the name FIDO. Some inves-
tigators prefer to use the name "passive heating" in order toindicate that in this case air is not blown simultaneously by
generators into the foggy space. Because the principles of
this method are of primary importance in fog dispersion it is
worthwhile to mention more details. This part will be basedin essence on the approach by Katchurin (1973) which represents
a simplified one-dimensional model. The main deficiency of this
model is the insufficient coupling of the foggy air mass with
the environment and the ground; however, it shows in a very
instructive way the evolution of microstructure of the heated
fog parcel.
Two equations describe the fog's microstructure: one the
time change of relative humidity f, the second the time change
of droplet size r.
df L dT 4?rD*n' P 00 2.- f - -+ kN [(f-1) fr 4(r~dr - ýT] (18)
dr 1 2av 2D*E i (f-I) D*iiE (19)if r~N' r TP-KkNTP
46! ,.,2~
where f is the relative humidity; t - the time, L - latent heatof condensation of water vapor, k Boltzmann's constant, T -
absolute temperature, DO - corrected diffusion term for water
vapor, n' - total concentration of water drops Eg-'i, P - molec-
ular weight of water vapor, P - air pressure, N - Avogadro's _number, r - droplet radius, ;(r) - density function of the
droplet radius distribution, a - surface tension, and E - satu-
rated water vapor pressure in the environment.
Two basic assumptions are made:
1. The nuclei which remain after droplet evaporation willnot influence further physical processes in the foggy air.
2. The rate of temperature change dT can be related toUTthe rate of heating in the following way:
For a unit mass of foggy air the heat change is
-(Cpa cq+) dT _ d (20)p•- L , at20)
where c is the specific heat of humid air, c - specific heatpa
of water, q• - relative specific humidity (grams of liquid water
per gram of fog). In the case of military smoke the values ofq+, L and c have to be changed. The last equation can be ex-
pressed in terms of the size distribution of droplets
t 4cL) dTn fr 2 -(r) 3-t dr. (21)
Because one can assume cpa >" cq+ and
(Cp + cqý) ).> 41rlpn' fr2 ;(r) - drpa 0
the assumption of d dT is admissible.Burning an amount of fuel dm having a specific-heat-release
coefficient a and a water-vapor-release coefficient a' one
obtains two equations (cp Cpfg
adm cp ý dT L L d(
47
C'dm - dq' ds - dE LE dT. (23)T M1 T
M is the molecular weight of air and P is the saturated air
pressure. By eliminating dT in the equation one finds an
expression for the change of liquid water with the amount of
fuel burned:
dq+ LE. - c Na'
- ~ cpN + LN (4
The change in liquid water content can be easily related to the
visibility which in turn can be approximately expressed by
eq. (3) which yields for a total droplet concentration n 100 cm
t - 33 qT 2/3 (25)
and
3 qTT = - d2 , (26)
where L is the visibility range and qT is the mass of water.
Finally one might ask how the visibility improves as a function
of the amount of fuel burned for instance if a 10' cal g-1
and a' * 1.4 g(H20) g-1 (fuel), theng 2 LE N-
dL 2 M dm(27)c pN + LN R M TT
where 1, T, qTo and m are functions of time. However, qT is
related to t by eq. (25). Also, qT and m are related to f by
eq. (23). Disregarding for the moment the very important as-
sumption of adiabaticity leading to eq. (27) (which would mean
the underestimation of heat needed for fog dispersion), one
can calculate from eq. (19) F - F(t); from eq. (24) q+ - q+(t),
resp. n' - n'(t) and from eq. (27) L - 1(t) if the time change
of m (or Q, or T) is known. However, some simple assumptions
are needed for the heat transport above the (heated) ground.
48
-__~ ~ ~ ~ ~ ~ ~ ~ ~ ~~ ~~~~o ...... . ..... .......... ............
The simplest assumption is a linear time dependence for the
consumption of fuel which leads on the average to e.g.,dT /dt - 10"•3C sec 1. For the case of initial and environ-
0mental conditions (qo W 0.15 gm-1; n' - 4.0 x 105 g-1;
To - 2739K) and homogeneous temperature stratification calcu-lati6ns were made. The results are shown in Figures 3 and 4where the parameters n', •, . and qT are plotted as a function
of time. It is apparent that during the first 400 sec there
is no substantial change in mean fog droplet radius and invisibility. Then suddenly the mean droplet size starts todecrease and the visibility improves. The evolution of a real
droplet size spectrum is shown in Fig. 5. This scheme can beapplied for a rough estimate of the amount of fuel needed to
disperse a fog through heating and droplet evaporation. Usu-ally one assumes that an increase in air temperature of 10C issufficient for fog dispersion. This roughly corresponds to
0.75 tons per hr. if on, considers a foggy space above a sur-face 50 m x 1,000 m (Katchurin, 1973) and under conditions
similar to those depicted in Figures 3 and 4.Passive heating of the air has been practically applied
in the FIDO system which occasionally has operated at several
airports. Stewart (1960) reports on field trials which used3 parallel lines of FIDO burners 900 yards long and were per-
formed in 1959-1960 at Marham (England). The conclusion wasthat a shallow dense fog can be cleared. However, probably for
economical reasons there was some hesitations in continuing theoperation. Details of the FIDO operation at London and neces-sary requirements for successful fog dispersion were outlined
by McDonald (1960).
Even disregarding the economic aspects, which certainlywould not play a decisive role in a military operation, one has
the impression that the simple mechanism of droplet evaporationwould not be applicable for clearing of a military smoke com-
posed e.g., of phosphorus particles. One feels that it might
be applicable for a natural fog or fog of highly hygroscopic
particles which are in equilibrium with the environment.
49
S. .. .. = -i-d- - . . .
I(m) q(gin3)-0.4
3-3
r0.2
- -4 0.1
0- io 0-J00.2 0.4 0.6 08a T (OC)
I I ,'I0 400 800 t(sec)
Fig. 3. The change of the mean drop radius F [pm], specificdroplet concentration n' [g-'], liquid water content[gm- 3 ] and visual range Z [m] with the time. Thefog with the initial liquid water content q, - 0.15gm'- was heated steadily at the ground. All otherparameters are identical with those under 4.3.1.Katchurin (1973, p. 238).
50
to ATB:I¶9
II ------
2325287 289 T (K)
Fig. 4. The relationship between temperature, altitude andtime for turbulent heat exchange above the heatedSround, environmental temperature T.. - 2831K;eated ground 2temperature To 3230K; q -0.1 gm-3
and D = 0.1 m2 s-1, All other parameters are iden-tical with those under 4.3.1. Katchurin (1973,p. 224) .
r~(r)
120 c
240
480
02 4 6 8 r O~m)
rig. 5. Time change of the fog droplet size distribution -
if one assumes the following initial parameters:0.1S gm"' T0 a 2739K, no' a 4.10' -1
-0 ̀ [Os']. All other parameters are the sameas under 4.3.1. Katchurin.(1973, p. 237).
However, in a practical applica-..on, the heated air always will
start convection and the mixing of air masses results in dilu-
tion of the aerosol cloud. This will be discussed in the
following paragraphs.
4.3.2 Combined Heating with Dynamical Method - It is verydifficult to imagine passive heating of a large air mass without
setting the whole environment in motion. One feels intuitively
that the blowing of heated air over a large distsace can be made
to cover efficiently a large area if oriented in a proper wind
direction. This is the basis of systems currently operating orin preparation at several airports in France and in the U.S.A.
In France they are named the TURBOCLAIR system and were designed
and installed by Societe Bertin et Cie at Orly and Charles de
Gaule airports. The principles of this system have been de-
scribed by Bertin (1964) and its operation analyzed by Serpolay
(1960), Cot and Serpolay (1966) and Fabre (1971) mainly from
the point of view of the results obtained during systematic
field experiments on the airport Orly.
The physical description of propagation of exhaust from ajet engine has much in common with rising hot plumes in the
atmosphere. One can treat both cases simultaneously, distin-
guishing only between the buoyancy force and the force generated
by the jet engine.
Two basic equations govern the behavior of the hot airjet: first, Newton's equation for the air mass in the jet,
characterized by its density p', mean velocity w' and the cross
section of the jet S', and second, the relationship for the
conservation of heat in the plume (e.g., Katchurin, 1973):
(p'S'w')g T-T . (P'S'wI) dwt * w, d (p'S'wI) (28)
:dT" -" T + (T'-T) d (1S'w'] -'1 (29)
One assumes that the buoyancy force prevails in both cases and
that cp outside and inside of the jet (plume) has the same
value. The second equation was obtained by including the change
52
S, , . ...... . . .. . . . .. . . .
of the heat content of the parcel of air inside the plume
cý P' S' WI dT' with the oxchange of heat from outsidec p (T-T')d(p'S'w') and the work from Archimedes and gravitational
forces during the ascent of the :'ir mass- (p'S'w')g T- + (p'S'w')g dz. Primed symbols are not related
to the outside air. The first term of the second equation is
the change of the air temperature in the rising plume due to
the adiabatic temperature change. The second term describes
the heat exchange with the environment which depends on the
temperature difference (T'-T) and on the entrainment termi
at (p'S'w'). The heat exchange in eq. (29) can be replaced by
humidity exchange with the environment. For a change in specific
humidity in the unsaturated air inside the plume one can write
= -(q 1 -q) ( S1 (30)
This equation may also be written in terms of relative humidityf'E'f' by replacing q' with -p or, more accurately, with
f-E' [1 - -p-r 0 where E' is the saturated water vapor
pressure in the plume at a temperature T'
If condensation occurs an additional term related to the
released latent heat must be added:
L d (____ d dT'(-r (Q'-q) J--(p'S'w') 1" + a-- , (31)
where Q' is the specific humidity of saturated air inside the
plume.Many articles have been published on the subject of entrain-
ment of environmental air into a heated air plume. Some of the
results of theoretical and experimental studies are published
in monographs on cloud dynamics (e.g., Schmeter, 1972). The
approach often used is based on the qualitative comparison of
the mass change inside of a plume p'S'w' to the flux of the air
from outside. This can be expressed for a unit of length byt dthe comparison 3- (o'S'w') V i'vl, where V' is the mean circum-
ference of the plume cross section, v, is the mean velocity
53
into the plume and depends mainly upon the air turbulence along
the boundary. Then
1 d tVlI it4 A' A•S•'T ( 3 S2w'P - -r T (32)
which for a circular cross section where R is the
radius of the cross section) can be expressed in the form
1 d , CT'PS'w (p'S'w) - T (33)
(C is a constant which has a value between 0.18 to 0.24 inaccordance with measurements made in the atmosphere). C - 0.22
(Katchurin, 1973) is recommended for artificially heated air.
This last equation also can be used to calculate the shape of
the heated plume. After differentiation eq. (33) yields
2 dR C T' 1 dw 1 dT' + (34)IT U-" IT 7- aT+ T- r TzB-+ 1W4
Simplification of the entrainment term enables one to
express the basic equations, for the force acting on a risingdw dw
air parcel ( w and for the exchange of heat and specific
humidity, in a simple form:
1dw T'-T C T' S(3S)
1 dT' C T'-T + (36)
1 d C T' (37S.(37)
This system of four equations enables calculation of the
four unknowns w', T', q' and R as functions of the coordinate
z in the rising plume. For droplet growth or evaporation the
last equation can be replaced by a similar relationship for
relative humidity f' inside of the plume
df- f'[(1 - f) CT' + . (38)
54
; ' .. . ;;•. " -:' - • -,. .. ... ... . ... .. . . - . T -- • . . .. ;.r .. ... -
The above mentioned can be simplified in the assumptiondR . tga 1 const. This is often used in technical applications
for the region not very far or close to the source of heat. In
addition, one can incorporate in the equations a horizontal windC Ccomponent v by replacing the expression by R .wi
The approach mentioned above represents a strorg simplifi-
cation of the very complex nature of heating of air which onone hand accelerates the air mass and on the other hand dilutes
the aerosol cloud. However, in the author's opinion it indicates
the direction for future research on systems similar toTURBOCLAIR presently limited to measurements of the temperature
distribution in the field and to attempts to deduce some semi-
empirical formulas for practical applications. Along similarlines has been the attempt to disperse fog by the use of jetaircraft engines (Smith and Wexler, 1959, Serpolay, 1960, Zarea,
1962; Appleman and Coons, 1970). A more sophisticated system
is currently under construction in the U.S.A. folloving a
thorough investigation of the heating of a runway by jet type
generators (Weinstein, 1973; Kunkel and Silverman, 1974; Kunkel,
1975 and 1977). In this experiment two model spaces (one60 x 150 x 800 m, the other 30 x 1S0 x 400 m) will be heated
by a system of 34 combustors located along two 600 m lines
spaced 155 m apart. Fuel consumption of 250 gallons per minute
is anticipated. For comparison, the technical parameters of
the French TURBOCLAIR engines are: the emission of 50 kg sec"1
of air at a temperature of 5009C. The mean speed of escaping
hot air is about 500 m sec-1 (Dubois, 1975).While the simplified theoretical approach could be refined
by theoretical studies of convection in a turbulent atmosphere,in its present form it still enables one to solve a problem
important to military smoke clearing: heating of the aerosol
air parcel, simultaneous dilution of particles in the unit of
volume and transport of particulates to higher levels. To the
author's knowledge, there is not a model currently existingwhich describes all three processes simultaneously. A numerical
solution might yield some results important for the modeling of
55
some of the most important interactions. The equation for the
force equilibrium (eq. 35) shows that for a given amount ofI
released heat (TO ) and a given entrainment (which will probably
include C as a function of z) one can find an altitude at whichdw will be zero. In other words, one should attempt to find the
optimum rate of heating compared to the entrainment in order totransport the particulates at a lower concentration to a desiredheight. This special non-steady state case requires the sudden
release of a large amount of heat. In addition, numerical
modeling of convection indicates that two- or three-dimensionalmodels can be established for quasi-steady situations (plumes,jets, and ordered convection) and known environmental parameters.
An attempt at a theoretical solution for the movement of
air in the horizontal and vertical direction as an explanationfor the dispersion of a fog has been made by several authors
(e.g., Zilitinkevich, 1967). The technology required to intro-
duce heat into a foggy space is the subject of many patents
(e.g., from the German Patentamt one can quote No. 1208754 and
1292153 of the Socidt6 Bertin & Cie, No. 1016049 by Baier, No.
1154136 by Partl and No. 2602878 by Romatowsky).
4.3.3 Dynamicai Mixing With Dry Air - Each of the methods
examined sets the environmental air in motion and facilitates
the circulation and transport action of the clean and dryer
air into the foggy space. Many attempts have been made to
introduce directly the dryer air into a fog by means of power-
ful ventilators. Usually the ventilators were installed along
the edge of a runway and sucked the air from it. This airmovement causes a downdraft of air which is supposed to be
dryer and which during the descent is pseudo-adiabatically
heated. A one-dimensional model for the time change of specific
humidity, including the effect of turbulent diffusion, can be
written in the following form
t ( ;a _-)w f IT (39)
The three terms on the right side of the equation are the
56
• V\ .
contribution by turbulent exchange, downdraft and heating of
the air (pseudo-adiabatic or adiabatic). Zilitinkevich (1967)came to the conclusion that fog dispersion by a dynamic process
is possibly accomplished by droplet evaporation. Hi3 theoret-
ical model is based on n, uation similar to eq. (39).Katchurin (1973) -med a calculation on the practical
application of this metl1 Lor thin layer fog dispersion. He
found a reasonable improvement in the visibility over a spacecovered by 25 powerful ventilators (jet-engines with an airflow of 100 ml sec-' and gas consumption of 2,000 kg hr' 1 ).
Another possible technique, considered in the U.S.A., is
to use the downdraft under a helicopter. As early as 1937, itwas observed that under specific conditions a thin layer of fog
can be dispersed by an aircraft wake (Katheder, 1937). In 1961it was suggested that a helicopter be used to disperse a
radiation fog. In the following years a series of experimentswas undertaken mainly by the U. S. Air Force. H':k (1965),
Buxton et al (1968), U. S. Air Force Laboratories at Cambridge
(1968 and 1969) and Katchurin (1973, pp. 220-226) describedthe potential and difficulties of this type of fog dispersion.
This technique apparently works for a fog layer thickness lessthan 200 ft. if the air above the foggy space has a relativehumidity below 90%. In other words, its successful operation
requires the presence of a very strong temperature inversion.
Remarks based on a numerical study of selected meteorological
situations on the use of helicopters to clear fog, were madeby Johnson et a! (1974).
The application of the dynamical method for clearing ofmilitary smokes has the limitations mentioned in the literatureand in addition, the disadvantages of the application of the
methods based on evaporation of droplets in general. One shouldtry, in the author's opinion, a simple experiment to see if the
downdraft under a helicopter causes intense impaction of smoke
particulates on the ground (covered often by vegetation) and
permanent removal of a portion of them.4.3.4 ',bsorption of Solar Radiation Many years ago it was
=. i57
suggested a warm fog could be cleared by intensive droplet evapo-ration by increasing the absorptive power of the foggy air
irradiated by the sun. Often mentioned is carbon black as a
substance, which can be easily dispersed, has high absorptive
power, is relatively harmless, and is cheap.
A very simple model of the effect of increase of absorptionpower in a foggy layer has been considered by Katchurin (1973,
p. 252-2S4). He assumed that the air above the fog would beseeded by carbon black powder (n g/cm' and of mean particle
radius r). He calculated the change in temperature of thestagnant air in the environment according to the formula
dT In a (0S 4 r '(4 0 )
*pCaP a
where I is the intensity of incident radiation, p p is the
particle density, c is the specific heat of the air, pa is theair density and a is the coefficient of absorption of theincident radiation (dependent on the natuie and size of the
particle). This simplistic model (interaction with waterdroplets is neglected) assumes further that the heat supplied
to the parcel is used fully for droplet evaporation
aInNrr 2 -LPa at (41)
(N is the number of absorbing substance particles per 1 g of
substance) and that the change in visibility can be expressedin accordance with eq. (3) in the Introduction:
di In(42)Z- q rop Lpa(2
p a
Katchurin (1973) used for his model the following values:
r - 0.04 urn, n - 10-5 gin', a - 0.50, I - 1 cal cm" 2 min-n.
The other quantities correspond to the normal properties ofcarbon black and the air in the fog. He obtained increasedvisibility of 4% min'n which is comparable with previously men-tioned and more expensive techniques. In the author's opinion,
58
however, this case seems unrealistic due to the assumption that
the conditions in the top larer of the fog are characteristic
for the whole space, that there is no interaction (optical) be-
tween the droplets of equal size and no air motion induced by
the heating. The interaction with the soil, so important to
fog formation, is completely neglected as in the previous model-
ing for other techniques.
Some of the problems of the absorption technique for fog
dispersion were discussed by Penn and Oser (1962). Qualitative
observations after the seeding of layer type clouds with carbon
black show that only insignificant changes of the top most
layer were observed if any (Katchurin, 1973). This technique
is not widely used today.
4.3.S Evaporation of Drops in a Laser Beam - The use of a laser
beam to evaporate fog droplets to improve the visibility has
been considered by several authors during the last decade.
Ferrara et al (1968) investigated laser beam scattering on fog
droplets and outlined a solution of equations for a polydispersed
fog. Kolosov (1969), using different wavelengths of laser
signals (0.63, 1.15 and 3.39 Vm), found that the coefficient of
attenuation in an artificial fog increased linearly with the
increasing liquid water content of the fog.
A systematic study on the use of laser beam for fog dis-
persion has been reported by several authors in the U.S.A.,
U.S.S.R., England and other countries. Bukatiy et al (1974)
used an SO2 laser (X - 0.63 pm) to attempt to disperse an
artificial fog. They analyzed the space and time distribution
of scattered light to include the influence of convective
currents. Belyaiev et al (1975) reported on experiments using
a CO2 laser (N - 10.6 pm and 1 to 10 2 W cm" 2 ) for clearing a
fog along a 4 m path. A second laser (with X - 0.63 Vm) wasused for attenuation measurements. The authors justified the
selection of the wavelength of 10.6 pm with the high absorption
of the transmitted energy by fog droplets and low absorption
by air. They found a slight but permanent improvement in
visibility preceded by a very short period (10 seconds) of a
59
very significant increase in visibility.
Direct evaporation of fog droplets in a real fog or asmoke cloud seems to face many problems. In particular, theuse of a wide beam will generate intense turbulent exchange
along a beam. In addition, wind will drift new fog masses into
the small cleared space. Some of these problems have been con-
sidered in the study by Belyaiev et al (1975).
4.4 Dilution of Aerosol Cloud
4.4.1 Horizontal Air Mixing - Horizontal mixing of air massesaccompanies many of the processes leading to droplet evaporation
by dynamic methods. Often it occurs as the natural phenomenon
of a horizontal wind blowing into the space covered by smoke.
The process can be described by an equation similar to eq. (39)
in section 4.3.3 (the coordinate system must be changed and the
last term on the right hand side omitted). However, the solution
is very complicated if one includes the settling of particulates
transported by the air motion and the wind velocity profile
above the ground. One for this reason cannot apply formulas
published in all standard textbooks on atmospheric diffusion
deduced for turbulent diffusion. The models are usually based
on a Gaussian distribution curve of pollutant concentration
(Sutton, 1953).
4.4.2 Convective Dilution of an Aerosol Cloud - Much attention
has been paid during the past two decades to the application of
artificial convection to dilute pollutants which are concen-
trated on the bottom of an industrial basin in a temperature
inversion. Systems of very powerful ventilators, chimneys withoverheated relatively clean air and installations similar to
the French Meteotron have been considered. For military appli-
cation a strong updraft generated by a sudden burning of fuel
has been considered.
In the case of a temperature inversion (typical for apersistent military smoke over a large area) one can assume an
initial smoke density n0 (g'-] equation in the form of an
exponential law no - noo exp(-,), where noo is the smoke particle
density at the ground at a time t * 0. A one-dimensional model
60
f -
will be described by a set of three equations, two of them
identical with eqs. (3S) and (36) and the third having a form
I dn' C T'"n --r 3z ""(43)
The dilution rate can be calculated from R = R(z) with a correc-tion for particle settling. Also one could attempt a more
complicated model for a sudden heating of the polluted layer at
a certain point on the ground and the subsequent dilution ofthe air mass by a rising and diluting aerosol sphere. The
starting point for such calculations could be the model used
first by Sutton (1953).4.S Other Techniques
Many other techniques, other than those mentioned above
have been described in the literature or suggested in patent
applications. Most of them, such as using water screens, areunsuitable for large scale operations (e.g., Aust:ian Patent
No. 37273, Sepulchre and No. 133646, Jahl). Other techniques
are in the author's opinion not proven. Among those one finds
e.g., a suggestion for increasing the precipitation of smoke
particles by mixing in a fine powder of silicia monoxide whichis thought will become negatively charged (Austrian Patent No.
29038, Potter, New York). Two Austrian patents (Nos. 195901
and 197807 by Waagner and Biro, and by Potocnik and Pointer)
deal with enhanced coagulation in high magnetic field "filters"
of smoke particles with particulates which have high magnetic
moments.
Kumal and Russell (1969) studied attenuation and back-
scattering of infrared radiation in ice and water fog at temper-
atures down to -40*C. The data were used to calculate
attenuation and backscattering for X - 2.2, 2.7, 4.5, 5.75, 9.7and 10.9 pm based on Mie theory. Several authors also suggested
using infrared radiation for fog droplet evaporation at airports
(e.g., German Patent No. 2005431, Industrie-Comp., Kleinewertens).
Many patents contain ideas for using hygroscopic particlesof different composition or surface active substances to enhance
61
the droplet growth by condensation or coagulation (e.g., Austrian
Patent No. 183749). Others exploit a special arrangement of
electric fields (e.g., German Patent No. 1159984) or acoustic
waves (e.g., Austrian Patents No. 154887 by Brandt et al, andNo. 162480 by Jahn and Himmelbauer) for particle precipitation.
Various installations for dehumidification or warming up of
foggy air also are suggested (e.g., German Patent No. 1016049,
Baier); however, the author found very few of the patents included
detailed analysis or results of pilot experiments.
5. FIELD EXPERIMENTS
The features of some of the experiments on warm fogs will
be discussed from the point of view of the potential use of
similar techniques for the clearing of military smokes.
The seeding of warm fogs with hygroscopic nuclei to increase
the condens3tional growth of droplets and the subsequent
vigorous coagulation and fallout of droplets has been performedin different ways:
Hougton and Radford (1938) suggested using a solution of
concentrated calcium chloride in water at a rate of 4 to Sliters of saturated solution per second. A similar substance
was used at Cardington in England with inconclusive results.
Ground and aerial seedings have been done by the staff of the
Cornell Aeronautical Laboratury, Buffalo (Kocmond and Pili6,1969) at the Elmira airport. A total of 25 ground seedings and
6 aerial seedings with NaCl were performed. Up to 700 lbs ofNaCl powder was dispersed over the foggy layer. The mean
diameter of the salt particles was between 10 and 30 pm. Air-
borne seeding was revealed to be more effective. The same group
performed another series of field experiments with warm-fog
seeding two years later (Kocmond et al, 1971). Several kinds
of particulates were used [NaCl, Na2 HPO 4 , CO(NH2 ) 2 and poly-electrolytes] and dispersed from the aircraft. All particles
improved visibility except polyelectrolytes even in a dense
62
valley fog. The improvement in visibility diminished duringa strong wind and intense turbulence and in a fog with a highliquid water content. Silverman and Smith (1970) came to asimilar conclusion after warm-fog seeding with the NaCiparticles. Routine warm-fog seeding with NaCl, polyelectrolytesand surfactants was done in 1969 at the Sacramento and LosAngeles airports and at Nantucket (Osmum, 1969). The resultsslightly contradicted Kocmond et al (1971) due to the higherefficiency of polyelectrolytes and surfactants in comparisonwith NaCI particles. The amount of salt solution dispersedfrom an aircraft flying at a normal cruising velocity wasusually higher than that recommended by Hougton and Radford(1938). Over the Naval Air Station, Lemoore (Cal.) White etal (1969) used 200 gal of solution per minute during theirexperiments with warm-fog seeding. Large scale experimentswith dispersion of giant monodisperse hygroscopic nuclei inItaly have been described by Montefinale et al (1270).
Several experiments made with urea (Silverman et al, 1972)ammonium nitrate-urea (St. Amand et al, 1971) and urea-bentonite(Depietri and Rosini, 1968) showed the potential for use ofthis substance for dispersion of fogs and clouds. However,some investigators stress the necessity of mixing bentoniteparticles of 3 Pm in diameter into the urea in order to preventfast coagulation. Several descriptions of the use of surfactantsfor the partial deactivation of hygroscopic nuclei or dropletsduring the field experiments are incomplete (Mihara, 1966;Bigg et al, 1969). In addition, the technology of nucleipassivation and the amount of deposited surfactant is largelyunknown.
For the evaluation of similar techniques for the clearingof military smokes the most important information needed iswhich parameter was measured before the seeding experiment, andwhat control of the final effect was achieved. The mostimportant microstructural parameters to be measured before thefog is seeded are: size spectrum of the droplets (or at least
median drop diameter) and their number concentration, liquid
63
.. _ n ~ m, _ _Lm | _ _ _I I I . ..A
water content and the same parameters for the seeding substance-particles (instead of liquid water content the amount of seed-
ing substance dispersed in a unit of time or over a certainarea). Occasionally the measurement of the background nuclei
concentration is quite useful. During the experiment it isuseful to make several measurements of the same parameters and,
in addition, the fallout rate of particulates e.g., onto asensitized gelatine sheet.
The meteorological macroparameters which one should measureare temperature and its lapse rate, wind speed and direction,
relative humidity, pressure, horizontal and vertical visibility,and total solar radiation (if needed). Very worthwhile measure-
ments are the air turbulence (at least two components), wind
profile or wind shear, some information about the fog (smoke)depth and the height from which the seeding agent is dispersed.
The main meteorological parameters should be recorded during
the entire experiment.
One should do preliminary modeling of the real situationand estimate the interaction between macro- and microparametersfor different size distributions and the total concentrationof seeding substance. One should realize that colloidal insta-
bility is a time dependent factor and that coalescence of
particulates becomes significant with increasing fog (smoke)
depth. The evaluation of a field test should include a detaileddescription of the terrain and its topography.
Any use of the thermal dispersion technique should be pre-
ceded by a detailed analysis of the possible thermal effectsaround the generator. Based on preliminary calculations a
network of stations measuring air motion, temperature, humidityand the corresponding gradients should be established. The
very extensive field test in France at Orly (Cot and Serpolay,1966) or the preparation of the warm-fog thermokinetic dispersal
system of the Air Force in the U.S.A. (Kunkel, 1977) are models
of such an experiment.
Special measurements, requiring the use of aircraft, areseeding the smoke or fog layer and measuring wind speed and •
64
direction, temperature, humidity and their vertical gradientsfrom above the layer. For successful dispersion of a warm-foghumidity less than 90% above the fog is necessary and thedispersing helicopter usually flies 100 ft above the top of the
fog (U.S.A.F. Rept., 1968 and 1969; Plank, 1969).
Experiments based on the artificial convection generatedby burning fuel would profit by the knowledge of the temperaturefield around and above the source, of its time change, and ofthe main meteorological parameters in the environment. Remotesensing detection of the shape of the heated air plume would bea possible solution to the problem.
6. RECOMMENDATIONS FOR FUTURE RESEARCH
This Survey of Techniques for Clearing Military Smoke Clouds
is based on warm-fog clearing techniques and is written by acloud physicist. This necessarily causes some inconsistancybetween the treatment of different methods and is reflected inthe structure of this report. The broadness of the field andthe limited time available dictated that the author could not
read in detail all cited articles. Several articles are knownonly from abstracts and some only are mentioned by title withoutkey words. However, these citations are considered important
and the author hopes that in the future he will have an oppor-tunity to comment on them and to suppliment this report. Inthe author's opinion key words cannot replace even an abstractof an article, but do facilitate orientation .,: the treated
subject for the reader.The recommendations for future research are based on this
survey and, in addition, notes are made on the military applica-
tion of some of the existing theoretical conclusions for thesystematic study of techniques for clearing military smokeclouds. In the author's opinion, the following techniques
deserve more attention:
The use of highly hygroscopic substances for military smokes
65
excludes to a large extent the fog dispersion technique of
seeding with hygroscopic materials. There is a real danger thatone could enhance the stability and increase the fog densityinstead of dispersing it. One possibility that should be checkedis the effect (if any) of covering a portion of the droplet
population of phosphorus particulates by an evaporation retard-
ing layer of surfactants. For this one would need to developa special technique to disperse the surfactant and deposit it
on the surface of smoke particles. Afterward, the colloidal
instability of the smoke could be enhanced by a supply of water
vapor.The coagulation technique merits attention for the reason
that it can be combined with dynamical and other techniques and
applied over the territory controlled by our or enemy forces.It is well known that the scavenging by spherical particles(water drops) is not very effective (Feit et al, 1970, p. 36);
however, little attention has been paid to the collectionefficiency of light, nonspherical particles similar to icecrystals. Their long residence time and their swinging move-
ment certainly contribute to the effective collection of micronsized particles. This is supported by many observations in
nature.
FIDO or TURBOCLAIR or any similar system hardly will beapplied in the manner used at airports. One simple, and for
military operations attractive, system is to use suddenly
incinerated fuels spread on the ground that could generate anupdraft and transports smoke particles upward and dilutes the
aerosol cloud. In the author's opinion, this simple technique
should be analyzed in detail because the supply of heat is
accompanied with a supply of water vapor. For instance 1 gram
of jet fuel produces l04 calories of heat and 1.4 g of water
vapor (Felt et al, 1970, p. 4). It would be worthwhile toanalyze the potential of this technique by a three-dimensional
model for a calm situation and for horizontal wind. In general,
turbulence in the atmosphere lowers the efficiency of systems
similar to TURBOCLAIR, and in the case of suddenly released
66
large amounts of heat, the efficiency of hiat transformationinto kinetic energy. Also the author fevls that the interactionof the parcel of the foggy or smoky air with the ground hasbeen grossly underestimated if not neglected in the majority
of the models.
The simplest technique for warm-fog dispersion--the useof helicopters--has a limited application in clearing of
military smokes. However, if properly applied it can enhancethe effectiveness of the coagulation (scavenging) technique.For thiz veason studies of the flow field and supply of heatby helicopter engines similar to the investigation by Plank andSpatola (1969) are vey useful. Also there is some evidencethat a helicopter downdraft could contribute to the smokeparticle retention on grass and leaves of shrubs and trees.
Some of the techniques mentioned above potentially can beused to enhance several of the mechanisms previously mentioned
(e.g., electrostatic charging or phoretic forces), b"t theauthor feels at this time that they will not play a decisive
role.
After finishing this survey, the author found severalimportant articles which could not be included in the literarysurvey. He hopes these can be added in the future because hiswish is that this study should be useful and up-to-date forthose who work in this field.
7. SUMMARY
This report documents the critical review conducted on thesubject of clearing military screening smokes. Table I suwna-rizes the methods available and presents an attempt to ratethe suitability of the various techniques for military applica-tion. However, the author does not feel qualified to be thefinal judge of the relative military worth of these methods.It is recommended that CSL have this table reviewed by militaryusers for a more informed assessment of potential militaryapplicability.
67
L!
IVALUATION OP THI VAXI•S 30011 ELINIMATION TICMNTONlS
Approach to the pgField sutIty
SRemoval of S6oke epel Technique Experiment RemarksParticles by: gosults ClelrLag
pr Filtration 23 Deposition of fog or Pilot experiments oa fair Most of the re-smoke particles on runways ted results"eliminators" (mesh, r-conlclueiv*doflectors, cooled "Hardly applicable"grids) (CHOulton)-"Appli-
cable for smallspace" (Ju42e)
Sedimentation i4 Simsiy sedimentation Pilot experiments in Fair or The author of thisor sedimentation en- 1ma1l scale with Poor report sug getshanged by downdraft stimulated downdraft using grin ad dry(water drops, hell- by water drops ice for inducedcopter) (Japan) downdraft
PhorItic rorces 2S to Ticrmophoretical Linited experiments Poor Combination with27 DiffuSsiophoetical in the atmosphere Poor other techniques
Photophoretical (very small space) Poor promisingEloctrophoretical mainly with electro- FairForces phoretical forces
Condensation of I ' to Hygroscopic Large scale experi- Fair Very limited appli-Vapors on Nuclei 35 substances ments with some c bilita= for hygro-
positive results. in scopic smokes. Forthe case of clearing Iother smokes a coN-fog biriation with
phoretical forcesor filtration pos-sible (addition Ofwater vapor)
Coaulatlon and 35 to Scavenging on Small scale eyxperi- Fror This author suggests
Sejýnentfation in 37 Isphrical an' non- roits--results con- Good using nonsphericalGravitational spherical bo ies elusive only from porous particlesF::lt I laboratory and combonationel wihlSafrcelseaxper imentswihesa.f
Coasulation and 35 to Biralarly and Large scale experi- Fair or Difficulties withSedimentation in 37 unipolar>y charged ments with some Good the applicationElectric Field particulates (or po, itive results in during military
bubbles) small scale fog operationsI dispersion (Vonnegut)
_ __ (inconclusive)
Coagulation and 39 to Sonic frequencies Large scale experi- Fair or Limited applicabil-Sedimentation in 4I ments with positive Poor itty (short range,kcoustic Field IIresults only in amount of energy)
Foall spaceU t a n cOn y l or t r ;................ ...... .... .................. .. . ... .
A Lltrae1o0nc IOnly laboratory BadLfre4uanclas experiments
Drop Evaporation 40 to Hygroscopic smokes Some positive re- Poor or Nuclei must bethrough Hleating 46 :ults with ware fog lad removed from theof Fog seeding space (difficult)
Oil smokes Not known Good or CbInatlon with
- -Pair dynamical methodC - - ---6
68
,•-l
I:"•
TAILI I
IVALUATIO4 OP THI VARIO WOKII • lLIMINATION TICWIQUS (Cv%'T)
Approach to the Fietld eSuitability a
Removal of smoke Technique xa rcsont for RemarksParticle* by: Ptcnulas Czloearing
Drop lvaporation 11 to Blovers or Inconclusive re- Bad Energy problem,throu h Mixim la ventilators suits fro clearing short rangewith Dry Air for on sall e areaS....................... . ..................... ............ •....................
Helicopters Some positive Pair Enhavncement of theresults dow, drs t
Drop Evaporition 46 to Continuous heat Som posFtive ra- Pair Enormous snergithrough Comblmed to asources suiti clearing necessary, hardly
Thermal aid the shallow fog applicable forCynealcal Mat~odl layer on runways clearzo| 4aok. inlarge scale
operations. ............... ..................... ....... ....................
Instantanueous heat Large scale experi- Good or Might be effectivesources ments were not yet Fair at low temperature
controlled and doc- inversion, howeverunented satisfactory technolo y of suc-
ce... two rstmucst...jJ_
DropEvaporation 52 to Carbon black seeding Showed little Bid Or The amount ofhy Inc!r.:resdtAib-'. $3 of fog changes it fog or iPoor energy obtained by
orption of Solari cloud structure :absrptic%. of solarý3diation (radiation isI insuff ic ient
Drop Evaporation SS to Use of laser to Known only In a Bad or For smoke or fog:y Usin a Laser S4 evap rate fog very limited space Poe- drop eVAporationhs• droplets unsuitable yet
(space, soCrcy)
Drop Evaporation S3 to Use of infrared Known only in a iBad to Unsuitable forthrough Infrared 54 heotcls limited space Fair la rler spaceqadi~tion
4 4-Dilution of 54t'imlt no hor.ebr Results of small Bad Unsuitable forkerosol Cloud by $S izontal wind effect scale experiments F large space opera-
Horliontal ixhing I by ventilators inconclusive tion
,ilution of SS Dilution of aerosol Not systematic ex- Fair Potential use inherosol Cloud by by vertical air periments with the case of shallowVertical ixiing motion (ventilsors, smokes are known smoke layer under
helicopters) inversion
Excellent * mothco demonstrated to work with reasonable logistical requirement
Good * method erzected to work with estimated logistical requirements
air,- method expected to dork, no estimate available of reasonable logistical requirements
-or . uncertain whether method would work
lAd - aethol Set oxpec.ed to work
69
\.
APPENDIX
References, Articles, Reports
Subjects of Literary Study
1. Sioke and Fog, their Microstructure and Composition. Par-
ticle Properties and Measurement:
1.1 Concentration
1.2 Size digtribution1.3 Chemical composition
1.4 Surface properties1.S Optical properties
1.6 Electric charge1.7 Behavior in humidity field1.8 Sampling by sedimentation1,9 Sampling by impaction1.10 Optical counters
1.11 Special identification techniques
2. Meteorological and Environmental Parameters Influencing thePhysico-chemical State of Military Smoke and Fog:
2.1 Temperature
2.2 Humidity2.3 Wind vector
2.4 Solar radiation and other parameters2.5 Atmospheric stability2.6 Turbulence-scale and intensity
2.7 State of the soil
3. Methods of Smoke and Fog Generation in the Field:
3.1 Continuous generators (point source)
71 .--
3.2 Line generator
3.3 Surface source
3.4 Chemical reaction in the air
3.5 Explosions - instantaneous point source
4. Improvement of the Visibility in Military Smoke or Fog by:
4.1 Changing the concentration and size distribution of
particulates by sedimentation, coagulation, particle
deposition under force field, etc.
4.2 Addition (decrease) of humidity
4.3 Addition of active and inactive particulates
4.4 Electrical charging of particulates
4.5 Coagulation in acoustic field
4.6 Changing of optical parameters of particulate cloud
4.7 Intensifying the cloud by dilution (air mixing)
4.8 Changing of the homogeneity of an aerosol cloud
S. Modes of Applying Heat Generators, Seeding Agents and other
Fog Dispersing Facilities in the Field:
S.1 FIDO type systems
5.2 Exothermic reactions
5.3 Absorption of radiation5.4 Quasi-adiabatic warming of the air
5.5 Air mixing
5.6 Increase of droplet mpss by condensation
5.7 Generators of acoustic waves
5.8 Electrostatic generators
5.9 Scavenging techniques
5.10 Filtration techniques
5.11 Generators of heat combined with intense turbulence
5.12 Other experiments
72
% .U"•\
6. Survey of Field Experiments (Description, Results):
6.1 Natural fog dispersion6.2 Artificial fog dispersion
6.3 Smog dispersion
6.4 Smoke clearing
73
M ae . ,W
WI
agu 0 i u
4Jh4
II* i,,U--°~ -,4
in1 0O
0 . u.. U
'"A 41SP-
4610 l 1 4 , v•.l.
* 1 C U&'Ue O.C
0 N 0 SC I0 0
U cy I- la a0 .20'*
N '.4 OUhC .0 ,,,
- .J .... .,, . .
a ; .. ,-. ,,
* -. , U , O." "+ ""
"" w a
1.0
- ,2 .,'cSId. -a
Ou
A Io7
V1 U 4 '"-' 0.
IsI o' ~a..g 4, A4 &I
6, 0 A- wIsIs b
r~I a 0 ICo
¶~ o.
'0 -A
V, u. ed*~~40 d -q
0-9 - 1w%.
In! ow.
76 M 4'
2~~'4~ L ,*
~V -4
41.
0 Oi
tN.4
*Q A
.J k 6vi 6
to 4.....
04 0
8 04 0 N0
wo 4
~ I..i-. 77
II
144
V 0
4 ewei.. N,
IV*
U4 c2 V-c
X u .44' 4
OU-~4 1"*
- -4 0 -9-
Y;.u 0. A.
78~
u 61
U~ CL *Mu
I Im
IA V * ,2 .
.4 $.
I.. ;6d
-.a 0- co
0~~4 a . v . 0
* .~ *- 0 a~~.UN
*d P. -4 S
£M au M U b0ad.6us* d-. . 0'
.8 9
1. ,
IsI
A o w U44 - u
7 'U
0~£ U. a.W; L
44
i* V 0 U 40 a)
a~ $.
V .. .
vr z
to a. u
*.00
-0 ILA 0'
.1. 80
* 41
- 0, ý11
0 >u
-4Ui
0. 0 444
0 -I
4 a 9
6j ..0 X 0
-0 a
dA 6A r6
b ma14 a . W .
03 5I. .. *,
~~~~4 VJ . h a
a ~10
0.0
16. A*
41 0 a
- I.. Q ....
~ .ht0401
41, = A -
6 SO a
z *- S
%6. a0 .
14 "14 o ~
no'.. .4 I 1 '
S~~~0 Alt W" I .V ,
• a
k~lr ID a. q. 1 .-0
C4-i go a6Mi C d0 )-.h4P
+'+ • I~~- ~N qII U ',***''
f, 1 $ - 0.o4 a 4Ja 424
a o3 0 0-I
-- a-
0 0 14 u -,..6-j•
0 w 4- a u a ac !C
Q -1 4*1 w* ba •*4 l- I'A - .
r. ) aaýa v
10
no""" "a""
a...4 46...*
-- ,' * .ll .,4.' •* l ,,,
.0•b 4 -0 o ti 6j03 of q Q
46- 0.0.
a- .0 a v u
0.1 a- d - '
'44- '4a4b
24..~~t j'a a2n. I .-.64 -4 . -4 V
a~0 Aa ~
3-4. .-4 000a a ~ u r n p
ILI - a - - * ..I
.0 all
83~a*'~
I..- U I.- - 3* U 64 -.
@4 -� U4*4 kUF a
i
I64
- 6464o.
UIJ U I� �
64 64W.� -4a UU
- 4.. u� .4
(.� �
I. I. .0 � :1
* 64 a o�.- .0 �
0. -.4 .4-4 64* *0 � .4.00 Lu � .4
'-4 a- 4,... 0. 64 -
64� �44 U*�I �4.Lu 14 * Lu
.4 0 .064* 0* 0.0 0.A
I a I -. u q�44� I�
* b. - * :2- -o.64 0- . .64 -� 0� 4.440 0 .-4 064
4. 4. --
4. 64. 4.� 640
40 40 64- I.. 4- -�64 4..*64 .64
0 0
2 64
- 064 '64
U .4a
00
a0 0 -. �
.� �- 4
I-
* .444- "4
- -.64 . 2.
o .� -
- 4-k8. 04- 4
I Eu . �>.-.I pa
* *4�* 40 4*� .464
� .J� .44
- 4424.Aa 4-4
4.44 I84 ___
I'-'.4. --4.
I. - I
.. IV k %[. Iso'**I
94 's'
93 . 4-~ A~
, 0 "0:x af 'cou
x a.0Y 4 0 W4
C: 1 'A to A " 9~ ~- ~.m4 40
onI. 0 .0 w 14"
106 C:U A4U M
a LL. 04'0 0 Q1U MU..
oo 6 *
.4V 00 91494 fn4
1 01 IO II
4144
CP- do 40
(AU
114a P6
414.
A
8s
~iW ~u14 9
-4.4
I.-a
v- 0
.6U p..
-4 u00A S.
.4 A.4 -
S+
ii:i - .-
- SS'.4
* O0Aaim 0 0
00 a
10 0 on V
o l A1 ]* 87
- , *eoe
M **
- 0-d 'S.
AdA
an 0 06
C3 0 4. ýoU
61
-4 .M a0 ~I "'
.. 4
0- i- 0a
I.NI 0
N a.'
-A*'- 40 ~
~~5';~ U~ 40 C W111- . W. g ~ 4 u1
C6 a
I-.M u.U 0 4' 0"I C 4
-~t a 6aa -~0. z .4
-. -0 !0
B C0 u a. ~ '
aob~ 01 14u C4U
4.. 4-Q 4'.~U.4 . " . ~ ~ u 0 - 4O-U@ ... ~*~ 4~4jA
u C L4
4.1 1 4' '
S .0 IL -
WI 0*4 0bO0. k a- g-
- a
-a-. Ad a ~ -- -
"U i
Ir "a. 0-0a0ii 34-
-If.. k. 8.
IIV .
ol 44 odd
94 wI ~%d Sk 'I U V %
-o r
on i, -u .0a A4
a U~ eaA1.. 416 u *0 a 41'dme Aa l CL A Up .4'
IA 0e r I
r4- ILa
-- u6. 0 j*
- ~ I.- V
6'-o M 'A 0 (
~. -. ,~ -0 abA
-4 W * 4
A. 1.
.4 6d
0- 94
-4-
i . !~ 1 4 0", A IA-.41II
.. 00
6 - 0 1 ,
49 . 41 r
k1 4.0 *0-0-96 9 0
1-0 .4 N
am 0 a*I-1.
L4 z CL.
u~ 0 a
64 4j-~
a - Nn
-00
* - 4'0. 0.41
00 0,
60 4.'n - -6
a h4'CUI- 090
u v-
a.
'0.0.
:1. 1' 4.
0. I
U00
Al 3 0 wa
a. . v9 00u0 04
! .394 I a
Is ao~ *040 u.-
S 110 14' - 4.
me A 'A % &- O n .
0 W c.6oo *-1.0 - ~..0 0
.06- .'0 00
*~g. -~.0- .1 %4
o 644 4.0U4
1O
04
0%AV 0 1. - I. -
., 4.4 1'40 0. -40 -. f
0 wU4Q *
.4 % i.~ 00
' .~ "1
~. .044. 0.
- 2 17- 7L - TTTI -ý - - 7- -
0*4 0
A d.
0 0~
0-- *j01;
~' 44~ S.N
lo lo=4
LIU4, L
m~~~c -Ur'-
O U N ~ O M ) .~ *Od~.0-
~wOO V.4
-0 V) 4' I.
~W i
4. ah
9% 36
- 92
* •
I. Imm.
o 'J MUoo .
.4!* - .4-
_ =m . . . - - -at
4w -SP u~i
.W, I.
a 30 .a
-0
a 0 4
.40r IP
30 c . '
44 M 0 1411
10
44, 40 'lla- we
0 o..' qa- ~ *
W; M 40~ "0
o~ g ~k 0-3 'a.~* .
60v 00uj~U~ 0
t-
13 4
4-~ 4, 4 Uad 4. 23 W .
* 0'.. ~ &0 w 06 .
- U 6 O..
j ~5 :4j W.
1..
%.*. %.~
v -
-. 0'. 43 1 0
94
* 40.4 w c aU.
. 08
0l V!g~ 4- 44. kA a.
14g~ 3
U v 0
~~a 0
00aI-Er hul 6 u~a 0 u -.
V 0 V 'A
1b. liIO
Os)
3 4
92 0
u - -
00
.44 0 0
.0 400
*0e4
1.0.o ~ad
'4 * a
O 03
404
me4
- .U U96
66, IIc log - 1
* ~4a
0 a 1
.0 1.
6 . A;
as u
iI -L
ti u "a :A 0
4pSA
16 Ohv 0
.x--
97
r, m.4-N 0 V
%M N . 41
0
AUr. vA w a 0
Nh u 10 1.
c 1~ . -. . 0 V
b. - ~ t N -R-
M .M
C6C
W A a 0 0.9 u 0
CA0 a ~ V aO )-0
M -6
00~~~~~ 0 0- . 0W
.- 0 . Iw 0U
- U i V i -4 Q.- sao~
*N 0
isi
. C- *a N
aa
u c
0 v~
4) w V 04.. 410 i- d3._GoO
o 0ý V 98
*1~~~~~~~~~~~~~~~~~ ___ ________________________________
In
do 0 a im 44 O 4
0 Im w
a. %4 I. wPM a 0 v IS 0
~~. 0-~ =~ -Av
o 0305 0 0
% .4 SU- 0 ~I ~ 4
U 3 '
0 0 1-. V -
Is5 Z U
U 00 aV -3 R. I - . o -
M3 a- a 0-0
0 0S N- >.00
r. 4 0 1-0 .W .6
00 00 $; .0
4-4 *- I"0
06 Id
- io
1:1-04 -4)
0 'A
'.4
a v.
0o w~
Ow lil'A 4'A
0 a
I-.a a 0
a 0 *ýa~ -ar
o.,., .~u ~ a u
I. l O L
U. a ux a of~ a 91. a .n0. 0.
C .. -- 0 11"L: *
4, 0
v 0 'A- C~ v, v~0 A A~~-. I-
j No 0'. *a C6C0. ~ .
0 -
40 a. --
- 4
X 44
0 A;iua
w- uC
0 l 0 0 u*
Uz 0
T 9- a ..
0 0
a-IU'- 0 alM aa' a
*~~~ u~**
0I Aa ou eýv
'.5 OU .
'A. AVa u .u P ka.U.0.. *0 0
Q 4 4 g
*~ fig a.
Ah~LA ,0U41 j
V. VAf V o 0 is.
fi.C 1- 4c, U 00
C>.4 .~... a U 0
- v a- M-0.
91. >~ 01 6 . 6C7 u a. 4- 0 1u 0A
V -5 A L2 0W
o; 4A 0 k
v 72 v M .01
~~ .1
1 0 5.*4 0- .. -21
A 00 a-z 0 -~ .g,~v
00 v- .
A 0 CIO
g~% 0.-
Aa 0 W 4
0.0 6'- o-~0
a A.0.
U 4.
4i v
101
UC6
-,ý 44 C P.-
'. UC .0 u
.~~6 0- C -
-. 0 0.gb
gz
CX 41,C o . 0402 6 14. 0.*J 4' .0- 4-0 .0 -- .0C "V -. 4. 0" C .
0 i 0C 0 .40Q
-00 o O IC.0 a-* > 04V0
Ci~2 4;U vC 20 po C
-" .0 04C.4
CUU 0U0.C
4) 0 CW
0~ 0~. U% !:00 -ý
u -c 40. 0% UC0 -- 0
0C. ; 0 C C U 0 . .
C6 - .C - - iA 0 0CC C
A 10U~ 0.N* A 1%
U ~ a- W OC
102UOC
tt
P04
L& #jF:
6j M4U 1A4N
w .0 )- a
o j6to
I1
a~ o 0 ). wC
U~~ 00"4
"c Q~ 4 0V 41 v a
1- *b0 c4ao
w. c 4~ 4)0.4adr- Ak
gu 0
u 04 0O, -V
~~~~~~ 4O~h" m 441~q
* 103 go n 6 4A.C
0 1.
C o~ a- . a
*0 v 00 0 4S a0
s..~ - -. . -r
04 -.- v..
* 14W aJ 14 00 ..-
"I.. E - l r-Cr16 0. 61 r 00 4r
J. - "
oo v}. .0 s". j)O..aa
o -c .211*2 "to 0 6u v a
1.0~= illflO ~ i r-. 6E w0 I
o0
0 0 00 U C-0 o 0. a,, IO
0.- 01 1. &W1
C 00
ex 4w00.v.- - U0 '0c z.1 0.
L)
u, -- u z 0
~I. .'m.
104
0 -U
V) -
I.-- CL0 16&-C LL .N 0 ..-
0L.~ U~ U0
91.43. £5 V5C -0toA .'v A I.-'fl
0~~
b. Awl bCU. - -U0 .3 ~ .o -'m v
16 w J 0
0
G. U%. .V~U. NUCo*2 0 t b
um 0U 455- ". 4- Li ýS -4 , .
3. 0C a ti 9: -
C -. a .0 40~ %- . it IUt.0 0 - Ll
U ~ ~ 1 5- .N9
- ~ . -.0 * 0
z% .0 A- U v"
0, 0, td -0 93
N~I 0. V)2AO U
oi~~4 - CU U 0 . A
u2 Ad-£C 0 0
S -a 00. 0..U 5all0l0d
Oj OR~' " I !IS
S. do 0 .0 4 'u'A 0
A a*
105
I r
ac 6
4.
4o-
o I- a0m r .
400 5 4~
OK -1
0-. l 0. *.
CO 0- '4 1 Z
LI U C ; AIl 1
106V
! -wl4
u .1 II-. t
0C 0 waa
U 04~ -: a ~C
ca> -~.0 be K
0 O
0 9: CO9
- l -a0~~6 a
40 12.V .j 9 0 U
"~4 .
0-00
U .0 wuC&.-
4 id~ A4 41 140go~~~1 r6 u.- ~1 '-IN i
MSU .~U5 U107~
cc
0- 00
0~ I. *
A 1..X. Ino.t.
c~ u~
4,J oI-~r,
-- 9 t
v *c 0
I.-. .- )
~ ~ 22r
-C i. r-
ww
n -, g ij
c c, 11 jr
01 0 0 J4)-
.01 MIS eiU .-'1-.C b-U ~U ~ 0
1% *C108
q* 0
w 0
49C14
o116,
4wc "9
4 -I.
bo~ j 0 r0
0 6 CO 11
4-a of~ ao 40 be
in vo % -a 0.
V r612 0 v~O
C'!- . "O 4~ r .
W Into o 1-4 ~ aaa
is a-' u-- -oU3
* . 0
V I VLlt
r- : .
Ad
It a~d 0
-. 4 U.40 v
.c 0. 1" Vof - I.. a
40 v 0 u - Vý
w 15 - 0 .--03- U O. 0 ID
I ýac
aic
10
a a -
0..
c 4
4 .4.
Si. wdI
job a
4.J -. f110
4.0 4.3~ b~ '.
*0.-
1,4fl
A9
00 0a V -.ud 0 w. E
4) . A 00A 4LI 0. V I- 0 o4 ..
L') CL A 0 u.
-4 0
1 1. 0
od ..4
aj- U .4
w ,v -.-
... a -41.
044 an P-4E-
u 446 E
vI~ .0
N* I* 0
.20- 0
06.
A IL
u
0 u1
0 w
A g 00
I- U * 112
4-4
1 AOh r
40
-4 .4
0 0m
vS 00
j- : A 4A.
0 A N
am *'.-0 $-a-4 INS-0 o
"" sa1 1 4. 6 0
I. *'~ a -113O
0 k a U. A' 0
U C
000
f4 wo 2 a ~a~
U u Z.'~O -
*.d~~l 0CD' a~ ~CV a- a- C 4 m
40 ' 0N 20 014-c
It ,m 0 ýo .0Wd~ ~'~.wL 00 b
oa Wl 0U~d
I- I- 0 b~one "a
c 0.
0 .- 0 11 0,4-i 4-. '9 *10 1L
so A 0* ~ 0 U
a u
" "A-
114
W1 q *1 4
4 4
.44
*00 4
2
-WI
M- 0 so
610~ 0.4 .4 )'g
V1C 1 1
401
... . 4- " W',r. ,v 0 '
- aja40 U
16 4 4 a601'01 0*
A u a
#A 164
M-N I
1.0 0 :8 N~ O '*
14~~ 000 b0, u1 a w~ a wN
I..- 0 1.6 -
Zb 0 0 S. 0
-C. 0U . V-0~N-0 -11.0 A' .O0 '
0 16 U
0 NN 0.S
0 00 Q3~ CN O 4) Q '.w.i
4) W 4AW -4W 0
CL,0 b. - : 1 1. A. m a '4 0 *
*~o ZO v.4 A N A I6,! N
m 3 cc mO~ "-~ .. a V 4 'A 1b.0 )-3 04
1, 2vi , CLaUd
V4Z .1:
12 62 OWCcW
L.j 0 w. 0 44'O
UY 4.0 . ~ g
V 0
LW .4 . 4 04.
116 a4 4 V
66Si
.4 S. in * ý Ui i§-~
QbU S *aS It
CA a uV V~ U u C1
* 0* 0.0 ~ A4 -d
S 6 0 % .'i4 IV, v:Id4 0~2Z ýM . .d61 -
1 .1 tc_~. .~40
~4 04 6 al.1-10 A 0.6i4S 0 14- - 0S 6~
a; U
OO'A 1 4
10 1.A~5 .0i~ A .4 i AAS
oww a-s u 'e-C 0 a.. 0 Un S SX O'0. S
0 6 4D S.S w i 0- * iu 0 .S~~~Z~ aJi r- vdi' a 2 .. 0 46 -
* - w as C 0 A -u ui
0i Go ai -SA * 01
.5. *e"0 ~ i Ol r. 0 ~0 -I
*~ 0 Si U UA Si *Z -Wd 0 V .~i
14 SaoiI 0d S )sl~ U~
*6 VA IL di.-. .S~.4 41. ~ 4 S5 4
-~~H Si.1-y4
Jii~ u, Si.4u *
* - -Sji i 0 W '~
-A .0 34~
* Mm
U-- - -. - IL
0aa
.00'
- ...
46-
03 W
Uow Ii-
I~~ k .
Pd U wtg 00,-
q5 4
-U$
4%4
;.4 wil. -.
*0u u0
,.jU 0 4-aU4s.0 It -C r
a.-. go--au 4
~...oa 0..
0 .00 M 61 0 5 -Ma..J 44u)Q
* ~ ~ ~ 4,0 *k.0-4 m- .'
~VO.( r.0
0*OK 00 a
0. 7 Z 'C4 -0 1" 1),*%
-~1 o- v4 .A~40 ..
a -aV isa 4. *aa
91 4444*4is 4. ia- N 'ai
o-~~~~~ .. ' %1-~ V a0. - %" A
'A 0a.A-0 4
.0A A4 * 4 a C.C A
4.4. 4vI 1U 44 0UI .~.
a4 6-3 N a a
0 61sa-
Ile 4..ý MMo 0
*11
Q. C 0
0 0
4-.)r
4,
*
0,*
c.
CC
a
v a
a
4- M
-aa. Go -
r-ar
Il 11
0 e0Z • ~ ~I. r .,
In
to 9: w ..
,116
"~ 00 V)
7.
* 4 0. -.
121.
-404~
0 '4..AM
__ __ -0 s w 0
rv1 0 u
a~ 0 -0 0-e
d eU U
01 U6l 0 0 I
In"r6 r
01 u0 -
u6 X o " I-.'
V . .- -.
0.V~ 6 C ~ 0.0
6us 0.0 1
93 t 0 *~ qU 46)~Ln ri It f ii..vIivIi
*00
a~ 0
U- 0
bebe
4-U V 4
a: 0 eV 'V
Ln 0 CL .4 10 Old0V
4' a 06 0
*00 4
0 a- O.
a 0.0 04
0 IV,
Lo*0.4*v
4, -0
'a .~ U~4122
- 'C6
~Pon
V4 0
S.l
f' u
~. a
:. vIC- 0 -'
-. 0 j
00 0
go )-A.-
-3 LO .
>,.v
.1 -- 0"S
L.J ab. $o "LW
40 ~ u
w o.w ' Co S. w .
u -4. *a
-0* u4 14
J 123
411, %dj~
UCL
L-
--r o o-4a
at va
0 61
A-IU h-Ej
alf 4.-4 3
Ad .
w
124h~
a 'A-I!v 10I41
-4 16 - - 40 ; -o aa
~~I. I-*
fi 4.4 41U
4..:
0 -. .. 0*0 A 0
*i. d* o 2 d00.4
*0 r .00
60 0's jA 0
'0 p" P-6 q
4..4 ¶ 1 2S- ,4
0 u - u:
44 0.
06 .
~ : 0
A O U
4 4
A ? 4 aa. U8.
6j U .a
- - 0. a-40kid
126, a
0 . 0 -e~~ý a 0 u 4
.4~0 #*Ai
C6 0 0,j.
1. C6.0
0
Q~ u
sm.4 , 4 1.~
4 ..
W 40 *c-
V 41
1 .410 (a 0.
4,L~0127
b. 0
od a
41
so-0 96
Isso
In hd 0.
to
044 of C: u
116
06. 0 A -VkA r-L 0%. 41:6 w 0 a
40
4-0 001I-bi 61
od k
6j Wa
u)No
aj
fe.Silo ri00
128
4F
0 3
0 -
* a. me .~ op ,
j 0 A C
-a 60o. 00
m~~a am 0 D *.
.4 r6 0. 2 a1-4 0,b
'I'
:0.. I-
IS iia 4.'
Oz,
0-4 -A A
129
040~
00 u~ I.- ~d.-44* 4.r~U* b
V~ 60.4 .0w
rn-i .~ 1 ,C
i~l 04J 0 * 44
- t- 0
0- 0 0 . ..
W~~~0 -4.J -4 U o
oWý - 0 60 LL,). a-
160 -0 U 0
4. U. v 4*2UU
04.4. I 3
*l ..43 I.
W-i a z 196b - *z4ý.
60 U .44*60.0
10a, a,
F0 - 00
z 0. a4* Q
a - -0-- A I.-6'~~~~ 40A' i a 9
4)961 - 001- - I
-0 biu 4 0 4* *'
4.) J1.0 v I v .
-O 0.4 44A -0
4"'I I4)60
dU a. -U
a 0 *
130
OV. 00 0 3s
a1 C41
.1 A -a '0 ro- 0 -.
-U a.J2 %
A 00. aa abI* 0..aie.
0 a Vv aU
0L.
)p a"VOto$.6 a~ uWi -4j In 1. ..
..4 0. v " r -, -,0v
0 I2; C_ 11111 a 2 .E
=- 5104,. 0 r1424 *i 1
POI 0000 461
4. 4)0 me ,
"'a
0041 h 44-*j bu. .e.,U~
1310Gý,
'[a 61 go-
wo .- a
la
66 Uk
S a 4
a
$. ox- k 4
- MU ~'4- I.
44.o do .
*00
a 4. 0 0~ '446a 116 0. 66 w
'Q -00 .4 U
lp u.A1 kA 41.- ;:.a
616. a 60 44
600
* --
.CA UU
4 Ala- .
0 0
3 - * .4
'A'
p.. In
aa a9-1 k'~ 6)*
a~~ ~ u f ~.
.01
ava
'4 0 T
133
a4 0
0 Cu
-03 'A 00 0
£0 v- -4 a
C6. '. 0 i
4 0 e. 4, .4 U
%. 04 -
NC Cm -k4 U
r~ r - a I%.-4
0 41 .01 = -SK 4 C: ON L ,
4"~~. 0. -om , i
3~10 M, 0. 3 4 4..
01- v 0. 0.0 0
40.0 a
aa. a -
4.0 ar -~- ~-ZA o1 I. 6. .- I
a ~ - " * 41344
0.1
00.
'A L- 46 .4WU
4.6. -1. C0,0 L. 0 - V
r an 0.-0 0.0
0
17 C " r-. I
2j0. r C q
v 0 -- r
v In,
0.0
0- 0 C L.& w -1
o u~a
4 v .0C
(A o *.9~1
C: 0 a
I1 - -
i n 4m. A '0 1I
* ap*Uu-.
w ___________________________am 0, wA r
0 0 0 1.
r 0 ia 0
a- r
- *
t a.o :-
r- 00
10
a* 0 AdI
0* I- la.
50 s. a'
0 v z
.0 0 0 0
10 41 .4
me C& 0
s* 6-U V1.9aa~ S.--'- ~
0 - C2 61 041 ~ -
* o~ua-1 Cr
4b
0. 13
C6 UU I.-ý0 -
a all,~ v
~~~l a.O ý. V . 0
0.0
0 054 6
-0 ^U IA 1
0 0 -
1.- 1:6 u. ?S:~ owr.
*0 k
.. -~~ ~ 4"5 4.
*0 0.0. z 9 :-~~~.6. 0 a O0C 010 .0 N .49-
0. 0
>1-0 0 0.0. 0 t0
0ý '4 0? 01
41~ .1v c> 0-~
L; 4 u 0
I, ~ ~. w -u 7 -
4 0-4 .d* u- 0 ,0
0 137
a '0 0 0
* .
C2 - ' *-4) 0. m .09
0, u
0 V.0 U~V 'A
-0. V1 1.
a *V .A !. 1
0 La ~
V. o 0 A' .
4. .94-
0 -. 4 a u r
oz u~ --m .. 0--0 O"c (J.
V ~ J04 W.,-.
.J 44-' '9 138
b_ __ - - - - - 1
-~4 U6 OR01P4 to 0 '0
* . M - 0 0b.4 -110do . 0 *0 ..-
-.2& .a2
e:0 a--U S. at "V -
0. 0 0 O
"a~ a a6 0
I. d bdU J 0 .
A 4J.0
u.. a., ý a .u I 2
a .2U b. -43 1: 4 . 1
41 .. S0 L - .4a. .fl 0 00 U. . &10* 4~ I12 a N~ 11l, 0.1* 0. 0 $. so k -- .00. '61 0 a 0. >~ va
Ln4 UOQ: 0 0 1
0 0: 'nI
6M I
-0- 6 a
4.A . .' a
i..b. A ,0 u -
611
.10P - .aoM v, U
ý 3 C
C3*:
&4 40v .
* 1 44.~4 U -
C6 la A t"
4j' au..A -0 pl a-
U. 0
-~ ~ ~ .a4 0 '''- ~a00000 Z3 .0 ý L.--4.
0 -
of OR~ w~ w 0 0 0A
0,40~40 a a.w
dO~ . 11.4
Ln g Ag 1- a A0
4
am~ .40 C
0 *0
-W w cO
a M
41 *U.0. ~
0 V-. 93 a~
It 6 .-- U -
A 0 . a~A
ajEW 3 v 60 6
F* I N&
I: I
140A
4i
CI. 'A 00 .0 5
U0 A. 0' C
5.00 %WU. u
U id Si 4U..04 .4-1
00 00 pn V. U)A .g JS I.
11 06w.S r. v30
lz a,0 u.a 06
6,~ c 1.~ ~p Id009iS 0 1I ). > b4~
Q P9ie~ 0S4~~~O so v ~I 0U.S~
'40U -, Sii
.. I aI a
1i0, 0 13, o
0 a *
S IL -
-1 0
.44
ow
4, -4141
V, Ll.'
- u 0
0Il. 06
.1j 0.0
0 ~ 646 0
A 0. a 0d>A 2 13 cmo
' 44 U.4m
.4 j. -m 04 a
1* to ~ ~ '0~
V UJ .0
4,.
lo .4 ; vU .0 4o
at 0
- 0 - a%-
o aa -0.40
-oC6 4.1...44~4
L-1-
44 9
J~~ 0'00gno
I- 0 0~.
w 0.
a-n ).9ie 0 1420
Par". V..'4. 6
W.4 , wta 1
WI .
4~~~ 06 a~ 4 U * *4
Goj .0r).vo"a
i0 4D0
44. al11
v 0 ku 60.4 %4W V
a 4 0. A *. 0 *f O 4*.
.4~ *.4 v
.40 .
42 aa I4*1 4*0 00 !O
G 0. W OR 440
a~.
*10 Vo.Z
0 a,- Vf 000 a c') 0 0
0 o ~~.4is
P., 4o 0w 0* a*~ *
9 6 I n I .Z* 6 so
A V
IA* 0..44W~ ~ 0 a sm0-
10 4*1 *,a]'0
A *
143 .
-I. a
0~ U A 0.0
~b u 4-44. 4.3 U44 . 3 *L .4 I
a S. "to 0~
ý 4.4 0 -h
a. a
In
u I. S
U VI 0 0 a0. IV .Z'
is 04 0 0
P I 5U -. 0.U 10,Au 4. UO in u 726
u .. LO -10
CA 4.-
WIm
~WI144
1 .41U4 0 4J
SL 0l 0 A r0b
.4 a. s. sab "
-4. Is .
'Ue
- a'
000
I . ug4W 4 ~is 0 w u IG4
04 U 41
ad4 U -0I -,0 'v
a 0 ; au-4O
0 Is 11 L5...4
14. 394.~
'V 0~ 400
P- 'u : b
00 *~* 14S JA-a
4044
4r4
10 In
Ul
ad
.. . ..
a 11
IL.
-a-.
LJ_ *.. ...4
.... .....~1 .14 All
* ..4 *N ý 00
.4W1
U 'A0 41
000
* .... 4WV,4 A % .4
& .40 04
"V41~0 4
*~ k0 A4
.m 4 . P-.4 61 .
U No to r 'v ~-0 v
7tM 0 ,. aa. .0
04 A4 !: -
~~.30
W. LL.4 -
14 .4 4.
0. . 6
V4 k, M4. 1
-0 0 W0
mgA I to
147
*~~- aca 61 3b-'~Sl~ .. W.
A~~ 0 ,.o .
a CL. 4,
to,
;Z 1.. 0 41 SL :
w.. " 4
v w
0cCL
04 ao .
CA*~' . 9:0-S to
I-- U 44 C
- a u
v a
u-p
-. ~ S. U
148
Mlot
14 ~4~ QU
0 r 4
U d3.a L U4 4
.m 0 w3t u%W~ v
u~U
w- 00 0
*~I 0. e~.5 0 1441 V4 A
'A1 0, of a-,
~~.lJ~~ 1.014 .
a 0 -
C 14.144149
-. -0
0--
0 0. 0 0.
fN a C u u4
6 0 a 0 N r.0t
6.00
a 3 3 0.. r .i
.0 0. .0 0 0.r
-0 .141
o. 0 ..
N~4.AN tj
UON '-
* *4 u 4N4
C6 C3 0 Q
N 002 0
u m
o 4 4-. 4.
6) U 0
0 . U.4
0
A CC.
U3 QN 4.4 . N
IISO
~0
f-4 .4- j
A -.4 44
c 3.
0*
O N4 00 0 0
4p a
.0 0.
414 4~
0 O.C0
I0 'v . '
o U~ W 04 ).-Uin4. 4.04- ~~~.4- 4.- 4 0 .U .-
I- o 40 U.0
On' .4J C *~44
CACOO14 aA .a~ aa 1-'CL
I. -j
10,$.
- -~ ~.. C~ 1431to 4
-- i a~ C L04 4l. C rC
u z ao a - IC.a - --0
S. CC #Ii.
4.s m.U .4 *8 IV qoa co0 0
LW C u a
*: U 0-C 0g
0a IN
u C. A a"IUI
-d C
0 .aC- a. -
.0 v 0 go-:W. w- a a v
10 2 10 0Ca A40-
-a -. ' 0~ OZ 2U *vV 0 OC O ".g w
Gfl C 1 4 U ax .0 06 w oGo 2U4 4 jA . ,a
2- 04 CL. C ,. u -00 44v.4 0 w0 X, fi
4..~~s 0..4 0a s-.4 C 4'''
a- 04. *.~44.4 - 4.0lb
I152
Sig
0~4
Af .0 .4
I4 X .
0~ 0
61 44 -1.- 0. v A93
-~~. 4'4 0u 0 Ua V 0 A 6J 'i
'A . ~ 10.10, . S auNC J*d
w. 0 'AO. U" t. V
A 0 u
0 41 6 ' NOb. u4 C
- ~ 034 -a)~ A.
o 3 *. U. .. '
as AO--. I 31 -a4 ý 0
A~~ A4-I
P.O z Vi a 004 0* '"~O*
IA
I-~~N~ 4'.4. u4
96 - 0 9.40
II; 'A
'0Mo
153
'04 0 . a 6
*I Z
CL am'-
v U 0
wo w
a -. -4..
0 -4
u .0
10 U £04 A
0 * 0
e4~ 0a.A 4 r u
A 56..20 a. u4.-01
a n - a 0
4-~~~l 00 -p4 '00L0 1
0.4 Ii.
S.,. - -
! j j: a' i . &
u ae aU~ aj
a- V 161.0
At V-- 4j
~ 4dMI .. 146
A lb O~U '
4. 0 6 4' ~ .-
.0 u~:o * *.,O ; ~. Ugo
U, ~b a.~ uU .a- *~.~i uaaa' 0
0 0
.- *- U1
jI~ 1111 p.
P1 . 4.
61.
SN~0 'm.
a.. 00 4 1
M"ý --
vu
aI 0- 0Zd
40 .1 u 1."I
Is I
W ' ld.
Iii VI C
V 0 g1 I a
6A6
561
EUI
"'oN 4"-•
.11i
p,.
j-N0 1.-w~ U
*S
00*
-W
a *a.4- N
64-
! v
A mu,
.4.. * 4
0' ~ O .4 4 I * d #16
0.4 ~ ~ ~ r v .4 4~.4
4-~ 0e~.&. - ~PMU a *~-~oa * *lot
* Uin.4m.~~ at . ' .. £
4-4- .44 4 .4 ~*~ u.41 Ohf
V0
mo 00
II .Iu2. -u4 o .4
4 0,
A N
0.0.4
. 1. 0 p ~
.4 pollbd0
M; ~0
* * u 1u 9
* .4~ .4 . 5 a.~.4.
ata
0 0 0
u a
@4
A *4
0 00 U
4., V
4 0 061 -0
- 160
IAA
.44
ý4. I i4A "4
j w~
hto
U: U býaA
*0 V
'a044
.4 56I Ii*P496010.
ut u 6S
~ !.""~b~~161
A.-'a-
I.~G FA u '.'
9..
444.0a
* l 4m,. 0 4?. ' r1 u~ 0 a
*~~n 61~ a~ -~
-, .0, 404 I=,,a
o. .a. ** W.a %I~.A. 0
at 4,U00 O~. . ~ .
0 ~ ~ 'G.A.
20 ' 0 4;
- '. a - 01
06 ... a. 0 14
0 J-'% C.! ~ U ~0 a0- 6NO b, * ja -
0m U 4A is V
no 0 lzao v
1624.0
II
r aa
44
I.. '4. u4 4 1~
a'-u
'1 44 a].
c Is 0A
~~. 604 -4 xW O
64 eU . .4v
o1 *0i how 144 0 4wwe 0 ~ 0 k~ is
P4
.1 *'41 1 *4-4M A
V 04 4 *'4 *
IL aidiU
j '4 . 4 02;'
0 014 10 4ý a U4444~44 h0U44'LW
'J " d
.4 u16
W -0a 0A
0 A0
4.0
ha a .0 u -
44g. O4 a A* 'V.4
a~ a i
~~1 0"d-~*~ '
Z. 6j kw
aJ 0 0 A V; k
* ~~ 04 J
a Ia
0 ,~ Q u4 -e .sp o @
'0! as -0
16
'AIVw-w _
IF .
a-u A
0U
*0 '.A
- I.
1 V)
A n
r- 0
04JD
Is I w% a
i6j Oak
se 0.~~. w~5. 6,
4..04 a i4. .r
4
a . o -a a. ~. e.
wu ag aý.d um4
..- ~1. 0 V 6N* ~ ~ k
A . Id;A N ~ 4 .
*~~~ u~.
U .. a. .A I.. ow
- a ~ , P.
166~ 4O4m
"",.A-1~
ad 6.co a 0 " AllC
Q -K - -4
*al
I. ,11
* ,..: 1 -. - h g-0~ d a
~U a. aa. o,4 .u .. =
u f
in .
.4- 0,u.~ji i*.
iIN 0
167~i
U N
41
g,q u g a
x e c
U "I.. "°
a4 4
UJ0
aa
0 - 1.
I.. - p
• *• *
ka.9 0 **II
9 i a ,,
4a74.S .I I+ +.-. + + + + +
LI.4 (b
IS.
0 61u
ýA 41
m U-
Id.
A aa
- a~
(A A. 0
9 08. a- ~j0
* .40 *I~i.
o sqi~ ~. -169
f...,
ol uI
6 0 I.-
-9CU
0 602 a~9~
CVU
0 4
Cu of ).4 a C
C rC
1 1
-, CC 0 W .. 'N .6 441-P, 2 AUU C 1
1. 0 U 0 -A. ý 4Co gow c do IV 6U 60
-u 0. .I 04.a
.j L)
017
0 _
i... -. U u. LI
AA
441 44j *a 1
'4 .. *0 414%.
V) 10 aa u
. a4 4 61:6 0. CLO .-u-*
o 0. : 0 " S. 0
co a, as d o j
a 0 0~ Q .0 0,
60 U- u 0 c A.4
~0.1 Q a
40 in - 6o
I0 *~ : 0
C4
-Vl .0 oa.W
-. 400 e .4 .-a 4F
44 0 a -
0V O40f al No a
6IA oA . 0
So, .4 14, V""640.~aM.
C6 U. 4
~0- 4- .
%4 . 0,
1; U
In4 -t 0
-0 ' -0 v -C
0 j41.4 0.
14 -0 -
VA Io am*- 5 1 00 r
AU ~ .0 aU -4.4'
0 '. a~a .0 .. o x~~ a.
0. 'AO~. ' .IM 0 - 1 .2 -0- V-
rl 0 r- C .~4N~ .0 u.' -0~.
* 1.00 -b~gC ~.4 .0 4 c 9 ...
A 91-0 -CO'U.J1 u 00 - C61 .@'-4. N
0N" ~ ~ ~ ~ 4 0 U. =~ r00- '4.0. 00 W .Q ~ -We
in u u >' a .0
0 - - .
uV 0.0
v 400 00
A Q - - -2L.4
3... a ~
11 .02.U -0,00=414060 No "
rI 172 -O
did
A 0
lb d.4
C C
0 lb
w' 440 1. 0
CC 6, LS0.
v- C6 w
OM 4u .0~~40
~4, 00 44' 0 1.-
ol 914. dE.0*U 0 a
1. 0
L/
0~ ...dO'
Li C. 173
-. - - ----.- * --- - -- ~-- .- 1k
dA-
* 0m *0
00
.d ,w . . ~ 44J 1" C
40 -.
goo- Cot,
6 0 CI
0 U.
r 0.u. * - U
I- 'm
4a .u --Z s
0. 0.-
00 v. 0
v6 a.
C 0 00...
fl: '- '.a
*~-. A.
a -M 0 0
~~ 1.-M A 0 0 -'
fti c , i
174
a C
4.7
a. u a 6W I
0. a. #j 0a
u a
13 C74is 0 1. A
44 a -'444
4.44 - *4* .
C' &
41~~* 16 ft '
C 4 . f 4. 1.-~~ 04.s'.
* * 0
-0 u 0 a .
a 175
*0~~ 391#l
Z0..
%.4 w 1%
0.e0
0 -41
14- ur
01.
04 9
0 ;
-- V 9;1
00 v- iL
d 4, C
0 H*1
IW
10 - '..'.
Ad' U1-u U U40
CL a4U l 1.0 p 4-01 ' N'
.. 6 1 ' -U 1..0 .1'd'
61 0 oA O.' Lni -
176U 44
Sk4 404
U u
-4 -1LA
d V3 4-
-- .4 q
91K .0. -$.. 4D
ýo um u , ':
.J4 a r
o so
bl 0*- 0~
CL ..JO 04
-'-W
000. u0- 4o m X I~
-4 0UCA -. - . 0 3
- 1 - -1w! .*~~~~ 05- *0
04.) do
* O.0. 01
el 5 0
61. 1444-0- 0 0C U ~J.4
- a * 4-I~ 177
- 04
- j w
> alI- ~ 5
44 im A C a L
LILI
~>0 '0.U 4's
to~'0
o 0 -0
- -
LPA
0-01
an o
-I0
U r..
to -"Cn
* 10.78
NFL .-- ~ -. r- .-
f4 0
o u .
Q. t.
0 *U 6
~~4;
C440. c CqLv I- P .
OO*
40 r11 IL 0
4)N .1 12
0140
0 00
~J0
AA
41 .0 0 w
-u 0. 4J 4U
X , 1..4 JN 00.-
0 o i .1dbd 41 113a~ ý4 jj
Fibi000m
17
100
It,. 0
%4 0 0~liiCLCL~
0 0.2
0 '0a*
to &'o0 "
0 0.o 0 0b- l. -.
40. " '. " v
a- o. qf U
r- 6 u 0a r-- ~~~ ,. Ia - *
a~*~ ~ .io 1 04.D. C 40
0 *- W-'r'
0.- 0X u~
).,a ov -9 .40 -V
1801
.1 1
X- no
61. 4A ai 041-'
J40r 4j b.f 4
6.~ .~A U * O
110 w4U.14 6, k0*1. 4 - 42
*OU~.4~.f4 Fn 0 4
q3 AU C1.9 IW *JI -0
V 14 10 Eu-44 ~-- 4,~ ~~~ 0S444 4U O C
3UVU U 4~.u 4.4L r CA'A, 1. 4,4.2 .1.S. 4) V04.6
41,4 I-A 4.0 C.. 44n~ z 4 va 0 Z 0 0. !~. 6 ec4 "A.~ 4~
0 r a. . 6 1 W
.fl.4 40m 404 34 .01U u 4
'A~ ~ a
t- ACu4w4
46~..' 4 ' 44- 4.C~ ~ .A~ 146
w 4*.4-C 0.-
u s 2 ~V)
64 0
pa f: 0) .0
-40 * ~ 4 0
jf
I, u a O w *
ot
A0 UD fl "4 4
4t. *. 4 "U 0000661 13- a b, .
W. 0 j4J4 CL 0 0 0._ :6 v .'
U'0 04. :1Ma IV, "a~U 11 0 xz
4 Qj .Vu to~ 9N
9 OK ~ 9 4J.' r~ r b. .1
ME L.~ . V 0 CIL "ro-0 ~ 0 ~ 4
N4~~~N .-. ' V0 "0 V - o)
N6O...2b cc* 4.0 C0C
.4~~40 0 -0.a0 a
w6 *6
4z 00N..4-.t
NA .;
a - 0 .. '.0 4
U 1 41 ~ dI. 0 1 4'
0 41 .. 4 .I
c 4.'
A:1v .1 82a o4 ______
a-a Al -40 W
1.41w u 0 .wo
-0 0.
*~C 00 LW 4
ua -0. .4
el Nk. 0 64 4 .
of"41 OWNt
614 - C~. earn u
0 a 4 . .44 v4 ý;4
1441
~v v
*ý *U 0 6- a.. c
-c 0 6
~* *g- ~. I I.ob I. j
40~4. 0 4.O 10 0I
so 40 41 4 6
0 . LW 414 -.0L
-6 34 t V
0~.
0 -'44 C440 .
-~~" U
4 . C~
44 0 *0>..~ 04i) ~ Q..C
a' r
I -C In. 4* - ~ ~ - x C.63 -
.9 9xh .4 r.~~~~~l 44-0 w a ~.~ 4
., -)4a a0 44
* ~ g ~C.. ~ 0Z i a44 C~Z ~ 4 U 0 4 44
£UU~ 01.44,~ -C 4..U h.0
S 14
.' 0
0 v a
r 16. U-
184__ _
0 £44
.- u. u u
,.d 3 .. U -, A4
U J
!V
Is
to ab. Q
"" C
O0
$. C ... 4 0 %4
*0 bo $. u00 L69 00 -
-u 6.
444
* 4, .- s,
C u * .* -n -, " --
44.. 0 .. .'
00' C IN
.,n
CL
7 gl
.0,4- .
c to5 0
4..0IA.I I-I•
• oo-
,ow
151
'Il
,qwA2~ oil
.4 aU0~I
"a 0. 0
*~I .4bj.0, -o aI
13 ". 14 N. iNO t I
40 ~ ~ V 4 * .
~00j 1. *
a L
99.4~~b 6.0 .
r4. -1 0 L)4. 6A .u-Im >
is4 'a. be
' 4 V4 m L
'.0~~~ a .44-.
N 4J-' . 0 ob 0-~~~
I944 0 4*99~In 0
ýM0 'A- 0 WC
0 UU a
's0 .9 lo h 0 l
-4 -d -0 0 ASf-94 0, 44 -, a~ a Li
4*J IL 4. 0.4 a
o .99.4 f
44 9600
4A4
4*'s 0 '-
99 * O'4~1874
* v
t 2t
I-U
An.-
@4 0 0
I- 'ý .! z %
A. v
j3 u0A
0~~ uP~h.~a..l 3aV4' x s-
-4tj ~0a
r4.
ON )a F 1,.N a~5- ~.4
via "0
I- - 4% A ~ b
4w 0. .a* a-
4.-
q*4
-4 9
-~~~ -,. 0
-- 1
*4 0 I 4 *
U ~ u 0
v .u .0 c4 r0.Q.
'.5.. 0
4d 0b b.~
Is 1- 0.1 X IXL
C2 Q
Cc 0 40 0
A v
4 -q Sin .4. -
- - $"-c u -i -4 .01
54.-0
2Ai* Wi.
*W
00. 0
0
LL .6,5. A
$.- 1; 0 0.
'I Q7A
2 Ip. aa
40 U
"" 46 0
I- U.- 61U
w~ *J4;-v
a %q0
~ s4 'a
U. 190
0 0
wj 4 5.0 :3 i
C: 0
.u 4 , .0 a~
.0 0
3~ U 0-
41. 40 A *
I~~~0~~~ '0~ --0-- g 0I 0* .*
U. 0 C. :!ýt 000.0
c x. W .0
L4.m 0 a4
~~.U~ a-. -0
0 0 w* U40u
O- -41
69
c "U .0 L w -90z t "V V~ P.ww 419
1. -
1.. *6
04 &.1
192
a I0 1
a
*0
-4
N �r4
a-4 0*0
a
U,.. a�w a-
-� '�* Or'4t�.
-�I' �
*0 �a -aEQ
- ae�Q a �--
U* a
U *�- * -0 -. 0 .0
3 � -U kW�
-- - hO 4�' a
�0.- NI
a
Q
a N
-U * Ia0
-a4 a -
0� AA O Uu -
-4 -4* U.:
-U-- a
hOa.
� I-,
.( 0 *eI-ILa w ---. �
.0 h-.4 - 0 .i�f -
193
I
, I.l~. x-*4l 1
q~ 0 -zo
It 0.o -a
0.
CL . 0 . 4.0 . 0I0
~ ~ . Uc *-. r . 0..~
C ~13 a
o~~a 1.. ~~aa 14- 1. 0
- 0 0-01l. a th ,044.4.~~u~ .o
- z-~ 01 v0 v,
2.0.00
-0 w I < 4 11.0~ U-W X0W 0d
0~~~~ 0I 0Q .4~~.0
c C6 041 . a0 1 U .0 'lU,
is Z0- .1~- -. I
- 14
4.
h ~~ ~' 0 0~0
OU 0* U *
L.l V~F,4 .1. "0. - cc 4.93 0 t5ci. C-0 Z.4 *
- w.2. 0 0 WdL - 1 =4 s
0 vUW 16 A0 V- .- ' w 0
.,-a 4.' g'tok gC 61
C6Q uc ad r .A'4
a 0
~a Cft0 C'4 r4 u*4'IA
C.C. .0 61
195
S1 0 a a k :I.I01 . C**A .
0.0 1. vU40N 0
wo .0 4 .
~W. 9: v N O
60 t
00
In x 0 3U .
a0 ON 0 EWNO
0 c k 0O w c L 0 v-
bw U- cCNU *.r
*0.-.C~ . N-a 14 vN I- b.l 0 . '
~ ~O 0 "V VN 40 .0'
"ro. *r 0. 0V~.L N Nu
1N I ON " N 4U - O_ 0-
- 02 4S) c 4u.. 4 x V
09N w 0i 04 C 1N a 9:vQr U0 u > 4 N -0" U,
I 40- I. C 44.0m fe t:t0 r N '0'
4 QN UV 0-4.0...CU U j60 "a 44 1 ' 40
o~~~~~ a4 v wi V'" 0") Na 404- N . O O 4 N.
U~~ ~ ~ ~ i N N O 4. q q '.0 *O N N 44 0.-. 4 40 N.U0
.1 Ia. *_ kOO 7 ýV
iU a00 Z - O O N ~.4.4 UM -U -4~4.U p
w 0.. 4,u
1960* N
00
1: 0
on --
u 4.I-a
96 oh' 0 00
ojw "4 'Ar li
011 0 0 I4-4
CV 0AV
0 .:A J 0
4 u. 6. d ~ ~
- * - n V, N M C 0 >1
-! ýc~ * e 0 I ~ ..
N0
'41 W0 C
o- r. u u
*61
ONa
A' u
SC6
0
10 f-Af
6 6 M 'a
~ 0' L/~197
- cr
0---
wii
*19
DISTRIBUTION LIST
Names
Chemical Systems LaboratoryAberdeen Proving Ground, MD 21010
Office of the DirectorAttn: DRDAR-CLG
Systems Development DivisionAttn: DRDAR-CLY-R
Biomedical LaboratoryAttn: DRDAR-CLL
CB Detection & Alarms DivisionAttn: DRDAR-CLC
i
Developmental Support DivisionAttn: DRDAR-CU-L
3
Attn: DRDAR-CLU-R
Munitions DivisionAttn: DRDAR-CLN 1
Attn: DRDAR-CLN-S
Physical Protection Divisien 1
Attn: DRDAR-CLW-P
Research DivisionAttn: DRDAR-CLBAttn: DRDAR-CLB-BAttn: DRDAR-CLB-P 1
Attn: DRDAR-CLB-TAttn: DRDAR-CLB-PS (Mr. Vervier) I
Attn: DRDAR-CLB-PS (Dr. Stuebing) I
Attn: DRDAR-CLB-PS (Mr. Frickel) 10
Systems Development DivisionAttn; DRDAR-CLY-A I
Department of Defense
AdministratorDefense Documentation CenterAttn: Document Procesang Division (DDC-DD) 12
Cameron StationAlexandria, VA 22314
199iI
DISTRIBUTION LIST (Contd)
Namescopies
Office of the DirectorDefense Research and Engineering
Attn: Dr. T. C. Walsh, Rm 3D-1079Washington, D.C. 20310
Institute for Defense Analysis400 Army-Navy Drive
Attn: L. BibermanAttn. R. E. Roberts
Arlington, VA 22202
Advanced Research Projects Agency1400 Wilson BoulevardArlington, VA 22209
Depa,-ment of the Army
CommanderUS Army Research Office - DurhamBox CM, Duke StationDurham, NC 27706
HQDA (DAMO-SSC)HQDA (DAMA-ARZ, Dr. Verderame)
4 HQDA (DAMA-CSM-CM)HQDA (DAMI-FIT)
: Washington, D.C. 20310
US Army Materiel Development and Readiness Command
CommanderUS Army Materel Development and Readiness Command
Attn: DRCDE-DMAttn: DRCLfICAttn: DRCMTAttn: DRCSF-S 1Attn: DRCDL/Mr. N. Klein"Attn; DRCBI/COLGearinAttn: DRCDMD-ST (Mr. T. Shirata)
5001 Eisenhower AvenueAlexandria, VA 22333
CommanderUS Army Foreign Science & Technology Center
Attn: DRXST-ISIAttn: DRXST-CE/Mr. V. Rague
220 Seventh Street, NECharlottesville, VA 22901
200
I
ei
IDISTRIBUTION LIST (Contd)
NamesC.p
CommanderUS Army Missile Command
Redstone Scientific Information CenterAttn: Chief, Documents IAttn: DRDMI-CGA (Dr. B. Fowler)
Attn: DRDMI-TE (Mr. H. Anderson)Attn: DRDMI-KL (Dr. W. Wharton)
Redstone Arsenal, AL 35809
US Amy Armament Research & Development Command
CommanderUS Army ARRADCOM
Attn: DRDAR-TSS 5
Dover. NJ 07801
CommanderHarry Diamond Laboratories
Attn: DRXDO-RDC (Mr. D. Giglio)2800 Powder Mill RoadAdelphia, MD 20783
Chief, Office of Missile Electronic WarfareUS Army Electronic Warfare Laboratory
Attn: DRSEL-WLM-SE (Mr. K. Larson)White Sands Missile Range, NM 88002
"Project Manager for Smoke-ObscurantsAttn: DICPM-SMK 2
Aberdeen Proving Ground, MD 21005
CommanderAtmospheric Sciences Laboratory
Attn: DRSEL-BR-AS-P 1
Attn: DRSEL-BR-MS-A (Dr. R. Gomez) 1Attn: DRSEL-BL-AS-DP (Mr. J. Lindberg) 1Attn: DRSEL-BL-SY (Mr. F. Horning) 1
White Sands Missile Range, NM 88002
DirectorUS Army Materiel Systems Analysis Activity
Attn: DRSEL-NV-VI (Mr. R. Moultron)Attn: DRSEL-NV-VI (Mr. R. Bertemann) I
Fort Belvoir, VA 23651
201
II"
DISTRIBUTION LIST (Contd)
Names Spe
US Army Mobility Equipment Reswarch and Dcvelopmept CenterAttn: Code/DROMD-RT (Mr. 0. F. Kezer)
Fort Btlvoir, VA 22060
CommanderUS Army Electronics Command
Attn: DRSEL-CT-LG (Dr. R. G. Rohde) 1Attn: DRSEL-CT-I (Dr. R. G. Buser) IAttn: DRSEL-WL-S (Mr. J. Charlton) I
Fort Monmouth, NJ 07703
DirectorUS Army Ballistic Research Laboratories
Attn: DRXBR-DL (Mr. T. Finnerty) IAttn: DRXBR-P (Mr. N. Gerri) IAttn: DRDAR-BLB (Mr. R. Reitz) IAttn: DRDAR-BLB (Mr. A. LaGrange) I
Aberdeen Proving Ground, MD 21005
Commander, APGAttn: STEAP-AD-R/RHAAttn- STEAP-TL
Aberdeen Proving Ground, MD 21005
CommanderUS Army Test & Evaluation Command
Attn: DRSTE-FAAberdeen Proving Ground, MD 21005
CommanderDugway Proving Ground
Attn: STEDP-POAttn: Technical Library. Docu. Sec. 1Attn: STEDP-MT-DA-E IAttn: STEDP-MT (Dr. L. Salamon) 1
Dugway, UT 84022
US Army Training & Doctrine Command
CommandantUS Army Infantry SchoolCombat Support & Maintenance Dept.
Attn: NBC DivisionFort Benning. GA 31905
202
*1
DISTRIBUTION LIST (Contd)
Names Copies
ComImandantUS Army Missile & Munitions Center & School
Attn: ATSK-CK-MDAttn: ATSK-DT-MU-EOD
Redstone Arsenal, AL 35809
CommanderUS Army Logistics Center
Attn: ATCL-MMFort Lee, VA 23801
CommanderIiQ, USA TRADOC
Attn: ATCD-TEC (Dr. M. Pastel)Fort Monroe' VA 23651
CommanderUS Army Ordnance and Chemical Center and School
Attn: ATSL-CD-MSAttn: ATSL-CD-MS (Dr. T. Welsh)
Aberdeen Proving Ground, MD 21005
Department of the Navy
CommanderNawal Surface Weapons Center
Attn: Tech Lib & Info Svcs BrWhite Oak LaboratorySilver Spring, MD 20910
CommanderNaval Intelligence Support Center4301 Suitland RoadWashington, F.C. 20390
CommanderNaval Surface Weapons CenterDahlgren Laboratory
Attn: DX-21Dahlgren, VA 22448Commander
Naval Weapons Center REPRODUCE FMAttn: Code 3311 (Dr. R. Bird) BESTAVAIL LE COPy'Attn: Code 382 (Dr. P. St. Amand)Attn: Code 3822 (Dr. Hlndman)
China Lake, CA 93555203
DISTRIBUTION LIST (Contd)
Names Copies
Commanding OfficerNaval Weapons Support Center
Attn: Code 5041 (Mr. D. Johnson)Crane, IN 47522
&ommanderNaval Research Laboratory
Attn: Code 5709 (Mr. W. E. Howell)4555 Overlook Avenue, SWWashington, D.C. 20375
Department of the Air Force
IIQ, Foreign Technology Division (AFSC)Attn: PDRR
Wright-Patterson AFB, OH 45433
CommanderArmament Development & Test Center
Attn: DLOSL (Technical Library)Eglin AFB, FL 32542US Army Materiel Systems Analysis Activity
Attn: DRXSY-MPAberdeen Proving Ground, MD 21010
Armament Concepts OfficeWeapons Systems Concepts Team
Attn: DRDAR-ACWAberdeen Proving Ground, MD 21010
REPRODUCED FROMBEST AVAILABLE COPY
204
•i-.