Rain Erosion Aspects of Aircraft and Guided Missiles

A. A. F Y A L L * , B . S c , R. B. K I N G * and R. N . C. S T R A I N f , M . S c , Ph.D., F.R.I.C.

(*Royal Aircraft Establishment. 1[Now at National Gas Turbine Establishment)

SUMMARY: "All-weather" operational requirements have added considerably to aircraft and

missile design problems as the effects of various meteorological hazards, predominantly rain,
are aggravated by high speed flight. The problem of rain erosion is reviewed with regard to
the functional requirements of the components, the flight plan of the vehicle and the occurrence
of rain over the geographical terrain of operation. The factors affecting rain erosion charac-
teristics such as type of material, surface finish, shape of component, speed and rainfall
intensity are discussed and empirical data derived. The influences on the problem of radar
limitation in rain, the probability of rain occurrence and the effect of wind and wind-shear
are mentioned. Details given of equipment for the simulation of rain erosion are concerned
principally with the R.A.E. "whirling arm" and rocket runway high speed sled techniques.
Correlation of test results from these facilities and flight tests are briefly discussed.
Degree of erosion of a material is defined in relation to the particular application of the
material and details are given of the characteristics of numerous materials, both metals and
non-metals. The utilisation of these materials for such applications as radomes, transparencies,
high temperature materials, de-icing systems are briefly discussed together with methods of
extending the rain erosion "life" of materials by design or by use of protective coatings.

1. Introduction NOTATION
predominant drop size in air
The development of high speed aircraft and missiles has steady erosion rate
R
confronted the designer with many serious engineering diffi- velocity
V
culties. "All-weather" operational requirements have added threshold velocity
considerably to these problems as the effects of various meteoro-
logical hazards, such as hail and rain, are aggravated by high
speed flight. 2. General Survey of the Problem
The effect of rain during flight has previously raised such Although the problem is not an entirely new one, little is yet
problems as the removal of rain from vision panels and trans- known of the mechanisms of the process of erosion and con-
parent canopies and the prevention of ingress of water. With the sequently the conception of suitable materials to withstand
advent of high speed flying, a new phenomenon has been en- rain erosion is extremely difficult. As yet, no fundamental
countered in the erosion of paint coatings, structural plastic properties or parameters have been shown to be necessary for
components and even metallic parts by the impingement of rain erosion resistance. Consequently, attention has been devoted
raindrops on forward-facing surfaces. The damage may be so to the testing of a wide variety of materials and components and
severe as to affect the performance and eventually the air- most development work has proceeded on this ad hoc basis.
worthiness of the aircraft or missile. For many radar applications, Rain erosion resistance must now be considered as a design
even superficial damage to the housing or radome may render the factor for many present and all future aircraft and missiles, whose
navigational or search system ineffective. The problem therefore operational speeds will exceed 350 miles per hour at comparatively
is a very real one and, together with other high speed phenomena, low altitudes. Many components incur damage during flight
such as kinetic heating, constitutes a threat to the safe advance of through rain; it is believed that all forward-facing areas will be
high speed aviation. affected at high speed.
Historically, the observation of the phenomenon of rain A multiplicity of factors such as type of material, surface
erosion dates back to 1945, when U.S. Air Force personnel finish, shape of component and speed, influence erosion character-
reported the deterioration of an antenna installation on B-29 istics. Present results indicate that over a wide range of differing
aircraft. This "Eagle Wing" antenna was made of plastic materials, properties such as surface hardness, strength, tear
reinforced with glass fabric. Laboratory investigations establish- resistance (in the case of protective coatings), resilience or high
ed the fact that water, free from dust or other abrasive particles, elongation at break may be contributory factors to rain erosion
would cause erosion of glass fabric laminates. The phenomenon resistance. Prediction of erosion resistance at higher velocities by
had obviously been present at much lower speeds but had only extrapolation from phenomena observed at speeds up to 500
been witnessed as removal of paint or madapolam-dope schemes, m.p.h. may be totally unrealistic as the erosion mechanisms may
and had been attributed to other causes, such as erosion by sand change radically with increased speed. At the high rates of
or dust, or stone impact on take-off or landing. The severity of rain straining involved, the process may be expected to become
erosion damage has increased with higher flight speeds and the essentially a single impact failure, with every drop causing
extent of the erosion has been aggravated with the wide use of individual damage.
non-metallic materials in aircraft. Precise criteria for resistance may be evolved when funda-
mental studies reveal the exact nature of the high rates of strain-
ing involved and erosion resistance may then be correlated with
Received 14th March 1962 and based on a lecture given before the the physical properties of the materials measured under such
Society on TAth January 1961. loading conditions.
447

https://doi.org/10.1017/S0368393100076963 Published online by Cambridge University Press

 448 VOL. 66 JOURNAL OF THE ROYAL AERONAUTICAL SOCIETY JULY 1962

The final criteria will be modified by data accumulated over 3.2.2. With distance
many thousands of flying hours. In estimating the rain erosion Considerable meteorological data exists (1> regarding the
"life" of a component, much will depend on its functional horizontal extent of rainfall of different intensities. Heavy rain
requirements. The rate of erosion of a material varies pro- (0-5 to 1-0 in./h.) may be widespread. In the most intense phase
portionally with the 3rd to the 4th power of the velocity and thus of the tropical summer monsoon, rainfall of this intensity may be
guided weapon systems may expect to encounter greater erosion occurring virtually everywhere along the southern and western
than manned aircraft. On the other hand, the angle of impact slopes of the Himalayas and Yunnan Highlands for 1,500 miles
plays an important part in deciding the onset of erosion damage. or more.
From this aspect, the weapon system usually has small impact Heavy rain, perhaps over an area 200 miles across, would also
angles because of aerodynamic and power requirements and so be expected with tropical hurricanes.
this factor may alleviate the erosion problem. Again, the weapon Rainfall of higher intensity is thought to be limited to one or
system will have a definite short life, often of the order of two neighbouring convection cells, and is unlikely to be found
minutes, whereas the aircraft will require to be serviceable for in an area exceeding 10-20 miles across at any time. This aspect
many thousands of flying hours. Terrain to be covered, and may be of importance, however, in considering the flight of short
operational altitudes will also affect the problem. range weapons.
The rain erosion characteristics of materials must be assessed
for the specific purpose for which they are used. For example, a 4. Rain Erosion Requirements
glass-cloth laminate used as a structural member may have a 4.1. GENERAL ASPECTS
sufficient factor of safety to allow a large amount of erosion to be Rain erosion requirements may be based on the flight plan
permissible. The same type of laminate used as a skin for a and history of the aircraft or guided missile. The plan may often
radome or aerial installation may only suffer very slight damage be divided into distinct phases on the basis of near constant
before the radome becomes unsuitable for the transmission of the velocity. Each phase must then be assessed for the possibility
radar signal. Another example is afforded by the erosion of of rain occurrence commensurate with the operating altitudes of
transparencies for canopies and vision panels, where superficial the phase. The possibility of damage to individual components
damage may suffice to render vision impossible. is then assessed on the basis of time in rain at the various speeds
of each phase and on the angle-of-attack presented by the com-
3. The Characteristics of Rain for Simulation ponent. The angle-of-attack may often vary during flights,
especially with guided weapons which may yaw or pitch. Without
Studies the performance for many flying hours for individual compon-
3.1. DROP SIZE ents, present-day assessments are made on the basis that the
The commonest parameters of rainfall quoted by workers on component should survive the flights with an acceptable maxi-
natural precipitation phenomena were examined in relation to mum mal-functioning. In a guided weapon system, this is
their applicability to rain erosion. sufficient to achieve the weapon's purpose. In the aircraft case,
In view of the wide distribution of drop sizes in natural rain, should such conditions be encountered then repair or replace-
a particular condition must be typified by a characteristic drop ment may be necessary before the next flight. It is obvious, in the
size. It is simpler in practice to simulate the characteristic drop latter case, that systematic inspection must be made between
size for each rate of rainfall, than the complete distribution of flights of all aircraft parts which are susceptible to erosion.
drop sizes, although some methods of simulation do give a range
of drop sizes. 4.2. RADAR LIMITATION IN RAIN
One definition of drop-size in air is the predominant drop The assessment of the rain erosion characteristics required by
diameter (£>„') which is the diameter of the drops which any component is made on the assumption that the aircraft or
account for the greatest volume of water in the air. missile is capable of operating under these weather conditions.
It would seem logical to typify the rainfall to be simulated Although it is unlikely that weather conditions would impose
in rain erosion tests by the drop size which contributes the greatest severe limitations on the airworthiness of the vehicle, the
fraction of the total volume. Thus a compromise is made between efficiency of the radar system may be drastically reduced because
the greater individual effects of the large drops and the greater of the presence of precipitation particles in the form of raindrops,
number of impacts of the small drops. The predominant drop snow or hail. For example, the attenuation of A'-band energy in
size in air (D p ') has therefore been considered as the character- 2 in./h. rainfall is 1 • 25 dB/km. In cases where accurate guidance is
istic drop for any given rate of rainfall for use in simulated essential this reduction may be sufficient to cause mal-function or
erosion tests. For one in./h. intensity, Z>p' is 2-01 mm. non-operation.
The rain erosion requirements of a missile must therefore be
3.2. VARIATION OF INTENSITY compatible with the capability of the radar system in rain, and
flight plans are now assessed on this basis. The data is usually
3.2.1. With height reduced to the maximum time that the weapon can remain
Because of the lack of data on drop-size distributions above operational from the radar aspect in the maximum intensity of
the ground it is impossible to do other than speculate on the way rain in the particular geographical terrain to be covered by the
in which these distributions will vary with height. Many factors sortie. This time may differ greatly from that assessed on the
may operate to cause a change, the more important of which may basis of observed naturally-occurring precipitation over the
be coalescence of drops of different sizes owing to their different entire flight path.
rate of fall. High rates of rainfall are almost certainly associated
Some exploratory work at the Royal Radar Establishment
with cumulo-nimbus cloud activity which may extend to the
has shown that rain may have serious effects on the more
tropopause (1) . The height of the tropopause, therefore, will
sophisticated aspects of radar guidance, such as variation in the
determine the maximum operating ceiling for rain encounter if
Doppler frequency spectrum.
this is higher than the height of the -40°C isothermal, the latter
height defining the upper limit for the existence of water in the 4.3. PROBABILITY OF RAIN OCCURRENCE
liquid phase. In instances where a calculated risk may be taken on the
Similarly a chart of mean freezing level heights will give probability of occurrence of rain, an estimate may be made of the
the heights to which raindrops may be encountered in the rainfall conditions which must be simulated for rain erosion
gentler continuous rains up to 0-5 in./h. assessment.

https://doi.org/10.1017/S0368393100076963 Published online by Cambridge University Press

 FYALL, KING and STRAIN RAIN EROSION ASPECTS OF AIRCRAFT AND GUIDED MISSILES 449

A typical example is the flight over a 500 miles trajectory For a satisfactory test method, the following essential
starting from London. It may be deduced that there is a 0 • 01 per conditions must be simulated or controlled:
cent probability that for 277 miles out of the 500 miles flight it (a) Speed of specimen relative to the water drops.
would be necessary to simulate one in./h. rainfall. This does not (6) Shape of specimen or angle of strike of the drops.
mean that 277 miles of such intensity would occur naturally as the (c) Water drop-size and rate of rainfall.
estimate is made up from many discrete rain cells, but for Test methods can be divided into two categories in which the
simulation purposes this is a correct representation of the total raindrop or the specimen is stationary and the other is moving.
number of drops encountered. An apparatus in which water drops were projected at a
The statistical grouping of numbers of drops is, perforce, stationary sample would appear appropriate. However, the size
rather wide and consequently data for tropical centres of oper- of drops which are found in natural rain are such that they
ation may be such that the effect of the probability factor is shatter at relatively low speeds.
masked. Several techniques have been developed to project drops for a
short distance to strike the test sample before break-up occurs.
4.4. EFFECT OF WIND These have taken the form of interrupted water jets or special
In automatic guidance systems for missiles, it is important to guns, but their value is doubtful because of the uncertainty of the
consider the vertical structure of the wind' 3 '. A strong wind, shape, size and speed of the impacting water.
especially blowing over the take-off or launching site, can carry The most reliable methods are those which have a moving
the vehicle considerably off a controlled course. A gust or change test sample which impinges on water drops which are stationary
of speed or direction with height, that is, wind shear, can create a or nominally stationary with respect to the sample.
torque on the vehicle so as to put it out of control. There is, how-
ever, evidence to suggest that, despite the relatively low shatter 7.2. SINGLE DROP TECHNIQUE
velocity of raindrops (~50 ft./s.), precipitation is present in high A classic method for the study of impact by single drops has
winds. been developed by Jenkins'11' and others. A drop of water is
suspended on a vertical synthetic cobweb and is struck with a
4.5. ASSESSMENT projectile, the test specimen forming part of the projectile nose.
Summarising, it may be stated that each item or material must This technique has been used for speeds up to 3,700 ft./sec.
be assessed in relation to its function on the aircraft and to the
aircraft's role and field of operation. 7.3. WHIRLING ARM TECHNIQUE
For structural components, the rain erosion characteristics For the study of rain erosion under continuous impact a
may have to be assessed on the basis of "rain all the way" while "whirling arm" technique is most useful. The test samples are
for cases in which the guidance system is the predominant attached to the outer portion of an arm, or conventional pro-
feature, it may be assessed on a radar limitation or probability peller set at zero pitch, which is rotated in an area containing
of occurrence basis or, a combination of these two factors. simulated rain. The R.A.E. test rig' 7 ' is a tapered circular

5. Flight Testing
The evaluation of materials and components by flight in rain
is unsatisfactory because of the vagaries of weather, possible
hazards to personnel and the lack of suitable rain monitoring
equipment, particularly for the higher speed aircraft. Limited
flight testing(4) has been undertaken at the Royal Aircraft
Establishment at speeds up to 700 m.p.h. Missile flight tests are
also limited in value by the lack of detailed knowledge of rain
conditions.

6. Simulated Ground Tests

The bulk of rain erosion evaluation up to the present time has
been confined to classification of materials by simulated ground
tests'5-6- '• 8> 9>10). These necessarily employ standardised
rainfall conditions as natural rainfall can only be defined on a
broad basis of average intensity and drop-size. Under such
conditions, each material tested may be assigned a "life" at each
speed investigated.
The suitability of a material for a given project can then be
assessed from the flight plan and functional requirements of the
component, under the weather conditions defined from
meteorological data for the particular terrain and altitudes covered.
Thus, although the phenomenon of rain erosion has not been
placed on a precise quantitative basis, sufficient critical inform-
ation is now available to indicate the best choice of design or
material.

7. Methods of Simulating Rain Erosion

7.1. GENERAL ASPECTS OF THE PROBLEM
The process of rain erosion is not sufficiently understood to
warrant any test method which does not simulate flight con-
ditions as closely as possible. FIGURE 1. R.A.E. "Whirling arm" test rig.

https://doi.org/10.1017/S0368393100076963 Published online by Cambridge University Press

 450 VOL. 66 JOURNAL OF THE ROYAL AERONAUTICAL SOCIETY JULY 1962

section arm (4 ft. 9 in. radius) spinning in a vertical plane with

cylindrical samples attached near the tips at right angles to the
arm and perpendicular to the plane of rotation. In this rig, the
leading edges of the test sample are at constant velocity, unlike
horizontally rotating arms where a velocity gradient exists along
the length of the specimens. The arm is driven by a 500 horse-
power electric motor and Fig. 1 shows the installation.

7.4. SIMULATION OF RAIN

7.4.1. General aspects
As the size distribution of raindrops in natural rain varies
appreciably, a simulated rain must be in some way typical of the
particular rate of rainfall. Any attempt to produce a typical drop
size results in a range of drop sizes, and each method of producing FIGURE 2. Pendine sled with radome section attachment.
drops must be appraised by comparing this range with that of
natural rain. /ng of 110 of these jets relative to the track gave an intensity of
approximately 7 in./h. uniformly along its length. Studies at 500
7.4.2. Spinning disc technique for production of simulated rain m.p.h had shown that, with constant drop size, the rate of
Research into drop-producing techniques led to the adapt- erosion was proportional to the intensity. As the vehicle was to
ation of a spinning disc method* 12 '. Simulation of one in./h. rain spend an extremly short time in the rainfield, it was apparent that
intensity with drops 1-9-2-0 mm. diameter was achieved by repeated firings would be necessary to simulate the true flight
rotating a metal disc 2 ft. 4 in. in diameter at 180 rev./min. in a times of many missile systems. Thus, by using a high intensity
plane parallel to that of the arm and directing a controlled flow of rainfield, the effective time in rain of one in./h. intensity could be
water on to the disc centre. Suitable positioning of the disc substantially increased and correspondingly, the number of
relative to the arm enables the specimen to be rotated in a rainfall firings reduced. The rainfield has been shielded on both sides of
of controlled intensity. Over 50 per cent of the volume of drops the track and may be used even under fairly windy conditions.
are in the range of diameters 1 • 8-2 • 2 mm. and over 75 per cent in
the range 1 • 6-2 • 4 mm. Evaluation of rain erosion character- 8.4. METHODS OF SLED TEST
istics are generally made under standard conditions of one in./h. Various studies have shown that rain erosion characteristics
intensity rainfall, 2 mm. drop size and a speed of 500 m.p.h. at supersonic velocities may be influenced by the break-up of
Tests are also made under alternative conditions and speeds raindrops by associated Shockwaves. Thus, it is essential that the
which may be pertinent to particular problems. specimens and components which are tested should form the
Specimens are usually tested in pairs and, for uniform leading part of the vehicle. On the vehicles at present in use,
materials, good agreement in erosion characteristics have been facilities are available to test several materials simultaneously, or
obtained both between pairs and between duplicate tests. individual components, such as radome nose sections. Kinetic
heating may be simulated before firing. Fig. 2 shows the sled,
with radome attachment, before firing. Speeds of up to 1,700 ft./s.
8. High Speed Track Studies are achieved with this arrangement.
8.1. GENERAL ASPECTS
The major limitation in the use of whirling arms for rain 8.5. RESULTS OF HIGH SPEED TESTS
erosion testing is the high centrifugal force imposed on materials Preliminary results at supersonic speeds indicate that, with
at high speeds unless the arm is of very large diameter. As tests many materials, the phenomenon becomes a single impact
by the firing of missiles through natural rain have even greater process, with each raindrop causing individual damage.
limitations than those encountered using aircraft, facilities have
been developed for the simulation of high speed flight using a
track on which rocket-propelled sleds carry specimens and
9. Definition of Erosion
components through a simulated rainfield. Rain erosion resistance can be defined as the time under
The track is situated at the Proof and Experimental Establish- stated conditions of speed and rainfall, required to produce:
ment, Pendine, South Wales. The facility comprises a track of (a) Damage to the surface which can be detected under
one foot gauge, 3,000 ft. long, equipped with comprehensive defined viewing conditions.
communication, instrumentation and control systems. (b) A definite surface or structural condition, e.g. in the case
of a radome, it may be the time to remove sufficient of the
8.2. EQUIPMENT
material to seriously influence the radar performancs.
As rain erosion specimens must be recovered for examination Qualitatively, the progress of erosion can be witnessed as
and possible retesting on subsequent firings, a vehicle with two 5 follows:
in. motors for forward propulsion and one 5 in. motor for (a) Loss of the original texture or dulling of the surface.
retardation with an approximate total weight of 250 lb. is used. (b) Definite erosion or wear.
To assess rain erosion characteristics at near constant (c) Erosion to depth.
velocity, the positioning of the simulated rainfield was chosen after (d) Significant structural erosion.
study of the velocity-space curves for such vehicles. These showed At any one stage, deep erosion or pitting may occur, usually
that near constant velocity was achieved between 900 ft. and 1,400 at local defects which may lead to the cracking or tearing of the
ft. along the track and the simulated rainfield was consequently surrounding material by the hydraulic action of the raindrop
sited in this position. impact.

8.3. SIMULATED RAINFIELD 10. Qualitative Results of Whirling Arm Tests

Track to the extent of 500 ft. had to be covered by a rainfall 10.1. GENERAL ASPECTS
of fairly uniform intensity, and a Vee-type jet was finally chosen The bulk of rain erosion has, and will be, qualitative and
to give a characteristic 2 mm. diameter drop. Suitable position- subjective in nature. Results of such tests are well documented

https://doi.org/10.1017/S0368393100076963 Published online by Cambridge University Press

 FYALL, KING and STRAIN RAIN EROSION ASPECTS OF AIRCRAFT AND GUIDED MISSILES 451

and are most useful as a guide to many Service problems. Some values are given in Table III. Where electromagnetic trans-
materials have a wide variation in erosion properties: this is often mission characteristics are important, the coatings are, of
attributable to failure at local defects or poor manufacture. necessity, thin. Coatings seldom have a thickness greater than
Materials differ considerably in their resistance to erosion but 0-030 in. With the best materials, a thickness of 0-010 in. to
can be grouped into various broad classes. 0-012 in. is recommended but, with this thickness, the efficiency

10.2. HOMOGENEOUS MATERIALS AND LAMINATES TABLE II

Qualitative rain erosion test values at 500 m.p.h. and one EROSION RESISTANCE OF TYPICAL METALLIC MATERIALS AND PLATINGS
in./h. rain for various non-metallic materials are given in Table R.A.E. whirling arm rig at 500 m.p.h. and one in./h. rainfall
I. Under these conditions, nearly all materials erode to some
extent. It should be noted that unless identical erosion is quoted,
comparison between materials on a time basis may be invalid. Description min. Comments
Table II shows the erosion resistance of some metals and
platings. Stainless steel and brass both show excellent erosion Aluminium alloy D T D . 150 600 Very severe erosion
characteristics and although brass has been tested for a longer Aluminium alloy D T D . 900 Severe pitting on leading
period little differentiation may be made between brass, steel or 423B edge
the micrograin nickel sheath. Brass BS.250 2880 Deep pits locally
Copper BS. 1433 1100 Severe pitting, Eroded
10.3. PROTECTIVE COATINGS through 0-1 in.
The main function of these materials is to give protection to Magnesium alloy DTD.259 840 Severe pitting overall
Molybdenum 660 Eroded in layers to depth
plastics, particularly glass fabric reinforced resins. Typical
of 0-032 in.
Stainless steel BSS.521 1800 Visible erosion on leading
TABLE I edge, no weight loss
EROSION RESISTANCE OF TYPICAL NON-METALLIC MATERIALS Anodised aluminium 720 Superficial damage
R.A.E. Whirling arm rig at 500 m.p.h. and one in./h. rainfall Nickel plated aluminium 1200 Slight roughening
(polished)
Time (unpolished) 240-480 Variable—coating eroded
Description min. Comments through
Sintered molybdenum on 40 Eroded through coating
Paint (epoxy resin base) 7 Coating (0-005 in.) com- stainless steel (0 010 in.)
pletely removed. Micrograin electroform 1200 Undamaged. Test dis-
1 Glass fabric reinforced 5-15 Severe erosion with 6 to 8 nickel sheath bonded to continued
resin laminate layers removed aluminium alloy
ial

Nylon or Terylene rein- 5 Severe erosion with tearing

4> forced resin laminate of reinforcement TABLE III
Filament wound radome 5 Layer 0-025 in. thick
l\
a
section
Durestos 5
uniformly removed
Severe erosion particularly
TYPICAL EROSION TESTS ON PROTECTIVE COATINGS
R.A.E. whirling arm rig at 500 m.p.h. and one in./h. rain
u if resin starved Thickness Time to erode
Honeycomb (glass fibre/ 1 Eroded through outer skin Coating material Substrate
Pi in. x 1 0 3 through {min.)
polyester) radome section (0-030 in.)
i
Expanded Nitrile Ebonite Deep erosion depending on Neoprene to D T D . Glass fabric/ 11 25
*- the number and size of 856 polyester laminate
t/J cells (Goodyear 23/56) (poor quality)
"5 Borosilicate glass 5 Deep pits
<o (annealed) Neoprene to D T D . Glass fabric/ 11 100 to 150
Perspex (unplasticised) 20 Severe erosion 0-25 in. 856 polyester laminate
C deep (Goodyear 23/56) (good quality)
Tra

Quartz 6 Surface pitted

Polyurethane Glass/fabric 10 80*
VI r Alumina (Sintox) 4200 + Little or no erosion—Test polyester laminate
.3 discontinued Aluminium alloy 25 360*
•2 Steatite 290 Deeply pitted
Supramica 560 20 Severely eroded
a Pyroceram 9606 2580 Slight pitting only. Test
Maranyl (methyl Aluminium alloy 20 415
§< discontinued
methoxy nylon)
(Cast film stuck to
tempera

Sapphire 1320 N o visible damage. Test alloy)

discontinued
Boron nitride 40 Severe erosion to depth of Maranyl (sprayed Aluminium alloy 15 190
J= 0-25 in. from solution)
Silicon nitride 105 Deeply pitted
s Polyethylene, normal 270 Eroded to depth of 0-125 Flame sprayed nylon Aluminium alloy 22 30
grades in.
Polyethylene, high density 330 Eroded to depth of 0-125 Flame sprayed Steel or laminate 15 5
types in.
alumina
Polytetrafluoroethylene 25 Eroded to depth of 0-125
Epoxy resin Aluminium alloy 15 10
Polystyrene 15 Eroded to depth of 0-060 Aluminium alloy 75 50
Nylon 6 1080 Eroded to depth of 0-2 in. Hypalon Glass fabric/ 15 10 lAdhesion
Natural rubber 120 Eroded to depth of about polyester laminate 40 /failure
0-02 in.
Polycarbonate 70 Eroded through 0-15 in. *Little damage to film in general: local failure only.

https://doi.org/10.1017/S0368393100076963 Published online by Cambridge University Press

 452 VOL. 66 JOURNAL OF THE ROYAL AERONAUTICAL SOCIETY JULY 1962

of the coating (13> is more sensitive to local defects and to the TABLE IV
adhesion of the coating system to the substrate. RAIN EROSION FOR MATERIALS AT 500 M.P.H. AND ONE IN./H. RAIN
Goodyear Neoprene compound 23-56 is the usual standard (Cylindrical samples approximately 0 • 8 in. radius)
of comparison for new coatings. The excellent tear resistance of Erosion Erosion
certain polyurethanes may contribute to their good erosion Material rate Material rate
3 2
resistance and it must be concluded that they also exhibit this mm. /cm. h. mm.3jcm.2h.
desirable property at the high rates of shear which occur during
water drop impact. Perspex 786 Polyethylene. Hostalen 24
P.T.F.E. 336 Nylon 6 17
11. Gravimetric Assessment of Erosion Magnesium alloy 3-9
11.1. MEASUREMENT OF EROSION RATE
D.T.D.259
A more precise method of assessing the degree of erosion
other than by visual means is desirable if the effects of variation Polyethylene 33 Copper alloy B.S.1433 1-7
of material, rainfall and speed are to be studied in detail<9>. Alkathene 2
The weight lost by a sample of homogeneous material during Aluminium alloy 0-5
test is a measure of erosion. Because the samples on the R.A.E. D.T.D.423B
whirling arm are at right angles to the arm, erosion occurs
uniformly along the leading edge. where i?=rate in mm. 3 /cm. 2 h. and Kspeed inm.p.h. This means
Samples of various materials have been exposed on the test that below 208 m.p.h. Perspex will not erode.
rig over a range of speeds from 300 to 500 m.p.h. and in some There is insufficient evidence to show that this relationship
cases, in rainfall from 1 to 3 in./h., and the weight loss was between speed and rate of erosion is valid outside the speed range
measured at suitable intervals. Despite a wide variation in 300 to 500 m.p.h. and extrapolation to supersonic speeds cannot
erosion resistance, all these materials show an "initiation be justified, although the evidence available indicates that similar
period," during which the surface becomes visibly damaged but relationships may apply.
the weight loss is not measurable, followed by erosion at a
steady rate. 11.4. EFFECT OF ANGLE OF STRIKE ON EROSION
The smoothness of the surface is a factor in determining the If specimens in the form of wedges of different included
initiation period. The steady erosion rate is a function of the angles are tested at constant speed then there should be an angle
material and so if the area eroded during test is taken into below which no erosion takes place, i.e. where the normal
consideration then precise quantitative values for each material component of velocity is less than the threshold speed. This has
can be obtained. Table IV gives the erosion rates for different been clearly demonstrated for Perspex at 500 m.p.h. using wedges
materials. which give an angle of strike of 90° (flat specimen) 75, 60, 45 or
Stainless steel has been evaluated under these standard 30°<io>. These wedges tested for 15, 15, 30, 60 and 60 minutes
conditions but the weight loss is only very slight after 30 hours respectively suffered comparable erosion except for the wedge
and it would appear that the initiation period has not been giving an angle of strike of 30°, where only very slight erosion
passed. could be detected visually after testing for one hour. Here the
threshold velocity calculated from the normal component is
250 m.p.h. (500 (sin 30°) m.p.h.). This is close to the theoretical
11.2. EFFECT OF RAINFALL RATE ON EROSION
threshold velocity of 208 m.p.h. corresponding to an angle of
This method <9 > would appear to be a precise way of deter-
strike of 24 • 5°. Tests at or near this angle are not very practical
mining the effect of speed and rate of rainfall on erosion,
as it can be estimated that the time to achieve the same slight
provided the steady erosion rate is measured in each case.
erosion in one hour on the 30° sample would take 95 hours at
Unplasticised Perspex has been found to give very consistent
26°. It is therefore reasonable to conclude that the limited experi-
results and erodes in a short time, so has been used to evaluate
mental evidence does not contradict the assumption that it
these variables. When the rainfall rate is either 1, 2 or 3 in./h.
is the normal component of the velocity which is significant in the
(constant drop size distribution), the steady erosion rate is in the
initiation of erosion. If, however, erosion occurs, the local
ratio o f l - 0 : l ' 8 : 3 - 0 and so within experimental limits is directly
tangential component will become significant.
proportional to the rate of rainfall. These results are corrobor-
ated by work on other homogeneous materials.

11.3. EFFECT OF SPEED ON EROSION RATE

This is one of the most important aspects in the study of
erosion. The rate increases rapidly with increase in speed and it
has been suggested that the rate is proportional to a high power
of the velocity. Gravimetric loss measurements for unplasticised
Perspex at 300, 400 and 500 m.p.h. and one in./h. have been
made ( 9 ) . The steady volume erosion rate at these speeds is 16,
191 and 786mm. 3 /cm. 2 h. respectively. Similar experiments have
been done at 400 and 500 m.p.h. for Alkathene 2, and for copper.
Assuming that the steady erosion rate (R) varies as a power
of the velocity (V),
R=k(V-Vt)y (1)
where Vt is a "threshold velocity" below which no erosion will
occur and k and y are constants. If this is a true threshold
velocity it will be independent of rate of rainfall. For Perspex,
the general equation over the range 300-500 m.p.h. becomes FIGURE 3. (left) Typical erosion of polyethylene after 4i hours
test at 500 m.p.h. in 1 in./h. rainfall.
^=3-86xl0-6(^-208)3ii7 . . . (2) FIGURE 4. (right) Typical erosion of copper after 18 hours test
at 500 m.p.h. in 1 in./h. rainfall.

https://doi.org/10.1017/S0368393100076963 Published online by Cambridge University Press

 FYALL, KING and STRAIN RAIN EROSION ASPECTS OF AIRCRAFT AND GUIDED MISSILES 453

12. Correlation between Simulated Tests and is the higher. It may well be that the effect of drops less than one
mm. in diameter will be neglible, in which event the -20°C
Flight isothermal may be more realistic. In temperate zones, the
There is considerable evidence of erosion damage in flight. height of the freezing level (0°C isothermal) is sufficient to
In most cases the speed and time in rain are known but reliable determine the heights of rain occurrence.
information is seldom available concerning the rate of rainfall
encountered. In general, however, there does not appear to be 14. Conclusions
any flight evidence which is at variance with damage predicted
Experimentation has amply demonstrated the hazards of
from whirling arm results. Premature failures in flight can
rain erosion to many present-day and almost all future-gener-
invariably be traced to local defects or unsuitable material.
ation aircraft and missiles. Although little is yet known of the
While it is difficult to compare flight time with erosion
mechanisms of the phenomenon, erosion testing on an ad hoc
"life," damage occurring in flight is always similar in nature,
basis has supplied much information which will allow designers
extent and degree to the results of simulated tests.
to combat the problem by suitable choice of materials and design.
13. Choice of materials and design It is unfortunate that no fundamental parameters as yet exist
as criteria upon which new rain-erosion resistant materials may
For speeds in the region of 500 m.p.h. in rain and normal
be designed, but adequate test methods mean that all new
impact, the common alloys give good resistance, stainless steel
materials may be quickly assessed. It may well be that design
being excellent. Metals can be used to armour less resistant
criteria will emerge from the mass of data resulting from such
materials, e.g. electro-deposited nickel has shown good rain
tests before a satisfactory physical explanation of the phen-
erosion resistance and is a useful sheath material. Sintered
omenon has been achieved.
aluminas give excellent resistance and are most useful for
Mention must be made of the effect of hail impact as this is
electrical purposes and against the effects of kinetic heating.
obviously a most important consideration. Suffice it to say that
Polyethylenes have given long service on aircraft aerial masts.
the problem is more serious than that of rain erosion although
Brittle materials like Perspex and glass are not satisfactory except
hail occurrence is much less frequent and more restricted in
at low angles of impact. Glass fabric/resin laminates which may
extent than rain. Consequently, rain erosion must assume
be necessary for strength and electrical reasons also have poor
primary importance as a design requirement. Erosion resistant
resistance and often must have some protective layer. Porous
materials will help to minimise the hail problem and in some
or foamed materials have the least resistance of all.
cases may be adequate to afford complete protection.
With all protective systems, the nature of the substrate is
extremely important. Many coatings possessing intrinsically good Space flight will bring its particular erosion problems with
erosion characteristics often fail prematurely due to collapse of abrasion by inter-stellar dust and collision with meteors at much
the underlying material. This is particularly applicable to coat- higher velocities than those at present envisaged for rain erosion.
ings on reinforced resin laminates in which void-freedom and It is to be hoped that studies of rain impact will eventually
absence of resin starvation is absolutely essential. In special contribute fundamentally to all natural collision processes.
cases the representative specimens must include all items which
ACKNOWLEDGMENTS
will be present in the final scheme. For example, in de-icing
systems it must be ascertained that the heating applied in the The authors gratefully acknowledge Mr. R. F. Jones and
service operation of the system does not affect the rain erosion Dr. R. J. Murgatroyd of the Meteorological Office who supplied
characteristics. In the final evaluation of such a system, it is much of the meteorological information.
often possible to simulate typical heating for, say, 200 hours,
before testing the erosion resistance. REFERENCES
1. JONES, R. F. (1959). The Occurrence of Raindrops and Hail
In all cases, less damage is done if the angle of strike is small of Diameter 2 mm or Greater in the Atmosphere. A.R.C.
and all designs should take this into consideration. This is Paper 20, 765, 1959.
fortunately favourable for aerodynamic reasons. Armouring of 2. Use of Ground-Based Radar in Meteorology. World
vulnerable areas which have normal impact should be a design Meteorological Organization Technical Note No. 27.
3. SISSENWINE, N. (1954). Windspeed Profile, Windshear and
requirement wherever possible. Surface finish is of paramount Gusts for Design of Guidance Systems for Vertically Rising
importance and in many cases the benefits of reduction of impact Air Vehicles. Air Force Survey in Geophysics No. 57, 1954.
angle have been negated by imperfect surfaces. In the case of 4. BIGG, F. J., MCNAUGHTON, I. I. and METHVEN, T. J. (1956).
radomes, reduction of impact angle is usually achieved at the cost The Measurement of Rain from an Aircraft in Flight. Un-
published Ministry of Aviation Report.
of more stringent electrical requirements. Much depends on the 5. BEAL, J. L., LAPP, R. R. and WAHL, N. E. (1952). Final Report
functional requirements of the components to be protected. With on a Study of Rain Erosion of Plastics and Metals. Cornell
suppressed R.F. and V.H.F. aerial covers, electrical thickness Aeronautical Laboratory, Inc. Report No. PC-590-M-37. 1952.
considerations are usually unimportant whereas the protection of 6. STUTZMAN, R. H., LAPP, R. R. and WAHL, N. E. (1954). A
Study of the Rain Erosion of Plastics and Metals.
radomes, with strict aberration requirements, entails a correction 7. FYALL, A. A. and STRAIN, R. N. C. (1956). A "Whirling Arm"
or allowance being made for the applied coatings. With glazing Test Rig for the Assessment of the Rain Erosion of Materials.
materials, for items such as canopies and vision panels, only Unpublished Ministry of Aviation Report.
aerodynamic design can minimise the problem. The use of 8. FYALL, A. A., KING, R. B. and STRAIN, R. N. C. (1957). Rain
Erosion Part II—An Assessment of Various Materials. Un-
ceramic radomes to combat the combined requirements of published Ministry of Aviation Report
kinetic heating and constant dielectric properties may alleviate 9. FYALL, A. A., KING, R. B. and STRAIN, R. N. C. (1957). Rain
the rain erosion problem at high speeds. Even at much lower Erosion Part III—A Gravimetric Assessment of the Erosion
speeds, rain erosion must be considered as an important design Resistance of Various Materials. Unpublished Ministry of
parameter, as civil aircraft use more environmental aids, such Aviation Report.
10. KING, R. B. (I960). Rain Erosion Part IV—An Assessment of
as de-icing systems for leading edges, propellers, engine intakes Various Materials. Unpublished Ministry of Aviation Report.
and balance horns, and cloud and collision warning radomes. 11. JENKINS, D. C , BOOKER, J. D. and SWEED, J. W. (1958). Note
As has been previously mentioned, the flight plan or radar on a Laboratory Apparatus to study the High Speed Impact
between a Liquid Drop and a Surface. R. and M. Pub. 3203.
limitation of the aircraft or system will often decide the choice 12. WALTON, W. H. and PREWETT, (1947). Spinning Disc and
of material and design. For world-wide applications, an upper Spinning Top Sprayers for the Production of Homogeneous
ceiling for rain encounter is fixed by the height of the -40 C C Sprays and Mists. Porton Technical Paper No. 14, 1947.
isothermal (at which temperature all water has frozen) or the 13. SEWELL, J. H. (1958). The Evaluation of "Maranyl" Nylon
("Calaton") as a Rain Erosion Resistant Material. Unpub-
tropopause (being the limit of cumulo-nimbus activity), which ever lished Ministry of Aviation Report.

https://doi.org/10.1017/S0368393100076963 Published online by Cambridge University Press