A-A135 369 MECHANICAL ICE RELEASE PROCESSES I SELF-SHEDDING FROM I/)

HIGH-SPEED ROTORS(U COLO REGIONS RESEARCH AND

ENGINEERING LAB HANOVER NH K OTAGAKI OCT 83

NCLASSIFIED CRREL-83-26 F/G 8/12 NL

I EE

J -|! JL _'-lilA. .L..Q.. . I-.4400.

1.1 1k154

IIIIl 11111-4 111.6

MICROCOPY RESOLUTION TEST CHART

NATIONAL BUREAUl OP SAA*ASO- 193 A

!/' .3-

j .1

A A 115 3 1?.

©CIO1Z a:REPORT 83w26 US Army Corps

of EngineersCold Regions Research &Engineering Laboratory

Mechanical ice release processesISelf-shedding from high-speed rotors

.1U 234S6 78 9 3- 4 5 ' 67 a

-I -

DE .I

'-. DIm= TATEKNM JaA I ad i PinkUSS

For conversion of SI metric units to U.S./Britishcustomary units of measurement consult ASTMStandard E380, Metric Practice Guide, publishedby the American Society for Testing and Materi-als, 1916 Race St., Philadelphia, Pa. 19103.

Cover: Ice accreted on a high-speed rotor.(Scale is in centimeters.)

II

CRREL Report 83-26October 1983

Mechanical ice release processesISelf-shedding from high-speed rotors

K. Itagaki

Prepared lowOFFICE OF THE CHIEF OF ENGINEERSApproved for public release; distribution unlimited

I lng'oifiedSECURITY CLASSIFICATION OF THIS PAGE (Whefn Dae Entered

REPORT DOCUMENTATION PAGE B RED MsETINOS1. REPORT NUMBER 2. GOVT ACCESSION NO. I. RECIPIENT'S CATALOG NUMBER

CRREL Report 83-26 b A?_r I (94. TITLE (and Subtitle) S. TYPE OF REPORT &PERIOD COVERED

MECHANICAL ICE RELEASE PROCESSESI. Self-shedding from High-speed Rotors s. PERFORMING ORG. REPORT NUMBER

7. AUTHOR(s) I. CONTRACT OR GRANT NUMBER(s)

K. ltagaki

9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT. PROJECT, TASKAREA & WORK UNIT NUMBERS

U.S. Army Cold Regions Research and Engineering Laboratory DA Poject 4A61102AT24Hanover, New Hampshire 03755 Task C, Work Unit 002

11. CONTROLLING OFFICE NAME ANO ADDRESS 12. REPORT DATEOctober 1983

Office of the Chief of Engineers 13. NUMBER OF PAGES

Washington, D.C. 20314 13

14. MONITORING AGENCY NAME 6 ADORESS(i differmt faro Controlling Ofice) IS. SECURITY CLASS. (of this report)

Unclassified

ISo. DECL ASSIFICATION/ OOWNGRAOINGSCHEDULE

Is. DISTRIBUTiON STATEMENT (of this Repoti)

Approved for public release; distribution unlimited.

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

IS. SUPPLEMENTARY NOTES

Is. KEY WORDS (Continue on revere alde II necesary and identify by block number)

IceIce formationRotor blades

2L ANSRACT' (C . se vrge N . - .. eeMY 40d identify by block ,.mbet)

Ice accreted on high-speed rotors operating in supercooled fog can be thrown off by centrifugal force, creating severeunbalance and dangerous projectiles. A simple force balance analysis indicates that the strength of accreted ice and itsadhesive strength can be obtained by measuring the thickness of the accretion, the location of the separation, the rotorspeed,and the density. Such an analysis was applied to field and laboratory observations of self-shedding events. Theresults agree reasonably well with other observations.

DO F Ji W3 9Ot I , O MOW St SOOm.Ea Unclassified

SECUITI' CLASIWFICATIOW OF THIS PAGE ( n De Mete

PREFACE

This report was prepared by Dr. Kazuhiko Itagaki, Research Physicist, of the Snow and IceBranch, Research Division, U.S. Army Cold Regions Research and Engineering Laboratory.Funding for this research was provided by DA Project 4A161102AT24, Research in Snow, Iceand Frozen Ground, Task C, Research in Terrain and Cimatic Constraints, Work Unit 002, Adhe-

sion and Physics of Ice.The author thanks Dr. D. Sodhi and S.F. Ackley of CRREL for technical review of the manu-

script.

NT ' PA&I~DT!: T'!

Jh : - E

J

Dif',t ribut ionl/

Availability Codes

iAvnii and/or

Di-,t Special

MECHANICAL ICE RELEASE PROCESSESI. Self-shedding from High-speed Rotors

K. Itagaki

INTRODUCTIONdangerous to operate. One of the major causes of

Ice adhesion and its control have been studied danger is the self-shedding of ice built up on the rotor.through various approaches. Although many experi- When ice is shed from one blade of a rotor, a heli-mental studies stressed careful preparation of the copter may become uncontrollably unbalanced, andsubstrate surfaces and ice, the results always had con. dangerous high-speed projectiles may damage air-siderable scatter. Efforts were concentrated on mak- frames or nearby structures. Many factors are in.ing the results reproducible by refining the alignment volved in the self-shedding process: the balance ofand procedure of surface preparation. Since most of forces, the initiation and propagation of a local frac-the studies used a chemical approach, little consider- ture, and the eventual release of accreted ice. Thisation was given to the fracture aspects of the ice, analysis is limited to the examination of the balancethe strength of the ice/substrate interface, or the de- of forces. Such simple analysis, however, leads us toviation from pure shear or tensile conditions. Pure a new method of measuring adhesive and tensiletensile tests, for instance, cannot be carried out with- strength of accreted ice simultaneously.in the limit of finite dimensions because the differ-ence between the elastic moduli of the ice and thesubstrate creates a peeling effect at the edge of the THEORETICAL ANALYSISice-substrate bond. The weakest point in the bondsystem dictates the strength of the entire system. The balance of force for a simple case is shown

Most of the theoretical studies concentrated on in Figure 1. A rectangular bar of components A andcontrol of the surface energy. Initiation and propa- B is bonded at an interface having cross-sectionalgation of a crack leading to the final catastrophic area Sc perpendicular to the radial direction. Asfracture were of little concern. A vague notion that this bar is rotated around an axis located at the endstronger bonding will result in higher adhesive strength of component A, the centrifugal force F, acting onseems to prevail. However, as I discussed in a recent Se isreport (Itagaki 1983) the surface energy only indi- R

rectly controls the adhesive strength of ice through F: f iw2 p S, dl (1)surface contamination. On the very clean surface, I=rice tends to break before the bond fails.

In this paper, I will use theoretical and experi- where r distance from the axis to the interfacemental analyses to examine one "real-world" process w angular velocityof mechanical ice removal, the self-shedding of ice R radius of the rotor.from rotating struts.

Under icing conditions, high-speed rotors, such Component B will break away if Ft ; A t - S,, whereas aircraft propellers, main and taf rotors of helicop- A t is the tensile adhesive strength between substratesters, and turbines of wind energy generators, can be A and B. If component B is accreted ice having

M EN =. . . .. .. ' ,. . .. i - -, ,. . .

h outboard side of S exceeds the supporting force

b available from the combination of the adhesivestrength A, .S. and tensile strength T.S, where A,is the shear adhesive strength between the accreted

Accrete Substrate ice and the substrate, S, is the area of contact be-e Ice tween the ice and the rotor outboard of the cross

section S, and T and Sc are the tensile strength andS R the cross-sectional area of the accreted ice, respec-

Stively. The difference between the supporting forces

and the centrifugal force Fc is a supporting forcemargin M:

AM = (TSc + As.S ) -Fc. (4)

Shedding will occur when M becomes negative.- ,If the accreted ice has uniform width b and thick-

- - - w 0 ness h, then

Figure I. Schematic dia- Figure 2. Ice accreted on the M = Th + A, (R -r)b - (RI - r2) w 2 pbh/2

gram of a composite rotor rotor substrate. (5)having joint at S

where Tbh is the tensile supporting force andA,(R -r)b is the shear adhesive supporting force.The relationship between M and r for different thick-

uniform thickness h, width b and density p, eq 1 nesses is shown in Figure 3. The curves are a set of

can be simplified to parabolas having nodes at r = ± ,/R2 - 2T/o2p,M = A, (R - r). As the thickness h increases, the

F 2pbh (R 2 -r2 ) • A, S = Atbh. (2) part of the curves between the nodes lowers, and2 M decreases and eventually becomes negative. The

self-shedding most likely begins when the supportingWhen ice accretes on a rotor, the geometry is as force margin becomes zero. If M = 0, then eq 5 be-

shown in Figure 2. The ice accreted on the substratecoe

on the outboard side of imaginary cross section S is

supported by a combination of the shear adhesive A, 1A2 -2hW2p [AR -h(R2o2p/2-7)]force F. and the tensile force Ft. The accreted ice r =

will separate at S when the centrifugal force on the hw 2p(6)

40I I I

30

20 5mmM

(hPa)

r (cm)

Figure 3. Radial variation of supporting force margin with thickness.*1 2

The solution becomes real when the curve is tan- Solving this equation for T, we obtain the tensilegential to the abscissa, that is, when the two roots strength of accreted ice by knowing the density ofmeet; therefore, the square root term in eq 6 becomes accreted ice p, the angular velocity w, the radius ofzero: the rotor R, and the location of the failure surface r:

A2 - 2hW2pAR + 2h 2 W2 p (R 2 ca2p/2-7) = O. T = 0 2 p R - r) 2 (10)2

(7) Since we can measure the thickness where the failurestarted, we can calculate the adhesive strength of the

From eq 7 the ice thickness h is accreted ice from eq 9:

wpR ±11 Th~p (g2,, p - 27)h = /7p A.(8) As = (11)R~~p2T

The accreked ice will most likely start to shed at thisthickness. Since the square root term in eq 6 van- CALCULATED TENSILE AND ADHESIVEishes at this thickness, eq 6 can be simplified by using SHEAR STRENGTH FROM EXISTINGeq 8: EXPERIMENTAL RESULTS

r=R2 2 p- 2T The analysis described here is based on the assump-(coopR ± 1/2p7')o tion that the accreted ice has a uniform cross-sectional± 1p area along the rotor. This assumption is justified

R +(9) under certain conditions. Near the transition from

WP dry- to wet-growth regime the thickness of the

The negative sign of the square root should be used accreted ice across the failed surface and the thickness

to keep r 4 R. Note that no adhesive force term is of the ice on the other side of the rotor blade are

involved in this equation. reasonably uniform (Fig. 4). Smooth, crescent-

0 a. Perpendicular to the radial direction.

b. Parallel to the radial direction.

Figure 4. Thickness of accreted ice (shown between dashed lines). The thickness

is uniform slightly below the wet- and dry-growth regime boundary.~3

Figure 5. Crescent-shaped cross Figure 6. Rough, icicle-covered accretion grown in the wet-growth regime.

section of accreted ice grown underdry-growth conditions.

shaped growth is usually observed in the pure, dry- tensile and adhesive strengths calculated from thegrowth regime (Fig. 5). Rough, icicle-covered sur- data using eq 10 and II disregarding the thicknessfaces are formed in the wet-growth regime (when the variation are summarized in Table 1. Self-sheddingtemperatures are at or near the freezing point) (Fig. observed during field tests made at the summit of Mt.6). Numerical integration using eq 1 would be needed Washington (Itagaki et al. 1983) is also included.to obtain accurate tensile and adhesive strength val- Experiments 1, 2, 3, 10 and 11 had cross sectionsues when the surface conditions are not uniform. similar to those shown in Figure 4, so eqs 10 and 11

During laboratory icing experiments on the high- can be used. However, except for experiment 7 thespeed radar (Ackley et al. 1979) several self-shedding data agree reasonably well within the variation ex-events were observed. The observed data and the pected from experimental error. In experiment 7

Table 1. Calculated tende and adhesive strenibj of accreted ice.

Shedding RotorExp. Exp. Temp. Elapsed Density radius Thickness Growth rute speed T Asno. code Date (00 time (S) (S/cm3) (cm) (0puJ (mm/a) (prm) (kPa) (kPa) Shape of accretion

Laboratory experiments, R = 25 cm (Ackley et aL 1979)I T-12-11 8/26/77 -11.1 105 0.91 15.55 0.346 0.0329 3600 77.5 69.9 Uniform2 T-14-77 9/7/77 -13.3 123 0.90 16.5 0.374 0.0304 3600 462.1 78.1 Uniform but curved

fracture surface3 T-16-77 9/9/77 -15.5 15O 0.90 15.5 0.596 0.0397 3600 577.2 118.2 Uniform4 T-4-77-2 10/28/77 -1.0 123 0.92 10.8 0.21 0.0171 3600 1318.2 29.7 Irregular

T "-6-77-2 11/7/77 -5.0 114 0.92 15.4 0.399 0.035 3600 602.S 80.3 Irregular6 -8-77.1 11/8/77 -5.0 161 0.92 17.1 0.S2 0.0323 3600 408.0 116.3 Irregular7 TR-8250 9/10/78 -4.0 300 0.92 24.6 0.34 0.0113 1800 0.26 27.3 lregular; only

tip shedI TR-9303 10/30/79 -14.5 "61 0.9 6.9 O.S 0.0157 3600 2188.s 41.6 Crescent

Natuzd icbq at Mt. Wadhistou, R - 75 an (Itapgi et aL 1963)

9 14 Jan 1/14180 -5.0 470 0.9 49.2 1.0 0.0212 1800 1064 157.3 Uniform on801 lB the aitoil blade

10 14 Jan 1/14/80 -5.0 470 0.9 42.5 0.42 0.0089 1800 1680 S7.1 Uniform on

Sol PC peased cylinder11 14Jan 1/14/80 -3.0 350 0.9 50.3 0.38 0.0108 1800 975 61.1 Uniform

802 PC

4

rb l . . ..

2.5xI0 I O0Ci

e Uniform and irregular Laboratorym Crescent2 1800 rpm JExperiments

2.0- o mt. Washington

-. 150 -

0

1.5£ C

w,00

oo

50

0.5 -o Uniform Laoatr aand Trregular Lorty •

* C resc ExperimentsA* 1800 1 0

a Mt. Washington

0 0.01 0.02 0.03 0.04 0.01 0.02 0.03 0.04

Growth Rot* (rm/s) Growth Rate (rm/s)

Figure 7. Growth rate vs tensile strength of accreted Figure 8. Growth rate vs adhesive shear strength. Fast-ice. Slower-grown ice is stronger. er-grown ice has a higher shear adhesive strength.

the ice shed at the very tip of the rotor, so the re- rates (Fig. 8). If we place the Mt. Washington tests

sults are questionable. in a separate group, the tendency becomes moreThe tensile strengths of accreted ice at slow growth pronounced in both groups. Tensile and adhesive

rates were about 1-2 MPa, while those of more rapidly shear strengths show little dependence on temperaturgrown ice were 0.5-0.6 MPa (Fig. 7). However, ad- (Figs. 9 and 10). The shear adhesive strength valueshesive shear strength was greater at faster growth calculated from the thickness, the location of separa-

2.5xl0 3

2.00

and Uniform and Irregulara Crescent Labpeamort 0A 1800 rpm Experiments

E 1 -W an.tWoshington

0.5-

III I I_00

-20 -15 -0 -5 0

Temperture (*C)

F. r v 9. empetre vS tensile strengt u rei sho little resdonahip.

ice Sowe-gow ic i stoner ergrwn ceha a iger her dheiv sWenth

200 I

9 Uniform and Irregulor Laboratorya A a1800 rpm Experiments

- 150-a Mt. Washington

100-

0*3 0

50-

-20 -15 -to -5 0

Temperature (*C)

Figure 10. Temperature vs shear adhesive strength. This also shows littlerelationship.

tion and the density were within the range of 30-150 0.05-pm alumina. Ford and Nichols (1961, 1962)kPa, which roughly agrees with the values obtained by give even lower values of 750 kPa on stainless steelvarious researchers on slightly contaminated, smooth- polished by alumina of unknown grade.mirror-finished metal surfaces. Jellinek (1957a, b, 1960) studied the effects of

sample size, surface roughness and temperature ex-tensively. However, the substrates were immersed

COMPARISON WITH PREVIOUS in benzene before testing, which presumably lowersEXPERIMENTAL STUDIES the adhesive strength (Raraty and Tabor 1958). His

results were generally lower than those of the otherBeams et al. (1955) were probably the first to use researchers.

the centrifugal force field on the high-speed rotor as a Yano and Kuroiwa (1979) were apparently thetool for studying adhesive and tensile strength. When first to measure the adhesive tensile strength of icea uniform, continuous film is deposited on an ultra- on a metal substrate by using the centrifugal forcehigh-speed rotor, a combination of hoop tensile field from a rotor with a speed much lower than thatstress and tensile adhesive strength supports the ad- used by Beams et al. (1955). Their system has a well-hered film on the rotor against the centrifugal force, defined geometry similar to Figure 1, but it cannotwhile only tensile adhesive strength supports the measure tensile strength. They measured tensile ad-discontinuous film removed from the rotor. They hesive strength of pure and KCI-doped polycrystal-did not apply this method to ice adhesion problems. line ice on a polished but not mirror-finished brass

Many researchers have reported on adhesion of surface. Their results at ambient temperatures higherpure and doped ice on various substances with a va- than -10*C scattered between 50 and 900 kPa. Asriety of surface treatments and coatings. The results the temperature dropped, the adhesive strength rapid-ranged from an extremely low 50 mPa, attributable ly increased to 2000-3000 kPa. Within this tempera-to viscous sliding on a smooth, oil-coated surface ture range the breaks were mostly cohesive. Since(Baker et al. 1962), to high values of 2734 kPa (Jones the conditions of their testing are different from myand Gardos 1972), which should be regarded as a studies, direct comparisons of the results are difficult.confined shear test of ice itself. Raraty and Tabor Compared with the melt-grown ice usually used(1958) gave another very high value of 1960 kPa for studies like those described above, accreted iceat -10C on stainless steel polished by #600 Carbor- may have different tensile and shear adhesive strengths,undum and cleaned with an HNO 3 -alcohol mixture, since it has a strongly oriented structure. The fol-Bascom et al. (1969) obtained slightly lower values lowing studies were made by using accreted ice.of 930 kPa by using stainless steel polished with Kuroiwa (1951) studied ice accretions in natural

6

JI

conditions at the top of Mt. Nisekoan-nupuri using Their conditions were closer to my analysis than mosta 270-cm-diameter metal propeller. The ice did not of the other studies. They tested five materials com-accrete beyond a certain radius of the propeller; monly used in aircraft structures: aluminum, stainlessthis radius was proportional to the ambient temper- steel, titanium, Teflon and Viton. Teflon had theature. If the radius of the ice accretion limit was highest adhesive strength above -100 C. Stainlessthe radius where self-shedding occurs, the tensile steel was the lowest throughout the measured range.and adhesive strength of ice can be estimated from Generally the trend was for a stronger adhesivethe photographs and drawings in this report. The strength at lower temperatures. This trend does notresults at -10°C of a tensile stength of 1960 kPa and agree with my results. Their results depend slightlyan adhesive strength of 320 kPa are within the rea- on the substrate materials but ranged from 30 to 70sonable range. At -5°C the tensile strength calcu- kPa at -6.6 0 C, which is comparable with my results.lated from his Figure 10 resulted in an unreasonably Laforte et al. (1983) measured the adhesivehigh value of 17,600 kPa, either the assumption of strength of accreted ice. Their measurements in-a uniform cross section and thickness cannot be ap- volved uniform accretion around an aluminum cyl-plicable or some other mechanism, such as vibration inder at various wind speeds and temperatures. Theyof the blade or aerodynamic heating, may be causing measured the force required to pull the cylinder awaythe shedding. from the accreted ice. They found considerabl de-

One of Kuroiwa's observations is that the radius pendence of the growth conditions on the adhesiveof the limit of ice accretion seldom exceeded I m, strength. The mode of failure of separation waseven though the ambient temperature dropped be- mostly adhesive. Although the accreted ice in mylow -10C. The temperature rise from aerodynamic studies is grown at a far higher wind speed than theirs,heating due to friction at I m is I 5.8'C, and that their highest adhesive strengths (which ranged fromfrom adiabatic compression is 17.8 0C, according to about 40 kPa at -20 C to about 180 kPa at -0 0 C)the theoretical calculations referred to in his paper at the highest wind speeds are comparable to my re-

A0oI = 3.43x 10- 5 . 2°C (km/hr) 2 for frictional suits of 30-150 kPa.heatingand AO2 = 3.87x 10- 5 UP°C (km/hr) 2 foradiabatic compression (Pohlhausen 1921)] . Eitherof these heating mechanisms would be more than DISCUSSIONsufficient to bring the temperature above the freez-ing point. The results of the analysis show an interesting re-

The speed of the tip of the rotor used in the Mt. suit of the relationship between the tensile and ad-Washington measurements was 141 m/s, and the cor- hesive shear strengths and the growth rate. The higherresponding temperature rises from frictional and adi- tensile strength at slower growth rates indicates thatabatic heating were calculated to be 8.90 and I0.0°C, some strengthening process, such as sintering, anneal-respectively, using the Pohlhausen equation. However, ing or recrystallization, is involved during ice accre-I observed ice accretion at ambient temperatures as tion. Since no systematic difference was detected inhigh as -20 C. The temperature rise measured on the the densities of fast- and slow-grown accreted ice,surface of the leading edge at the tip was 2-2.50 C, far the density vs strength relationship observed by La-lower than the calculated values. My measurements forte et al. (1983) is unlikely to be the cause of themade on top of Mt. Washington indicated a rise of less difference.than 3.0°C under the dry conditions at 0.75 in from the The appearance of the fractured surfaces indicatedcenter. Under these conditions the heating can be re- that the grain sizes of slow-grown ice were smallergarded as purely aerodynamic. Extrapolated to I m, than the fast-grown ice. If the fracture proceedsthis temperature rise would be about 5.30C. Apparent- within the grain, the Petch effect may explain thely the calculations using the Pohlausen equation are too greater strength in the ice with smaller grain size.high. On the other hand the thickness at which self- According to Shuison et al. (1982) the tensile strengthshedding would occur at I m was calculated to be 5.4 a, which may correspond to our tensile strength T,mm by assuming that eqs 10 and I I applied and that of randomly oriented ice at -10*C is related to thethe tensile and shear adhesive strengths are the medi- grain size d by the relationshipan values of 500 and 200 kPa, respectively. This thick-ness is quite reasonable, so Kuroiwa's observation may at = Oi + k-d- (12)be caused by self-shedding.

Stallabrass and Price (1962) used centrifugal force where o = 600 kPa and k = 20 kPa-m . If the grainon accreted ice to measure shear adhesive strength. size for faster growth is 2 mm and the grain size for

. . .. -l,

slower growth is 0.2 mm, the tensile strength is cal- istics of ice in bulk, at the ice/solid interfaces and at

culated to be 1050 kPa and 2010 kPa, respectively, ice/lubricant/solid interfaces in a laboratory device.for randomly oriented freshwater ice. The tendency Naval Research Laboratory Report 5662,

seems to agree with my results but my measured Ford, T.F. and O.D. Nichols (1962) Adhesion-shearvalues were one half of their values. This difference strength of ice frozen to clean and lubricated sur-may be due to the difference in crystal structure, faces. Naval Research Laboratory Report 5832.The accreted ice is mostly columnar in the temper- Itagki, K. (1983) Adhesion of ice to polymers andature range of the experiments; the tensile force was other surfaces. In Physicochemical Aspects of Polymerapplied perpendicularly to the direction of column Surfaces (K.L. Mittal, Ed.). Vol. 1. New York:

growth, so the strength may be lower than for ran- Plenum Publishing Corporation, pp. 241 -252.

domly oriented crystals. Italaki, K., G.E. Lemieux, H.W. Bosworth, J. O'KeefeI have little explanation to offer on the grain size and G. Hogan (1983) Studies of high-speed rotor

vs adhesive shear strength relationship or the depend- icing under natural conditions. Proceedings of Firstence on temperature. However, self-shedding is a International Workshop on Atmospheric Icing of

dynamic process and little time was spent between Structures (L.D. Minsk, Ed.). USA Cold Regionsthe formation of ice by accretion and shedding, u. Research and Engineering Laboratory, Special Re-like the other studies. Droplets hitting the substrate port 83-17, pp. 117-124.at high speed tend to clean the surface; higher depo- Jellinek, H.H.G. (1957a) Tensile strength propertiessition rates may also result in cleaner substrate sur- of ice adhering to stainless steel. USA Cold Regionsfaces, which may give a higher adhesive shear strength. Research and Engineering Laboratory, Research Re-To prove or disprove such speculation and to under- port 23. AD-716 663.stand the accretion problem more clearly, further Jellinek, H.H.G. (1957b) Adhesive properties of ice.experimental studies are needed. USA Cold Regions Research and Engineering Labora-

tory, Research Report 38. AD-149 061.JeUinek, H.H.G. (1960) Adhesive properties of ice,

CONCLUSIONS Part il. USA Cold Regions Research and EngineeringLaboratory, Research Report 62. AD-638 344.

Simultaneous measurements of tensile and adhe- Jones, J.R. and M.N. Gardos (1972) Adhesion shearsive strength of accreted ice can be made in situ strength of ice to bonded solid lubricant. Lubrica-simply by measuring the radius of the rotor, the lo- tion Engineering, 28: 464-461.cation of failure surface, the ice thickness, the den- Kuroiwa, D. (1951) Icing of aeroplane propeller ofsity of the accreted ice, and the rotor speed. The real size. Low Temperature Science, 6: 11-22.results agree with previous measurements of adhe- Laforte, J.L., C.L. Phan, B. Felin and R. Martin (1983)sive and tensile ice strength. This method can be Adhesion of ice on aluminum conductor and crystalused to analyze existing data to obtain little-studied size in the surface layer. Proceedings of First Inter-values of shear adhesive and tensile strengths of national Workshop on Atmospheric Icing of Struc-accreted ice. Some new measuring methods based tures (L.D. Minsk, Ed.). USA Cold Regions Researchon this study can also be developed, and Engineering Laboratory, Special Report 83-17,

pp. 8 3-9 2 .Pohlhausen, V.E. (1921) Der Warmeaustaush zwischen

LITERATURE CITIED fester Korpemand Fliisigkeiten mit kleinne Reiburgand klemen Warmeleiting. Zeitschriftfur Angewandte

Addey, S.F., G.E. Lemieux, K. Itapki and J. OKeefe ?Aithematik und Physik, 1: 115.(1979) Laboratory experiments on icing and rotating Rraty, L.E. and D. Tabor (1958) The adhesion andblades. Snow Removal and Ice Control Research, strength properties of ice. Proceedings of the RoyalNational Academy of Sciences, Washington, D.C., Society of London, 245A: 184-201.Special Report 185, pp. 85-92. Shulson, E.M., J.H. Currier and P. Lim (1982) ABaker, H.R., W.D. Bacom and C.R. Singletery (1962) grain size effect on the tensile strength of polycrystal-The adhesion of ice to lubricated surfaces. Journal line ice. Sixth International Symposium on theof Colloid Science, 17: 477-491. Physics and Chemistry of Ice, p. 29.Bacom, W.D., ELL. Cottinton sad C.R. Singleteny Stalhbrass, J.R. and E.D. Price (1962) On the ad-(1969) Ice adhesion to hydrophilic and hydrophobic hesion of ice to various materials. National Researchsurfaces. Journal of Adhesion, 1: 246-263. Council of Canada, Aeronautical Report LR.350.Reams, J.W., J.B. Brezeale and W.L. Bart (1955) Yao, K. and D. Kurowa (1979) Adhesion strengthMechanical strength of thin film of metals. Physical of contaminated ice. National Academy of Sciences,Review, 100: 1657-1661. Snow Removal and Ice Control Research, Special Re-Ford, T.F. and O.D. Nichols (1961) Shear character- port 185, pp. 30-40.

t ... _ .8

A facsimile catalog card in Library of Congress MARCformat is reproduced below.

Itagaki, K.Mechanical ice release processes. I: Self-shedding

from high-speed rotors / by K. Itagaki. Hanover, N.H.:Cold Regions Research and Engineering Laboratory;Springfield, Va.: available from National TechnicalInformation Service, 1983.

ii, 13 p., illus.; 28 cm. ( CRREL Report 83-26. )Prepared for Office of the Chief of Engineers

by Corps of Engineers, U.S. Army Cold Regions Research

and Engineering Laboratory under DA Project 4A161102AT24.

Bibliography: p. 8.1. Ice. 2. Ice formation. 3. Rotor blades.

I. United States. Army. Corps of Engineers.II. Cold Regions Research and Engineering Labora-

tory, Hanover, N.H. III. Series: CRREL Report 83-26.

I