O-AO95A 174 COLD RE IONb RESEARCH AND ENGINEERING LAB HANOVER NH F/6 S/12

uuuuuuumIIII
ICING ON STRUCTURES, U)
DEC 80 L 0 MINSK
WCLASSIFIED CRREL-80-31 M

ImllllI
 7

Icing on structures

i< c:

P, 4

- ,,

V-A1 2 2.4 (EV1 21-

Cover. Mountain-tap radar site in Alaska en-
crusted with rime accumulation. (Photo-
graph by author.)
! ]+-"-]

CRREL4't 80-31

Icing on structures,

L.D. Minsk

DecoM8O

Pr.pai fod r

F I Of MARINI (AOLOCY
()f1(
IS (A 01 ()(,I( At SURVEY
1

UNIHIt) STATS ARMY

(()RPS () I N(.INf IRS
((011) RI(,I()NS RIStAR(H AND FIN(INIEIRING IABORAIORY
HAN()VI R NF W HAMPSHIRI, U S A

t I l
A, .... . , .,d .. .b. I,
r ' dII ,r n
-Icj n irui

r" /"
 Unclassified
SECURITY CLASSIFICATION OF THIS PAGE (When Data Entered)
REPORT DOCUMENTATION PAGE READ INSTRUCTIONS
BEFORE COMPLETING FORM
1. REPORT NUMBER 2. GOVT ACCESSION NO. 3. RECIPIENT'S CATALOG NUMBER

CRREL Report 80-31

4. TITLE (and Subtitle) 5. TYPE OF REPORT & PERIOD COVERED

ICING ON STRUCTURES
6. PERFORMING ORG. REPORT NUMBER

7. AUTHOR(e) 8. CONTRACT OR GRANT NUMBER(&)

L.D. Minsk

9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT, PROJECT, TASK
AREA & WORK UNIT NUMBERS
U.S. Army Cold Regions Rtsearch and Engineering Laboratory
Hanover, New Hampshire 03755

II. CONTROLLING OFFICE NAME AND ADDRESS 12. REPORT DATE

Office of Marine Geology December 1980
U.S. Geological Survey 13. NUMBER OF PAGES
Reston, Virginia 22092 22
14. MONITORING AGENCY NAME & AOORESS(iI different from Controlltng Offtce) 1S. SECURITY CLASS. (of thin report)

Unclassified
ISA. DECLASSI FI CATION/DOWNGRADING
SCHEDULE

16. DISTRIBUTION STATEMENT (of this Report)

Approved for public release; distribution unlimited.

17. DISTRIBUTION STATEMENT (of the abstract entered In Block 20, If different from Report)

IS. SUPPLEMENTARY NOTES

19. KEY WORDS (Continue on reverse side if necessary and identify by block number)
Deicing materials Ice prevention
Deicing systems
Ice
Ice accretion
Ice loads
2ij AST RACT (17o6afe revers es tif nesay d Identify by block number)
ILCaccretion on structures built on the earth's surface is discussed. Sources of water are the atmosphere or
water bodies near or surrounding the structure. Ice types include frost, rime, glaze, and spray; properties and
conditions governing their formation are presented. Methods of estimating accretion rates and total accretion
on structures ,are given, and extracts from U.S. and Canadian codes for ice and wind loads on structures are
included. Techniques for preventing or removing ice accretion are presented.

DO I0' 1473 EoITion OF I MOV 6S IS OBSOLETE Unclassified

SECURITY CLASSIFICATION OF THIS PAGE (When Data Entered)

................................................................... •.,
 PREFACE

This report was prepared by L. David Minsk, Research Physical Scientist, Applied Research
Branch, Experimental Engineering Division, U.S. Army Cold Regions Research and Engineering Lab-
oratory.
The report was prepared for the Office of Marine Geology, U.S. Geological Survey.
Technical reviewers of the report were John M. Sayward and Stephen F. Ackley of CRREL.

, . :- --

V_

A-I

: |
m 1
CONTENTS

Page
Abstract. ......................................................................................... i
Preface ............................................................................................ ii
1. Types of ice accretion.................................................. .......................... 1I
a. Frost ........................................................................................ 1I
b. Rime................ ..... I.................................................................1I
c. Glaze ........................................................................................ 1I
d. Spray ice .................................................................................... 2
2. Conditions governing type of accreted ice...................................................... 3
a. Meteorological.............................................................................. 3
b. Structural ................................................................................... 3
3. Accretion rates................................................................................. 3
a. Fundamentals............................................................................. 4
b. Effect of height............................................................................. 8
c. Geographical distribution.................................................................. 8
4. Spray icing..................................................................................... 8
5. Structural design factors ......................................................................... 13
a. Dead loads ................................................................................ 13
b. Wind field in the boundary layer ......................................................... 13
c. Wind loads ................................................................................ 14
6. Techniques for minimizing structural icing..................................................... 16
7. Data collection needs...................................... ...................................... 17
8. Literature cited ................................................................................. 17

ILLUSTRATIONS

Figure
1.Crushing strength vs temperature for commercial artificial ice .......................... 3
2. Relationship between meteorological conditions and type of icing ..................... 3
3. Collection efficiency of a cylinder .......................................................... 5
4. Collection efficiencv f- is determined by an inertia parameter K and the parameter 4) 5
5. Critical radius of cylinder above which icing theoretically does not occur.............. 6
6. Variation of collection efficiency' of a 15-mm-radius circular cs'linder................. 6
7. Icing efficiency .......... *......................
-........
S. D~ependence of icing effticiency on wi'nd "speedfo vai uiui*l w.tevi conotents ...
9. Variation of average diameter and weight of ice accumulation ........................... 8
10. Mean annual percentage of hourly weather observations with free/ing rain), Nor th
Am erica ...............................................................................
11. Regions with similair gluie ice characteristics...............................................9
12. Icing severity as related to air temperature and wind velocit\ ............................ 12
1.3. Correctiorn factor for use in estimating radial ice thickness...............................IS1
TABLES

Table Page
1. Characteristics of icing sources............................................................. 2
2. Types of ice from atmospheric sources.........I........................................... 2
3. Relative frequency of types of ice formed according to cloud type ................... 4
4. Relative frequency of ice accretion according to cloud type and air temperature .... 4
5. Occurrence of type of ice by height on meteorological tower in Obninsk, U.S.S.R.. 10
6. Probability of at least one occurrence of an ice storm in any year at a representative
point in the region..................................................................... 10
7. Probability of at least one occurrence of an ice storm of stated intensity in any
year at a point in the most severe part of the region................................ 10
8. Number of ice storms >2.5 erm and >5 cm in 50 years and the probability of at
least one occurrence in one year in the most sev-,re state in each region.........1I1
9. Distribution of ice incidence for Soviet ships in various seas as related to air and
water temperatures....................................................................1I1
10. Distribution of icing incidence for Soviet ships in various seas as related to wind
direction and wave height............................................................. 12
11. Frequency of ice intensity related to wind speed ........................................ 12
12. Ice thickness for different return periods at a representative point and at a point
in the most severe location for each region.......................................... 14
13. Ice thickness combined with wind gusts in the most severe location in each region.. 14

iv
ICING ON STRUCTURES

L.D. Minsk

1. Types of ice accretion small compared to other types of ice accretion. There-
Ice formation on a surface exposed to the atmos- fore we will not consider frost further.
phere can occur in one of two ways: by flooding of
the surface by a large volume of water or by accretion b. Rime
of discrete water "particles." Flooding of a surface Discrete water droplets in the atmosphere can easily
and the subsequent freezing of a ponded volume become supercooled because of their small volume.
applies only to a horizontal surface. We are concerned When the supercooled droplets strike a surface they
with the process of ice formation on both vertical and will freeze as soon as the latent heat of fusion is dissi-
horizontal surfaces, and therefore this discussion will pated. Hard rime will form when the heat loss is rel-
be confined to droplet or particle accretion processes. atively slow, allowing "wet growth" tu occur whereby
The two sources of liquid particles are atmospheric some flow of freezing droplets can occur before com-
processes (which produce hydrometeors) and wind plete crystallization. Soft rime forms when droplets
shear of, or splashing from the surface of, a large body freeze very rapidly upon deposition, resulting in char-
of water (droplets arising from this source are also fre- acteristic granular structure. Hard rime is the denser
quently considered hydrometeors). Atmospheric and harder of the two; its density ranges between 0.1
sources result in the formation of frost, rime, or glaze, and 0.6 Mg!m 3 , in contrast to 0.01 -0.08 Mg/mr for
whereas spray sources can result only in the formation soft rime. Hard rime appears milky or translucent,
of rime or glaze. Characteristics of icing sources are depending on the amount of air trapped within the
listed in Table 1. structure. Soft rime, because of its much lower den-
sity, is more delicate in structure and can appear quite
a. Frost fluffy, though lamellar and needle-like forms also exist.
Water vapor in the air may deposit on a surface Rime grows principally into the wind, as individual
which is at or below 00 C to form frost, or more prop- small water droplets impinge one on top of another
erly "hoarfrost" (hoar is the generic term for ice after coating an accreting surface.
crystals formed directly from the vapor phase). Hoar-
frost can form only when the air is still or very nearly c. Glaze
so. The resulting bond between the crystal and accret- When the water droplets striking a surface have sui-
ing surface ranges from very strong (e.g. frost formed ficient time to flow in a continuous film over the ac-
on house windows or on automobile windshields) to creting surface prior to freezing, a hard, nearly homo.
very weak (e.g. surface hoar formed on a snow surface geneous ice is formed called glaze ice (or "glare ice"
or on road pavement). The growth of hoarfrost is or "black ice, " because of its characteristic specular
limited by the amount of water vapor in the air; since reflection of light). Because the surface has been
there is little vapor at subfreezing temperatures, the wetted nearly completely prior to freezing, a very
total amount accumulating on an exposed surface is strong bond results. It is generally bubhle-iee, and
 Table 1. Characteristics of icing sources.

Droplet diameter Mean droplet Liquid water Droplet

range diameter content concentration
3 3
Source (JAm) (L) (g/m ) per cm Reference

Sea spray
Breaking waves 1000-3500 2400 4600 Borisenkov and Panov (1972)
Wave crests 60.1000 (150-200) Borisenkov and Panov (1972) (Wu 1973)

Fog
Advection 6-64 20 0.17 40 Kocmond et al. (1971)
Radiation 4-36 10 0.11 200 Kocmond et al. (1971)
"Sea fog" ?-120 46 0.13 Houghton and Radford (1938)

Cloud
Stratus 1.5-43 4.9 (0.05-0.25) Pilii and Kocmond (1967)
(Borovikov et al. 1961)
Cumulus
(cumulonimbus) 4-200 40 2.5 72 Weickmann and aufm Kampe (1953)

Table 2. Types of ice from atmospheric sources.

Density
3
Type of ice Appearance (Mg/m ) Conditions of formation

Glaze A hard, well-bonded, generally 0.7-0.9 Supercooled water droplets at

clear homogeneous ice a temperature close to freezing
(00 to -3°C) and wind speeds
of 1-20 m/s
Hard rime A hard, granular white or 0.1-0.6 Supercooled water droplets at
0
translucent ice growing in the a temperature of -30 to -8 C,
direction of the wind wind speeds generally 5-10
M/s
Soft rime A white, opaque, granular ice 0.01-0.08 Supercooled water droplets at
0
with delicate structure only a temperature of -50 to -25 C
loosely bonded, growing in and low wind speed (1-5 m/s)
the direction of the wind

therefore its density approaches that of bubble-free the shattering of trapped air bubbles as they rise to
ice (0.917 Mg/m 3 or 57.3 lb/ft 3 ). It is also very hard. the surface of the water, and by wind forces. But the
The crushing strength of bulk ice was determined by sp(ashing of waves against an object in the water (such
Butkovich (1954) (Fig. 1). The degree to which ice is as a ship hull or a fixed structure) produces the most
bonded to a substrate is measured in terms of the in- and largest droplets. Only an insignificant amount of
terfacial shear strength. Generally, the lower the tem- water from spray sources has been measured higher
perature, the higher the shear strength. The difference than 16 m (52.5 ft) above the waterline on a ship
in coefficients of thermal expansion between ice and (Anon. 1962). Spray ice orginating from seawater will
materials to which it is bonded is one reason for the generally include brine pockets, since salts must be re-
widely varying values reported in the literature. jected before the ice will freeze at O°C, and the increas-
A summary of ice formed from atmospheric sources ing concentration of the brine may depress its freezing
is given in Table 2. point well below the ambient temperature. Sea spray
ice, therefore, is weaker than freshwater ice, but it may
d. Spray ice adhere just as strongly to a surface when no brine pock-
Droplets can be generated by breaking waves, by ets are present at the interface.

S I
 £000 T0--

so.

-
-4001
40 Oo

*
:: iO 4£
t' 05
0

to 30 0 HRO VI
U0- L 1 0
too--
W
5I R /
0

0 0 /
La - - 0 6 -10 -C
-40 o 0 TEMPERATURE,-
00 -0 -0

:30o
(6r32 14 -4 -t -40 -4 I
TEMPERATURE 32 20 10
TEMPERATURE, -F
Figure 1. Crushing strength vs temperature for corn- Figure 2. Relationship between meteor-
mercial artificial ice. Load applied normal to long axis ological conditions and type of icing (Kur-
of prismatic crystals (Butkovich 1954). oiwa 1965).

2. Conditions governing type of accreted ice of high wind speeds and air temperatures and large
droplet sizes, glaze ice will predominate because flow
a. Meteorological of the impacting drops can occur prior to freezing, and
Since ice forms by the accretion of cold or super- most or all air can be excluded. At the other extreme,
cooled water droplets which must dissipate their heat when wind speeds and air temperatures are low and
of fusion, the type of ice that results will be governed droplets small, soft rime will predominate, because of
by the temperature of the droplets and their size, as the rapid freezing of the more supercooled droplets
well as the rate at which they st: ike the surface. The upon impact before any flow can take place. Thus,
factors governing this rate are the wind speed and the considerable air can be entrained in the ice, resulting in
number of droplets in a unit volume, i.e. the liquid its characteristically low bulk density. In between these
water content (LWC). The air temperature is also im- extremes hard rime will form. It will still exhibit a
portant in two respects: in its influence on the rate of grainy structure, but some flow will have taken place,
heat dissipation by convection to the surrounding air, excluding some air.
and its possible influence on the temperature of the The liquid water content varies with the type of
accreting surface. That air temperature in itself is not cloud; in fact, this is one of the principal determinants
a strong indicator of an icing condition is suggested by of cloud type. Ice accretion should therefore bear some
observations of glaze ice formation at surface air tem- relationship to the cloud type present during ice torma-
peratures as low as -20oC (-4 0 F) and as high as 5 0 C tion. Glukhov (1972) has shown this relationship from
(41- F), though this type of ice occurs most frequently data obtained from a 300-m-high instrumented tower
bctween -3 and 0°C (27 and 32 0 F). Kuroiwa (1965), in the Soviet Union (Tables 3 and 4).
however, has shown that the type of ice can be corre-
lated with three wind velocity-air temperature regimes h. Structural
ard three air temperature-droplet diameter regimes, in The type of ice formed is not influenced strongly by
6bhscrvations made on Mt. Fuji (alt. 3776 m) and Mt. the structure itself, though the amount and shape of the
Neseko (alt. 1300 m) in Japan. I he graph in I igure 2 ice accretion will depend upon the structure's geometry
illustrates the general principle that Under conditions and exposure. A rigid, fixed c, linder will retain a fixed

3
-T

Table 3. Relative frequency of types of ice formed according to

cloud type (%) (Glukhov 1972).

Type of ice accretion

Hard Soft No. of
Cloud type Glaze rime rime Mixture observctions

Stratus (StO 18 48 4 30 227

Stratocumulus (Sc) 6 66 - 28 75
Nimbostratus (Ns) 28 36 - 36 89
Fractonimbus (Frnb) 26 33 6 35 65

Table 4. Relative frequency of ice accretion according to cloud type and air temper-
ature (%) (Glukhov 1972).

AIr temperature (c C)

0.0 -2.1 -4.1 -6.1 -8.1 -10.1 -12.1 No. of

Cloud Type -2.0 -4.0 -6.0 -8.0 -10.0 -12.0 -14.0 observations

Stratus (St) 30 16 16 11 15 10 2 227

Stratocumulus (Sc) 14 8 31 24 13 5 5 78
Nimbostratus (Ns) 52 28 10 5 3 1 1 92
Fractonimbus (Frnb) 47 19 19 9 6 - - 65

orientation into the wind, and droplets striking the where M = mass of droplets causing icing in time t, g
windward side cannot cause accretion on the le ward R = cylinder radius, m
side unless water flows there. Thus rime will ' ,rm only E = collection efficiency, 0 < E< I
into the wind-the lee side will remain ice-free Glaze V= wind speed, m/s 3
ice, however, may form both on the windward and lee- w = cloud liquid water content, g/m
ward sides when flow takes place around the cylinder.
In strong winds, a "tail" or plume of ice may be driven Not all droplets in the volume of air equivalent to the
by viscous flow around the cylinder. If the accreting projected area of the accreting surface will strike the
cylinder is free to rotate, as is the case with guy wires, surface, and of those doing so some may not freeze but
some antennas, and transmission lines, gravitational will be blown off. The collection efficiency accounts
forces acting on the accretion on the windward side for this; if all droplets in the airstream strike and freeze
may cause the cable to rotate, exposing a fresh surface on the surface, E = 1, whereas any loss is reflected in
for accretion. Thus, a cylindrical deposit will build up, E < 1. E is defined as the ratio of the mass of water
and the type of ice resulting may be shifted if the con- droplets striking the surface in unit time to the mass
ditions are on the borderline between two of the three of droplets which would have struck the surface if they
growth regimes. had not been deflected (see Fig. 3). According to
Langmuir and Blodgett (1946), E is a function of two
3. Accretion rates parameters, K and 4):

a. Fundamentals K 1.29 x 103 (Vr2IR)

The amount of ice deposited on a unit length of and (D 0.175 VR (2)
cylinder in unit time is given by
(for sea level and 00 C conditions), where r = droplet
dM/dt 2REVw (1) radius (cm), V =wind speed (cm s), and R c\linder
radius (cm).

4
 TANGEN TRAJECTORY
Figure 3. Collection efficiency of a cylinder.
Water droplets folio wing trajectories off the
center line approach will be deflected and, at
S-
I -....... .
j
a certain distance from the axis, will not im-
pinge on the cylinder. Collection efficiency

is the ratio of the mass oi droplets striking the

cylinder to the total mass that would impinge
if they were not deflected. If a constant mass
flux is assumed, this can be represented in terms
nt the linear measures S and /), so that L S,)D
-STREAMLINES (Stallabrass and Hearty 1967).

/0
08 '0 /
0-/ 4 /0o

/ 2x/, '
5i0

20.6- /

o0.4
0

02-

0.1 3 4 6 0 [ 2 3 4 6 810 2 3 68 100 2 3 4 6 8100C

K

f iqure 4. Odlection efficiency tL is determined by an inertia parameter K and the parameter i

A graphical solution for E, once K and 4D are calcu- vary depending upon turbulence. The relationship has
lated, is given in Figure 4. Note that K varies inversely been developed for small (atmospheric) water droplets.
with cylinder radius, in agreement with observations Soviet observations of icing of a 1 5-mm-diameter
that the greater the cross-sectional area of the accreting circular cylinder have resulted in the data plotted in
cvlinder the lesser the accumulation. This holds true Figure 6, where the collection efficiency E is given as
for shapes other than cylinders, but in those cases a function of droplet diameter and wind speed. Ilhis
other factors, such as wind direction with respect to show,, the si/es of droplets that will impact on a c liin-
the surface, become important. If K< /8, L = 0 for der at a gien wind speed.
an ,,. Maripulation of these equations can give the Not all the droplets striking the obstacle, whethc
critical radius of a cylinder above which icing theor- supeicooled oI not, will frecie and he retained on the
etically will not take place. A graphical solution is obstacle. Stallablass and I leart\ (I 907) deline 'icing
ivcn in f igure 5. The actual si/e of the c',linder will eflicienc'" a> the product of the collection cnliienc\
 100 r-T-r-T -rTT T r rTT"r'm-

E
Z. 50
40

0 20
I 401 2 045

E
o ' 4 5 10 20 30 4o loo
I
V, Wind Velocity ({m/$)

Figure 5. Critical radius of cylinder above which icing theor-

etically does not occur.

U. ' u, wind speed (m/s)

0
12S
2 0

S28

N 4

0 _ 2

U I 16 2u 24 7-

r, cloud droplet radius (iml

Figure 6. Variation of collection efficiency E of a 15-mm-radius circular

cylinder with cloud droplet radius r and wind speed u (Glukhov 19 72).

and the freezing fraction. They conducted ice accumu- mum and then drops off gradually. The trend, plotted
lation tests in an icing wind tunnel and measured a one- from Soviet observations and sh, ,wn in Figure 8, re-
hour average icing efficiency. The results, some of suits not from a reduction in the collection efficiency
which are shown in Figure 7, graphically depict the but from change in the freezing fraction. (In fact, as
great influences of cylinder diameter and temperature is evident from Figure 4, collection efficiency E ap-
upon ice accretion. (Test conditions were: air speed proaches 1 as K increases, and K is a linearly increas-
50 mph, air temperature between -160 and -50C, liq- ing function of wind speed.) At low wind speeds
uid water concentration 3.2 g/m 3 , median droplet di- nearly all the droplets which collide with the surface
ameter 200 tim.) will freeze because the heat of fusion can be dissipated
Wind speed influences not only the collection effi- to the surroundings. As more and more droplets strike
ciency but also the freezing fraction. With increasing the surface with increasing wind speed, however, the
wind speed the icing efficiency (or capture coefficient heat of fusion cannot he dissipated rapidly enough.
as it is called in Soviet literature) increases to a maxi- Some droplets will remain in the liquid state and

L ~.......
 1 IS . . 15

iS' OIAI
ISO
I60_,__ 0 0 i"
0 a S

V Y 12"

120 c --

3" DIA

o - D
12" DIA

10"DIA ,. .

-24 -20 -'6 -12 -8 -4 0

AIR TE61KRATUREC.

Figure 7. Icing efficiency, i.e. the fraction of water droplets

that could strike a cylinder if not deflected and which will
freeze on the surface, as related to air temperature and diam-
eter of cylinder. Small cylinders, below 3 in, in these ex-
perimental results, will form ice with thickness exceeding
the original diameter of the cylinder when temperatures
are low enough (Stallabrass and Hearty 1967).

t5

,O) * 'o?
01 0* 0

0 00
o. _ .

0 0. .'.o . - .-. 00
00

CL i - I I. , ,
2 4 8 tO 1 f4 u m/sec
Wind speed (m/s)
Figure 8. Dependence of icing efficiency (or capture coefficient) on
wind speed for various liquid water contents w (Glukhov 1971 ), where
3 =
1) w = 0.12-0, 16 g/m , 2) w = 0. 17-0.21 g/m', 3) w 0.22-0.26
g/m 3 .
7
 80 800

70 -700

80 -600- --

50 o500 /
S40- 400
/EI, I,..i
30 - . 300

20 200 - /'

10
I0 0 I I

0o 200 300 100 200 300 1"

Height (m) Height (m)
a. Average diameter vs height. b. Average weight vs height.
Figure 9. Variation of average diameter and weight of ice accumulation
(1-mixture, 2-hard rime, 3 -glaze ice, 4-soft rime) with height on
meteorological tower at Obninsk, U.S.S.R. (Glukhov 1972).

flow to the periphery, where they are entrained by the actual occurrence of glaze ice formation will not al-
air flow and carried away. ways follow these conditions. This analysis is of
little value in estimating potential ice accretion rates
b. Effect of height by region and season. The Climatic Atlas of the
Both the liquid water content of the air and the wind Outer Continental Shelf Waters and Coastal Regions
speed directly affect the accretion rate (see eq I). Both of Alaska (1977) contains monthly tables of precipita-
of these factors also vary with height; LWC varies with tion associated with wind, precipitation types, air tem-
cloud type which varies with altitude, and wind speed perature (dry and wet bulb), fog, cloud cover, wind
in the boundary layer generally follows a logarithmic speed and direction, and several associated occurrences
profile, so accretion rates will vary with height. Obser- of these elements. Tattelman and Gringorten (1973)
vations made from a 300-m-high tower in the Soviet have estimated probabilities of glaze ice storm occur-
Union show the variation of accretion with height for rence according to climatic region, based on data from
the three types of ice and a mixture (Fig. 9). Frequen. Bennett (1959). The climatic regions are shown in
cy of occurrence of the ice types at this same location Figure 11, and probabilities in Tables 6-8. Region VIII
is given in Table 5. incurred no glaze ice storms in the analysis period, but
experienced the only two rime ice storms in the country.
L. Geographical distribution The National Weather Service does not measure ice
Bennett (1959) made a comprehensive search of accretion, and a climatology of ice accretion is not yet
available data on glaze ice occurrence. Principal well established, so estimates of icing potential are poor,
sources were railroads, electric power generating and and any predictive capability for estimating icing se-
transmission companies, and telephone companies, verity according to season or geography is meager.
Most reliable data, therefore, are confined to the con-
tinental United States, the populated parts of south- 4. Spray icing*
ern Canada, Europe and Scandinavia. Bennett pre- Statistical analysis of more than 3000 cases of ship
sents a map made by the U.S. Army Air Force in icing (Borisenkov and Panov 1972) indicates that the
1943 which includes Alaska; it shows the mean annual principal cause of icing is spray from ocean water (89%).
percentage of hourly weather observations with freez- The combined sources of spray and fog, rain, or drizzle
ing rain or wet snow (Fig. 10). Wet snow was consid- accounted for only 6.4% of the cases, and water spray
ered as snowfall occurring at temperatures above 320 F
and freezing rain was assumed to occur whenever li- This subject is covered in more detail in CRREL Report
quid precipitation fell at temperatures < 32 0 F. The 77--17 (Minsk 1977).

8
Fgr 1gra

o S o . S. Aoe

on amappreare
in794 by he eater nfomaton Srvie o th U.. Amy Ar Frce (Bnnet 759)

Fgr "ue

~~~~~Figure
0 11.
enana ecae Reg siia
ywete lzfc
osrionsh
with racerisics aimnorhAeiaBsd
onap
a pre are
d 9 7e3).fnomto
Ging3byten 7ete h .. AryArFocs(enet15
90

,;
j__ RE ION9

! O N
R G 77:1111Z
 4

Table 5. Occurrence (%) of type of ice by height on meteor.

ological tower in Obninsk, U.S.S.R. (Glukhov 1972).

t ypt. (it hce

ht rnc
H'... fm) Soft rime (hi,e FIord rime Mi tire

0.100 23 5 S 2
100-200 29 31 25 23
200-300 48 (4 70 75
No. ot Lass 227 237 633 396

Table 6. Probability of at least one occurrence of an ice storm in any year at a repre.entative point in
the region (estimated from Bennett 1959, in Tattelman and Gringorten 1973).

Regional Regional I Probability

Regional Probability Average Probability Average of an Ice
Average of an Ice No. Storms of an Ice No. Storms Storm
No. Storms Storm (Any _>0. 63 cm Storm > 1. 25 cm > 1.25 cm
in 1-Year Thickness in 9-Year > 0.63 cm in Q-Year in 1
Reeion Study in 1 Year Study in 1 Year Study Year

1 8 59 5 .43 3 .28
[1 18 .86 7 .54 3 .28
Ill 20 .89 10 .67 6 .49
IV 15 .81 5 .43 3 .28
V 8 .59 4 .36 2 .20
6 .49 4 .36 2 .20
VII 4 .36 2 .20 0.5 .05

Table 7. Probability of at least one occurrence of an ice storm of stated intensity in any year at a
point in the most severe part of the region (estimated from Bennett 1959, in Tattelman and Gringor.
ten 1973).
Reizional Re'cional
Ieid, onal Maximum Maximum
Maximum Probability No. of Probabilitv No. of Probabilitv
No. of of an Iee Storms of an Ice Storms (if an Ice
Storms in Storm (Any >0. 63 cm Storm 1.25 cm Storm
9 -Year Thickness) in 9-Year - 0. 63 cm in 9-Year > 1.25 cm
Recion Study in 1 Year Study in 1 Year Study in I Year
24 .93 14 .79 8 .59

11 35 .!8 10 .67 4 .36

111 44 . w) 19 .88 9 . 63
IV 35 .918 11 .71 8 .59
V 20 .89 9 .63 6 .49
VI 16 .83 8 .59 5 .43
Vtt It .71 4 .36 4 .36

10

=NO
 Table 8. Number of ice stornib > 2.5 cm and > 5 cm in 50 years and the probability of at least one
occurrence in one year in the most severe state in each region (Tattelman and Gringorten 1973).

Kumlbe-r ihe Sturyirs

2. 5 cn iiIn A0
1'r-ohahilit ' of
an Ice Storm
Numnber Ice Storms Probability of
> 5 cm in 50 an Ice Storm
itegitu I Year 1 Year

4 .08 2 .04
1I5 .04
II I.16 3 .06
lv4 01 .04
V4 - 1 .02
OH.0 1 .02
.06 1.02

Table 9. Distribution of ice incidence H ) for Soviet ships in various seas as related to air and water temperatures
(Borisenkov and Pchelkso 1972).

It perattaI-
ul, it

5.0 n. 1' 1.H.11 1), 0 0it 1tt 12oI/~

1) 1.ia b. ) 0 6. U ds.

"C4
t:li 't 11 140 20 2t. so 4 297

l~as. ti.. iI2' 2' IS I- 4 4 192

.1W1. H I 2' lo H
191 239

a ~~r .. .tt~ 2 1 4S 49 1 S73

9
tai,. r , IItI ht 36
It,10 sr t 4~~g 44 8 82

I 22 21 4 2521

.11. iI ii t' t ia~ , .0it'lsrIah , 1, :'1', it's, 'c.t , l 0 r (1

20 it) 5u0 ')1 the I,00

ni rr t, tg
i .1 /' .,.'titlrr2 " tk IDedk Id irrlti.ilio i r -rii spfi% tn1U snmsttli leached

ltt ' *hliti3! t1)itf.1 i iitit1I1 1 r,1g1101. 'ttt Ci c\perimenis and obswii dtionsovter three
as., ii ~ tr.ini w.r
wui~~ klitpt! iow, A.~ Il 1jL- wisitC! *m i in tht-Sca ot ljapadn, Barents Sealand
.fill] dirt 11-fl
titti %dir ir'gr WiIt 1114, tic t1d
tiL-A111 ianges it cig dctis i (Borisen-
1 .11rtalv i .tl iiIlt-ljtair iii kt ind P.telktr 19)'21. Air ttrnperaturescidown to -3C
anid tws ht-cri 'CL'rttt it tiniperaturt it, 1.% 2~ ii n~it
tt (tnpair d b% signitsi.int it inA. At relJtieJ
sht'rkhirriar I I 't681~his, tclttd it ii. r it , t rusS54111 tcedt
1,k i( rn s)i and airtenhpenatures beloS
WiritA p,-cdh nrra ~i ! Ij.
l'bl "ra i ( , ileic [),\ it iii. %hip htL ne icaed I he rest ot
Rt'A- rh irn Iipar I iiij i i it I
i'#(I j-
.1'" r ,itt t tht stqi .at rcic% mtl\ %fight aiiuintN tit ice, tor does
ic iiv "eci ii,. t ittiai teritio l wi .i % rIAtl sItri
t I no
rig tt uurrii t e ani\ at .all, Stroing \-wint ( I 2.5 rn, %)
I 2) 1 hi ct , I-su, t~i',ml, i silr, I ot
ald t d M iil 1CnlPpe lJrVtS Iit -1 S ( t-cslt In iC .ALLreticr on
t
hiass o ,cpt'Id I"t. it, t).5i ol isI i !Istl the mialin ickk, i tggirng, msit and Tpars, how comnpanioln-
LJattIh Vi I, 1 . t t . i i,%'
5-l wrd hl dgc i U s irriir ,ttpttitikin itLors con the
aI't
'71-111f''r' *tIT~ Iriaiurgl A1101 I he Lipper hridge and
., g ~ ~ ,* . ~ ., ,,Iit~ , ~ 01 h.o l JCt I, iLL I,1,11Cai %cr' little ice, I he %hip's
 I4

Table 10. Distribution of icing incidence (%)for Soviet ships in various seas as related to wind direction and
wave height (Borisenkov and Pchelko 1972).

Wind diret tion and iped 1Uve vi t

(rn/i) (m) P, ,od

271 -60' 00-90 91./0' 181.27U o/ ihip% "

SU <10>10 -10,10 <10--10 < I0>10 ause- 1-3 -3 casis iciq

Bering Sea 12 43 16 21 I 1 2 2 567 5 45 481 O.L-Mar

Sea ol Okhotik 19 34 1 I0 I 3 10 20 323 73 27 25: Oct-Mar
Sea ol Japan and
latarskis Stlait 2; 5S 5 9 I 7 199 82 18 I I DCL-I b
Western Pa~ilik ()can Ii i2 I 8 I 2 6 15 232 71 29 Is No 4 -cb
Balents Sea,
Norwegian Sea 2 9 7 141 19 20 9 20 638 78 22 525 I)et-Mar
BaltiL Sea 24 12 22 22 45 79 21 45 jan-Naiu
labrador fegiun 21 48 2 2 I 2 12 12 87 67 13 .13 DL-Feb
lauLkSea and
Sea ol A/o ii IS 11 jan-Mar

Table 11. Frequency (%) of ice intensity related to wind speed

(Shekhtman 1968).

h mq Hind speed (knl ots) spr,"d ol

I'll, ~lt 0.1 Iot 1-20 21-29 30 and otier (knots) u,,,

I J.l gr utlh 2 I 12 42 40 29 i
slow "ti w It 1 8 29 43 19 23 304
No ,hangc 2 22 39 24 I3 17 54

.30- 4

SAA 0---6
+"6
01

Ca 0o
SA4 1
4-.1
M
--
0 Figure 12. Icing severity as related to air temperature and
0-eo-wind veoiy(oiekov and P'anov 19 72, taken from
• ~~ ~ ~0 ~ .. ~a ......
------ ~~ JIapanese data in Tabata et al. 1963) where a = 450.ton
0
" A . '6 " displacement ship; b = 350-ton displacement ship, 0 no
1 s - 1 " ,icing; I significant icing 2 heavy icing.

A. 0

Air temperature (C)

12
 stern does not become iced even when the heading is runoff water freezing slowly because of high rainfall
downwind. rates, and by direct contact of supercooled water drop-
The distribution of ice over a ship tends to be erratic. lets on the sloping or vertical surface.
Investigations on the Professor Somov in January-Feb- A relationship between the precipitation measured
ruary 1968 established the following pattern of ice dis- by a precipitation gage and ice that will accrete on a
tribution: 30-70% on the horizontal surfaces, 15-40% vertical surface (assuming necessary negative tempera-
on the vertical surfaces, 5-30% on the surfaces of com- tures prevail) has been proposed by McKay and Thomp-
plex configuration (instruments and equipment), and son (1969):
0-30% on the round surfaces such as mast, spars, and
rigging. In the presence of spray, ice accretion com- T = 0.785 VP ° '- (3)
mences immediately at temperatures below -3 0 C on
the metal surfaces of the ship and on canvas equipment where T = rate of precipitation striking a vertical sur-
covers. However, ice does not form immediately or. the face, in./hr
wood decks; instead, a slush develops which mixes with V= wind speed, mph
seawater and flows overboard through the freeing ports. P = precipitation rate, in./hr
For a period of 1V2to 2 hours after formation on metal
and canvas surfaces the ice is loosely bonded and can This relationship assumes that the collection efficiency
be easily knocked or scraped off. After that time the E is unity.
ice becomes tightly bonded to the surfaces and can be Loads due to ice accretion are generally less than
removed only with great difficulty. design snow loads, but the prevalence of structural
Shellard (1974) summarizes conditions for icing due collapse of antenna towers, powerline cables and support
to seawater. These occur whenever sea spray is present structures under conditions of no wind suggests that
at the same time that the air temperature, and therefore ice loads can be high. For example, a glaze storm on
the temperature of most exposed surfaces, is below the 27-29 January 1940 struck Great Britain and resulted
freezing point of seawater. The freezing pcint will vary in some of the greatest accumulations of ice ever re-
from a little below 0°C for only slightly saline waters to corded: 6 in. of ice on an automobile and 4 in. on
-1.9°C for ocean water. A small vessel is likely to begin twigs in Worcestershire, and a 2.4-in.-diameter accum-
generating spray in a sea corresponding to force 5 ulation on a telegraph wire in Wiltshire. The maximum
(17-21 knots); at force 6 (22-27 knots, with wave thickness reported in a series of observations in the
heights of 3 m or more) most small vessels moving Soviet Union in the 1920's was a diameter of 114 mm
against the waves will be showered in spray. Spray on a 5-mm wire in [okmak in the Kirghiz (Bennett 1959).
blown from wave tops is not likely to become a serious Ice that accumulated to a diameter of 25 cm on guy wires
source of icing, however, until much higher wind speeds of a tower in Newfoundland was estimated to have
are reached, because such spray is patchy and at a low weighed over 40 kg/m (Boyd and Williams 1968). Tables
level when it begins to appear at 17-27 knots. As a 12 and 13 present estimates of ice accumulation on a
consequence, large amounts will not likely reach deck cylinder for the regions defined in F igure I I (Tattleman
level until at least force 9 (41-47 knots), the point at and Gringorten 1973). These values are based on 61
which visibility begins to be affected. These observations storms and required subjective development and mani-
can be applied to fixed installations which may not pulation of the statistics. Caution is required in applx-
generate quantitits of the larger droplets from impact- ing these estimates in design.
ing waves.
Woodcock (1953) reports that foam patches result- b. Wind fieldin the boundary layer
ing from whitecaps at the sea surface remained visible Wind speed in the lower atmosphere varies with
from an aircraft for more than two minutes in a force height as a consequence of the frictional drag developed
5 wind in the Hawaiian area, which supports Shellard's by the rigid surlace on the flow. It has been found that
comment regarding the initiation of spray icing at that a logarithmic decay expresses this change quite closelv
wind speed.
u,= (u./k) In (//z) (4)
5. Structural design factors
whereu, - wind s1-eed at height z
a. Dead loads u. friction velocity
Freezing rain falling under calm or low wind speed k = von Karman's constant (z0 40)
conditions will coat horizontal surfaces with ice; the z. roughness length (heigh, at whith the
windward side of a structure will also accrete ice on neutral wind profile extrapolates to zero
surfaces oriented other than horizontally, both from wind ,peed)

13
 Table 12. Ice thickness, estimated to the nearest 0.1 cm, for different return periods, at a representative
point (AVG), and at a point inthe most severe location (MAX) for each region (Tattelman and Gringor-
ten 1973).

7 Return Period (Years)

2 5 10 25 50 100
Ie,,.ron AVG MAX AV(; MAX AV( MAX AVG MAX AVG MAX AVG MAX

1 0.4 1.4 1.4 2.1 1.6 2.4 1.8 5.0 1.9 7.1 2.1 >7.5
I1 0.7 1.0 1.4 2. 1 1.7 3.3 2.0 5.0 2.2 6.0 2.4 7.0
11 1.2 1.6 1.6 2.4 1.8 3.8 2.0 5.8 2.1 7.2 2.3 >7.5
1\ 0.5 1.4 1.4 2.1 1.6 2.4 1.8 5.0 1.9 7.2 2.1 >7.5
% 0.2 1.2 1.2 1.8 1.5 2. 2 1.7 2.5 1.8 5.0 2.0 7.0
% 0 1.0 1.2 1.9 1.5 2.4 1.7 3.8 1.9 5.0 2. 1 6.0
1 0.4 0.6 1.8 1.0 2.2 1.3 3.4 1.5 5.0 1.7 6.3

Table 13. Ice thickness, estimated to the nearest 0.1 cm, combined with wind
gusts > 20 mjs in the most severe location in each region (Tattelman and Grin-
gorten 1973).

Return Period (Years)

Region 25 50 100

2.5 5.0 6.5

I3.7 5.0 6.5
Ill 2.6 5.0 6.6
IV <2. 5 3. 5 5. 6
\ <2.5 e2.5 3.6
VI <2.5 3.0 4.5
VII <2.5 2. 5 4.1

U. can be calculated using the relationship that the 1 s measured wind speed at reference height zr
slope )fthe In / vsu plot sk It intcreSeS with = roughness length.
,ughness rf surfLse and with mean wind speed, but
it is tppr .indatel[ equal tol '10. . I'lnd loud%
I he Aner uan
, Nat i(nal Standard BLuilding code
S ome value%(of/0 In feet are (ANSI 1972) states that wind hr)ads (in buildings and
oither stiuttuies shall he determined by seleuting a mean
-(5 -4
wa ter 3 x 1 t) 1 s 10 reuurrette intersal based on usage, anticipated structure
ue x 11 .4 lite s,ensitisit, tor wind. and life or propertN risk in case
to
,,(v, II t() , 1 si0 falure. A hrsi w rid speed is selected, based on

.irnutal -trLt'nlc
i v tte -nlilk.
,,pett 3:0 Itabo,,tCg! tlul'(i
v.isurenirrit if w rid speed it one height enables tlt
Ut,iri )of tie spetd at art)tht height' frr a 0- ( 00 f Cer re surrre iriersal
tlosing a published small suale nap itgood lr.al dati
I
i,/ l/ ,ire not available). I hen the seles ted hasiu wsind speed
I is ci verted to effeu i vel it,pressUreS j lror build.
rrgs .td sIf tines ,l c/i)for palts and portions of
Whert widl atsesir
dr, ures, ,a i US igrts, el irIed .s
whe.rte d t nrr peed ,tihtight /
sl(i

14
 6

4- .2 2

-- \
0

0
12 '3 4

Diameter (inches)

Figure 13. Correction factor K, for use in estimating ra-

dial ice thickness (Chaine' and Castonguay 1974).

The wind pressure coefficients for flat and cylindri-

qF = Kz GF q3 0 cal surfaces were calculated from

and flat surfaces: P, = 0.0042 Gm 2 lb/ft2 (9)

qp = K z Gp q 30 cylindrical surfaces: P. = 0.0026 Vm 2 ib/ft 2 (10)

where Kz= velocity pressure coefficient dependent on where Gm and Vm are given above.
type of exposure and height z above ground
GF, Gp= gust factors dependent on type of exposure Equivalent radial thickness Ar is defined by Chaine'
and dynamic structural response as the probable ice that would accrete uniformly around
q 30 = basic wind pressure (lbf/ft2) = 0.00256 V02 a 1 -in.-diameter cylinder equivalent to the asymmetric
where V30 = basic wind speed in mph. natural formation:

TablesofqF andqp are included in ANSI A58.1 (ANSI Ar= Ah2 +Av2 )+r2J (11)
1972). A procedure is also given for calculation of gust
factors. The National Building Code of Canada (National where
Research Council 1977) also includes sections on wind
effect, wind levels, and icing factors. K = correction factor based on cylinder size
Chaine' and Skeates (1974) have prepared tables
showing ice accumulation on a horizontal surface and actual accumulation
the related vertical surface accumulation, based on ob- theoretical accumulation
served maximum wind speed and computed gust speeds,
and the equivalent radial accretion, transverse wind load- A h = accumulation on horizontal surface
ing and wind pressure on a cable and tower. Nearly 150 A v = accumulation on vertical surface
locations in Canada are tabulated. r= cylinder radius

The maximum computed gust is calculated from The correction factor K is the ratio of experimentally
measured ice thickness to theoretical ice thickness.
Gm 5.8 + 1.29 Vm (8) Chaine has prepared a graph of this factor as a function
of accreting cylinder diameter and air temperatures
where Vm = maximum wind speed, mph. (Fig. 13). The experimental measurements were those

15

C .. , . . . . . .= -
made by Stallabrass and Hearty (1967) for a very limited alone is usually insufficient to reimoe well-bonded ice,
range of conditions (they measured the accumulation A poly(dimethyisiloxanc )-biphcn ol-A-puld carbonate
after a one-hour exposure to wind of 50 mph and liquid block copolymer developed for CRREL (Jellinek 1978)
water content of 3.2 g/m' and a median volume diameter has been used successfully to reduce tire <rdhesion ot
of droplets of about 200 pin). This equation is therefore ice to nasigatior lock walls (I ankcnse incl .I ,i 1976)
only a guide and not to be accepted as a precise determi- The same nraterial has been applied to a dish radar
nation of expected ice thickness. antenna to reduce heat input necesar', to control
iceaccumulation (Hanamoto 1980); tests showed that
An approximate relationship between accumulation on the 145 minutes required to eliminate 0.6.4 cm (0.25
vertical and horizontal surfaces was developed by Chaine' in.) of ice on the antenna surface, heated at 2.3 Wim2
and Castonguay (1974): at an air temperature of -2.5 C could be reduced to
20-44 minutes. Vibration of the coated, unheated
A, :A h V/I10 (12) surface did not eliminate the ice accurnulation. An
ogani, free/ing point depressat, Mtonsantt, Santourelt
Transverse wind loading Lc as used by Chaine' gives the 990-CR was effectli in keeping the fhi4ht deck, turn-
maximum load on a cable (powerline) occurring during buckles, shackle, and lines :caritice (inthe USCGC
an ice storm as a combination of ice accumulation and Burton Island (Bates 1073). [he use , and sCaLh for
wind; it is calculated by additional effreeise icephohiL sur t4rrcaments ire
diLLIsed h% Sa(19std 1979).

L = 0b0026 V 2 llb/ft (13) Stallabrass (1970) investigated a numbei of methods

12 of reducing accretion of ice on ships. I he methods
utilized, in order of their deicing effectiseness or ease
where Or ice removal, cec:

d z conductor diameter, in. I. Pneumatic deicer

Ar equivalent radial thickness, in. 2. [reeling point depressant (ethvlene glycol)
:. Rubber-surfaced plastic horant on steel panel
6. Techniques for minimizing structural icing 4. Gras deck paint on wooden panel
No completely effective methods of ice accretion S. Spar varnish on wooden panvl
control or removal have been found. Mechanicr (impac - 6. bar,,polhethylene foain on stecl panie I
tion) methods are the most common techniques, but 7. Black rust-preventive paint on steel panel
experiments have been conducted utilizing heated surfa ce,, 8. [,rra deck paint on steel panel
icephobic surfaces, deformable surfaces, and freezing point
depressants. Also, no device for unequivocal remote I hese qualitative tests were made in an iing wind
measurement of ice accumulation rate is available, tunnel and on outdoor (land) tet Sites. In addition, .r
though some techniques are available for such measure- polyethylene-sheathed parallel filament rope was com-
ments under experimental conditions. In critical areas, pared with a steel cable for ice accirtion and removal.
where no ice accumulation can be accepted, or where Because of the lower torsional stillncss (f the rope, it
accumulation beyond a certain point may lead to cat- tended to twist during the iing esent and ice accreted
astrophic failure, active control methods must be used. around the entire cirt uriel enIc, in (011nrat TOthe steel
Application of heat is the most dependable but not cable which prese'nted . LVirStat ft e to, the airborne
necessarily the most convenient, practicable, or cost- droplets and Ihiirie ac unrulited ice assn metr ically.
effective method, because irregular surfaces, surfaces Little diflference was noted in the ease ot remusal Of ice
exposed to very high heat losses, and extensive areas all from pol, elh\ lene-slicalied rope compared to the steel
present problems. cable 'Vcae of the encapstlallon of the plastic.
Methods of removal of ice accumulations reported in I lie 4 x t pneunatit dICILr svurkcd effecti\e l,
the literature are conventional: baseball bats, sledge wIen atta hetf 10 both I I-ft-diameter c\ lini ical form
hammers, axes, hammers, picks, and other impact instru- srinilatin! a mast andi when s'rcad ralt on itpanel.
ments are frequently mentioned as the only tool available, Bet,use stimail di,metOI L\ljnders aSc uuLal I Mire
but icephobic coatings and heated surfaces have been ex- it C than large dirrMeter k finde s, Icrtr 11- n carn redLcCd
perimented with. By itself, an icephobic or low-energ\' hs gathering tirgether iald enclosing iriunher ioi m,ill
surface may be insufficient to reduce ice accumulation, cables in ishcath, Or installing ,rlarge diameter sleath
but in association with some other mechanism such as arfnd a singlc sirall cable. A British-designed ice-
vibration or a deformable surface ice removal may be resistant trr It-i uses, lai g cdiarreter ugigus, ed tripod masts.
accomplished effectively, Vibration or deformation rather thaIrn a ast guiecd h\ ItulliplC
Migi, r iggiri cables,

16
 as a means of reducing sailarea and the amount of super- sensing apparatus and tested the Rosemount ice detector
structure icing potential, as a candidate. This instrument, an automatic contin-
An item in a popular Soviet publication (Soviet Life uous-recording device (Rosemount Inc., Minneapolis),
1975) reports that their scientists have suggested the was developed initially for measurements on aircraft,
use of an anti-icing "shirt" for protection against ship but has been adapted for detection and measurement
icing, and that tests have proved the effectiveness of the of land structure icing. Detection is based on the
method. It is quite likely that this is merely a flexible change in resonant frequency of a small vibrating rod
plastic sheeting covering portions of the superstructure. in the droplet stream. Sensitivity can be selected to
Ice and snow accumulation on power line towers in initiate a deicing cycle when ice growth has reached a
Norway has been reduced by enclosing monolithic predetermined mass; the number of deicing cycles is
concrete structures as well as open-lattice steel towers a measure of accumulation rate.
with sheets of solid, corrugated plastic. This technique The most comprehensive data on altitudinal varia-
may be applicable to ocean installations. tion of icing have resulted from the two Soviet instru-
Since sea spray is the principal source of icing in mented towers. A Soviet network of cable icing measure-
maritime regions, reduction of the supply of spray from ments has also been established to measure the accumu-
breaking waves is an obvious control strategy Tor fixed lation on a standard diameter cable (5 mm), though
installations. Wave damping would be required for the other diameters are also used in order to determine the
maximum distance that a droplet would travel. Large collection efficiency variation with cable size and with
(I to 3.5 mm) droplets generated by wave impaction, meteorological conditions. A similar icing accretion re-
rather than spray whipped off breaking swell waves, porting network inAlaska and other northern states
are of primary concern. Droplet size of the latter averages would provide much useful data. The construction of
20 0/mm (Wu 1973). Borisenkov and Panov (1972) state instrumented towers with provisions to measure ice
that the flight time of droplets generated by a ship plow- accretion variation with elevation, and associated means
ing into waves is 1.3-1.4 s until impact on the ship. Since to measure cloud liquid water content and droplet size
this presumably is the forward part of the ship, total at a limited number of locations, would provide the
flight time until the drop reaches the sea again is reason- basis for establishing a predictive capability.
ably estimated as 6 s. In a 30-m/s wind, a drop would
travel 180 m. A small droplet, 200gm and below, would 8. Literature cited
travel a much longer distance. It is not likely that it American National Standards institute (1972) American Na-
would be economical to install wave damping for such a tional Standard building code requirements forminimum
large area. design loads in buildings and other structures. American
Deflection of droplets from the trajectory that will National Standards institute, New York, N.Y., ANSI
A58.1-1 972.
carry them to a surface to be protected may be possible. Anon. (1962) Precipitation measurements at sea. World Me-
This may be accomplished by an air curtain, or by another teorological Organization, Geneva, Switzerland, Technical
surface whose shape can be given optimum aerodynamic Note No. 47.
design or which may be permitted to accumulate ice. Bates, C.C. (1973) Navigation of ice-covered waters: Some
This approach
specific isexperimental and requires deslign for
conditions. new initiatives by the United Slates of America. in 23rd
Report, International Conqress of ,\arilutior. Permanent
international Association o Navigation Congresses,
Brus.
Heated water is used for flushing decks and other sur- sels.
faces for ice removal. However, large quantities of sea- Bennett, I. (1959) Glaze, its meteorolog, and climatolog\,
water even slightly above the freezing point can be used geographical distribution, and economic effects. Quarter-
for removal or protection if cooling due to wind is not master Research and Engineering Center, Technical Re-
excessive and if proper drainage is provided to prevent port EP-105.
Borisenkov, E.P. and I.G. Pchelko ld.) 11972) Indicators
ponding. for forecasting ship icing (inRussian). Arkticheskii i
Antarkticheskii Nauchno-issledosatel'skii Institut, Lenin-
7. Data collection needs grad. Also CRREL Draft Translation 481, AD 03011 3.
Geographical, seasonal, and altitudinal distribution of Borisenkov, E.P. and V.V. Panos (1972) Basic results and
perspectives on the investigation of hvdrometeoro logical
icing ispoorly known and at present the National Weather conditions related to ship icing (inRussian). Arkticheskii
Service reporting network makes no observations or i Antarkticheskii Nauchno-issledavatel'skii Institut, Lenin-
measurements of ice accretion. The occurrence of freezing grad, Trud,, vol. 298, p. 5-33. Also CRREL Draft Trans-
or frozen precipitation is reported, but there is only a lation 411. AD A003215.
loosely defined relationship
loosly efiedbetween
etwen ice formation
elatonsip ie and
frmaionandLeningrad: ct al. (1961) Phv.sics ot Cloud (inRussian).
Borovikov, A.M.Gidromieteoi.,dat.
the amount of precipitable water, as was discussed in Bod, D.W. and G.P. Williams (1968) Atmospheric icing of
Section Sa. The NWS has considered including direct structures. Division of Building Research, National
measurement of ice accretion in an automatic weather Research Council, Ottawa, Canada, I echnical Paper 2"-'5.

17

S7
Bu tkovich, I .R. ( 1954) Ultimate strength of Ice. U.S. Arm N aifona I Researchi. ounsi I (977) Nat iori 3ui IWing Co~de oI
Snow, Ice and Permafrost Research Establishment (SIPM Canada 1977, NRC,( No. 55. ijniatkc Infioirmation for
Research Report 11. AD 0505 14. bu ilding Design in Canada 19,7. Suppleroii No. Ito the
Chain(!, P.M. and G. Castonguay (1974) Ness lPPrOaCh to National Building ( ode of Canada, N RC(: No. s~s.Comn
radial ice thickness concept applied to bundle-like conduc- nientary B, Wkind Loads. Commentaries on P irt . 1ofhil NA-
tors. Atmospheric Environment, Enivironmient Canada, tional Building Code ot Canada 19 77. Supplement No. 4to
I oronto. thle National B~uildinrg Code or Canada, N RCC 1;5565.\
Chains, P.M. and P. Skeates (1974) Ice accretion. handbook tional Research Council Of Canada, tittassa.
(treaing precipitation). Industrial Meteorology -Studs VI. Pili , R.J. attd W.C. Kintond (1967) Project tog drops. (or-
Atmospheric Environment, Environment Canada, Toronto. tell Aeronautical Laboirators , NASA CR-675. National
Frankenstein, G., 1. Wuebben, 1F.H.G. jellinek and R. Yokcota Aeronautics and Space Administration, Washington, [.(
(1976) Ice re-moval fraim the swalls of navigation locks. Sayrward, j.M. 11979) Seeking lows ice adhesion. CRRi I-
I'rorerdims ot is rpos/to ont /,ltid II uters for \oeiqu- Special Report 79-1 I. ADA 071040.
tioti, / Iood Control und Water fjit rd/on, Colorado State Seanty, IT.W., . L. Wise, W.A. Ffro%.. i, Jr., 11.1. DI)ai auG A .5
Unisersitv , August 10-.12. Prechtel ( 1977) Clinratic atlas or thle oauter oiritinsental[
Glukhov, V.G. ( 197 1) Evaluation of ice loads on high strc- shell isaters and coastal regions Of Alaska. Vol. I Gali:
tures front aerological observations (in Russian). Gidro- ot Alaska; Vol. 2, Bering, Sea; \ 01. 3. ChUkchi-Beaiil''
meteoizdat, Ti ude, vol. 283, G(;D, p. 3-11. (Translation: sea. Arctic Ens ironmental Informatlion and Dati Cenier
Soiiet l/sdro!ocit, Selected Puprs. Issue no. 3, 1971, Publication B-77, Univer sit (oif Alaska.
p. 223-228.) Sheckhtman, A.N. (1 9681 Ifhe pioibabilits' and intensits (it h
Glukbo%, V.G. (1972) Meteorological conditions for formation icing-up or ocean-going sess (iii Russian) . NaUchloi-issl-
of ice on htighlistiuctures (in Russian). (;idrometeoizdat, dlovatel'sk ii Insitut Aeroklini.itologii, Moscis, I tads -1
[rudy, vol. 3 1 1, GGO. 50, p. 5.5-65.
Hiananroter, B., F.]. Gagiion and B. Pratt (1980) Deicirig a Shellard, H.C. (1974) -file meteoirological aspects ot ice c
satellite conmunication antenna. CR RLL Special Report tlion on ships. World Meteotological Organiiaricin, Genes a,
80-18. A[)A 085397. Mlarine Science' Aftairs Rcport no. 10 WNID-No. ;971,
Houghton, lA.G. and W.H. Raeford (1938) (On the irreasU1re- Stallabrass, I .R. 119701i Methods for thle ailes anon or snip
1
ment of dirop si/c and liqusIiswiter content in fogs and icing. National Reseatrch Co rtcli., ttais ., Cinad,
Clouids. M.I.T. Paper, on PfltSsical (Jceanlogriaphs' and Mechanical F ogineccing Report MI3-51I
Metecoroilogy. Stallabrasss, I.R. and P.F . Hearts {11967( The icirg of \sjindcr,
jellinek, [1.11.6., H. Rdii, S. Kittaka, %I. lee and R. Yokota in conditions or sirriulated lice/tog sea SPiasf . N.Ational
(1978) Ice releasiitg block- cso nter coatings. Colloid Research Council, Oltasa, Canada, Mecchanical I ngiriefi i.
arid Prs./snier c ic(e, vol. 25 6, p. 544-5 Report MD-5O, NRC ito. 4-82.
Korond, W.C., R.J. Pili6. W.I. EFadie, E. I. Mack and IgI' 1 atata, T., S. Isvart aind N. Onio (I it6,) Studies roilt ice
Leonard ( 1971 ) Investigation o1 warm log pr~opert;v, isandl curtulation on sItips. I. I its 1,,rr;rria;rsi Si ;1 M
log modification conicepts. NANSA CR-I7731, Natinal .1, vol. 21, It. 173-221.
Aeronautics and Space Adm inistra,ion , Wash ingtoin, I).C. Estiated gluej 1,ce
Vattelniasui, P. arid I.I. 6, Incrit)IrI (1971)?SI
Kuroiwa, I). (1965) Icing and snose accretion on electric artd ssrod loads it t1C Cartlts SLurlace for tire ciiitigiUOUs
wires. CR RE L Resear ch Report 123. AD 611750. Uniited Stales. Ai IiFoce Caimtbridge Rewdiar cl Eiatwric'
Langmut~ii, 1. and K. Blodgett (1946) Mathemnatical insesriga- Reporrt Nit. ATICRII-R-73-0646, [.G. HlanScoir I feit.
tion of "sater droplet traj~ec tories. U.S. Arms Air Force Bedford, M'.
[echoical Retport 5t I8. N elckrlranrr. IA..iand 11.1. .int Karipe 15)Pis Sici pT,,P
McIay, 6.A. anl FLA. Thomptison (1969) Estimating the Cit.15 Of ctL lIUS clOULIS. 1(orinal i01 tIrOicI Lv 1(sli,
ha/ard of ice accretion in Canadamfromr climatological no. 3, p. 201I-21 1
damta. /outy~u/ of lpp/ietf lrti'ori/ois , sol. 8, no. 6, Woeodcock, A.. 553 Sill nuiclei in nmasrineai' as a iuniion
p. 927-93i. of altitide ind isiid fiorce, /.icrl 11Cit. i, ut!.

Minsk, I.D. 1 1977) s~eacctumulation Ott Ocean structuires. 10, no. .. ii t.

CRREL Report 77-17. ADA 0-14258.
A facsimile catalog card in Library of Congress MARC
format is reproduced below.

Minsk, L.D.
Icing on structures / by L.D. Minsk. Hanover, N.H.:
U.S. Cold Regions Research and Engineering Laboratory;
Springfield, Va.: available from National Technical In-
formation Service, 1980.
iv, 22 p., illus.; 28 cm. ( CRREL Report 80-31. )
Prepared for Office of Marine Geology, U.S. Geological
Survey by Corps of Engineers, U.S. Army Cold Regions Re-
search and Engineering Laboratory.
Bibliography: p. 17.
1. Deicing materials. 2. Deicing systems. 3. Ice.
4. Ice accretion. 5. Ice loads. 6. Ice prevention.
I. United States. Army. Corps of Engineers. II. Army
Cold Regions Research and Engineering Laboratory,
Hanover, N.H. III. Series: CRREL Report 80-31.
-U-S. GOVERNI"FNT PRINTING OFFICE 1981 700-660/309
/